Water Sensitive Planning and Design PDF

HANDBOOK OF

WATER SENSITIVE
PLANNING and DESIGN

L1562 FMFrame Page 2 Tuesday, April 9, 2002 12:53 PM

HANDBOOK OF
WATER SENSITIVE
PLANNING and DESIGN
Edited by
Robert L. France, B.Sc., M.Sc., Ph.D.

Associate Professor of Landscape Ecology
Science Director of the
Center for Technology and Environment (CTE)
Graduate School of Design
Harvard University
Cambridge, Massachusetts
LEWIS PUBLISHERS
A CRC Press Company
Boca Raton London New York Washington, D.C.
Library of Congress Cataloging-in-Publication Data

Handbook of water sensitive planning and design / edited by Robert L. France.
p. cm. (Integrative studies in water management and land development)
Includes bibliographical references and index
ISBN 1-56670-562-2 (alk. paper)
1. RunoffManagementHandbooks, manuals, etc. 2. Landscape ecologyHandbooks,
manuals, etc. 3. Ecological landscape designHandbooks, manuals, etc. 4. Water in
landscape architectureHandbooks, manuals, etc. 5. Buffer zones (Ecosystem
management)Handbooks, manuals, etc. 6. Watershed managementHandbooks,
manuals, etc. I. France, R. L. (Robert Lawrence) II. Series.
TD657 .H36 2002
333.91dc21
2002016075
This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with
permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish
reliable data and information, but the authors and the publisher cannot assume responsibility for the validity of all materials
or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior
permission in writing from the publisher.
All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or internal use of specific
clients, may be granted by CRC Press LLC, provided that $1.50 per page photocopied is paid directly to Copyright clearance
Center, 222 Rosewood Drive, Danvers, MA 01923 USA The fee code for users of the Transactional Reporting Service is
ISBN 1-56670-562-2/02/$0.00+$1.50. The fee is subject to change without notice. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.
The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating new works,
or for resale. Specific permission must be obtained in writing from CRC Press LLC for such copying.
Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation, without intent to infringe.
Visit the CRC Press Web site at www.crcpress.com

2002 by CRC Press LLC
Lewis Publishers is an imprint of CRC Press LLC
No claim to original U.S. Government works
International Standard Book Number 1-56670-562-2
Library of Congress Card Number 2002016075
Printed in the United States of America 1 2 3 4 5 6 7 8 9 0
Printed on acid-free paper
Series Statement
Integrative studies in water management and land development
Ecological issues and environmental problems have become exceedingly complex. Today,
it is hubris to suppose that any single discipline can provide all the solutions for protecting
and restoring ecological integrity. We have entered an age where professional humility is
the only operational means for approaching environmental understanding and prediction.
As a result, socially acceptable and sustainable solutions must be both imaginative and
integrative in scope; in other words, garnered through combining insights gleaned from
various specialized disciplines, expressed and examined together.
The purpose of the series Integrative Studies in Water Management and Land Development
by Lewis Publishers, part of CRC Press, is to produce a set of books that transcend the
disciplines of science and engineering alone. Instead, these efforts will be truly integrative
in their incorporation of additional elements from landscape architecture, land-use planning, economics, education, and environmental management, history, and art. The emphasis of the series will be on the breadth of study approach, coupled with the depth of
intellectual vigor required for the investigations undertaken.
Dr. Robert L. France
Series Editor
Harvard University
Foreword
One can first of all simply wonder at the fact that it is only recently that humanity has
begun to ponder the evolution and fate of water in the world when the very survival of
our species depends on it. So begins the report from the UNESCO World Commission
on the Ethics of Scientific Knowledge and Technology (The Ethics of Freshwater Use: A
Survey, Lord Selborne, 2000). The report goes on to challenge our current culture in which
water, in public policy agendas or in the news, is most frequently represented as either a
hapless victim of pollution, as a malevolent foe causing flooding or drought, or as a guilty
instigator of regional conflict. In contrast, the report stresses that we need to take a
constructive approach to water, where it is treated as a foremost priority in every
community from the local to the global.
Water infiltrates every aspect of our biological, cultural, and spiritual lives (Deep
Immersion: The Experience of Water, Robert Lawrence France, 2002), yet we continue to
regard it with derision if we even think about it at all. We imagine our lives as if somehow
they were uncoupled from the hydrological cycle, giving little thought to the ramifications
of our actions on those living downstream from our drains and streets. For many urban
dwellers, the experience of water remains restricted to the distance it falls from the tap to
the sink or that which carries away our waste upon flushing. Those championing the
beauty and emotional benefits of water are, for the most part, little-heard voices in a
deadening silence (Water-logged-in.com, Robert Lawrence France, 2001). The time is ripe
for a major paradigm shift in how we regard water on both the exterior landscapes of our
environments as well as the interior landscapes of our cultural sociology.
We urgently need to move toward achieving ecological integrity through water sensitive
planning and design. The earths hydrological cycle has been altered through centuries of
human activity. Due to its corporal form, water is an important integrator of environmental
disturbance across spatial scales from small subdivisions to regional drainage basins. As a
result, local site-specific actions are cumulative on larger landscapes. This behooves us to
exercise caution and to employ wisdom in the planning and design of individual development projects. Given the demonstrable decreases in environmental integrity already manifest,
proactive planning and design can provide positive environmental benefits.
Landscape architects have become active participants and leaders in developing lowimpact stormwater management, wetland park creation, riparian buffer-corridor planning,
and watershed management. It is hoped that this book will come to be regarded as a
landmark effort in which to shape and advance these emerging fields of truly integrative
water management and land development. In that light, this book has three goals:
1. To present promising procedures, options, and limitations in water management
across the spatial scale from parking lots to bioregions
2. To introduce novel approaches and explore future directions in water sensitive
site-specific design, watershed planning, and waterside development policy
3. To examine case studies employing innovative techniques used by a variety of
disciplines representing those concerned with water sensitive development
Koichiro Matsuura, the Director-General of UNESCO, concluded his message on the

occasion of World Water Day 2000 with the following apt statement: There is a fundamental truth which I would like to emphasizethe water supply does not run dry when
it is drawn from the well of human wisdom. Readers will find a wellspring of such
wisdom contained within these pages, arising from the contributors whose work is documented here. Therefore, drink deep, think hard, and act imaginatively.
Robert L. France, B.Sc., M.Sc., Ph.D.
Associate Professor of Landscape Ecology
Science Director of the Center for Technology and Environment
Graduate School of Design, Harvard University,
Founder and Principal
W.D.N.R.G. Limnetics
Preface
Origin of the book
In Spring 2000, a symposium was convened at the Graduate School of Design at Harvard
University. Titled Water Sensitive Ecological Planning and Design, this gathering was
sponsored by the Department of Landscape Architecture and brought together more than
50 presenters from North America and Europe (see www.gsd.harvard.edu/watersymp
for further details). The symposium was perhaps most successful and instrumental in
demonstrating the extremely broad range of interests and professional approaches encompassed within water sensitive planning and design. In addition to landscape architecture,
the following disciplines were represented by the presenters: governance, history, fisheries
resource management, environmental economics, ecotoxicology, limnology, urban planning, hydrology, land-use planning, civil engineering, environmental conflict resolution,
soil science, citizen activism, ornithology, environmental arts, cultural geography, extension school outreach, conservation biology, education, performance arts, and watershed
policy management. The symposium was attended by over 300 practitioners, academics,
students, and interested private citizens. Based on the formal written responses and
informal comments received in the time since, many attendees felt the meeting to be one
of the most significant in which they had ever participated. The present book was developed from a subset of 27 of the oral and poster presentations at the Harvard Design School
symposium, in addition to another eight works that were independently solicited.
Purpose of the book

The book is designed to introduce, describe, and demonstrate new interpretations to water
management in a form to engage the broadest audience possible. A few readers will find
familiar elements here, but even the most educated will be surprised at the variety of
novel approaches explored and developed within these pages. This collection of papers
pushes the frontiers of standard water management toward new directions, challenging
readers into abandoning the comfortable safety of conducting business-as-usual within
narrow disciplinary confines, and instead directing views outward to the exciting and
incompletely mapped regions of true interdisciplinary water sensitive planning and
design. Represented in these pages are the pioneering efforts of those individuals working
at the intersections of paradigms, those very regions where as Thomas Kuhn informs
us visionary insight and conceptual evolution are most likely to occur.
Structure of the book

The book is arranged in two parts that loosely pertain to the physical scale of research
and study scope. In Part I, 17 chapters address the subject of site-specific water sensitive
design, and in Part II, another 17 chapters focus on issues relating to the water sensitive
planning of riparian buffers and watersheds. Even a brief perusal of these chapters,
however, will strike perceptive readers that this structuring is somewhat artificial in that
one persons concept of design might very well be anothers idea of planning. In
other words, the thematic structuring of chapters in the book really has as much to do
about convenience as it does about trying to conceptually confine an element such as
water, which by its very nature is known for its ability to defy rigid compartmentalization.
The Overview papers leading each of the design and planning sections of the
book are intended to provide a brief roadmap to navigate through the 17 subsequent
chapters. At the end of each of these sections are two point-form summaries arising from
taped discussions with presenters and a few invited participants attending the Harvard
Design School symposium. These discussions provided opportunities for leaders to elaborate on past problems, current challenges, and future directions in their respective fields.
Following each chapter is an accompanying Response whose purpose is to succinctly
tease out several salient features from the chapter in more detail, as well as to emphasize
cross-linking to other chapters in the book.
And finally, many of the figures in the book are available in color on an accompanying
Web page: www.gsd.harvard.edu/watercolors. Use the password lentic-lotic.
Readers are advised that the color versions of images may be particularly helpful in
appreciating subtleties in the design projects and in understanding some of the planning
maps.
Acknowledgments
No conference at the Harvard Design School is ever the product of the efforts of a sole
individual. George Hargreaves, Chairman of the Department of Landscape Architecture,
is to be credited for having the wisdom to insist that both design and planning be
represented here together. His continued support, as well as that from the Dean of the
Design School, Peter Rowe, were instrumental to bringing about the success that the
symposium has since come to be regarded in professional circles. In terms of logistics, it
is no exaggeration to state that the entire undertaking would have been impossible without
the dedicated above-and-beyond efforts of a few individuals: Doug Cogger; Trevor
OBrien; and most especially, Lora Nielson, to whom I am ever indebted. I also wish to
acknowledge the help of Francois de Kock in designing the Web page, Markley Bavinger
in preparing the slides for the introductory lecture, and that of the moderators at the
symposium: Nicholas Pouder, Francois de Kock, John Felkner, Todd Gilens, Mike Flaxman,
and Markley Bavinger, in addition to all participants in the taped discussion workshops.
Transcription of a symposium to a final book is never as easy a process as one might
imagine. I wish to thank the acquisition and production staff at Lewis Publishers, in
particular Randi Gonzalez for her help in the cat herding process, Brian Kenet for his
enthusiasm in initiating this new series, and especially the late Arline Massey, to whose
judgment this present product owes its existence and to whom I would like to dedicate
this book.
Finally, I would like to express my sincere thanks to the contributors to this publication
from whom I have learned so much: Glenn Allen, Jim Bays, Diana Balmori, Judi Barnes,
Brian Bear, Tom Benjamin, Catherine Berris, Mike Binford, David Blau, Margot Cantwell,
Kelly Cave, Mike Clar, Larry Coffman, Tim Collins, Phil Craul, Jay Dorsey, John Felkner,
Bruce Ferguson, Mike Flaxman, Chuck Flink, Wendi Goldsmith, Dennis Haag, Stephen
Hust, Neil Hutchinson, Gail Krantzberg, Tom Liptan, Jim MacBroom, Kaki Martin, Frank
Mitchell, Amir Mueller, Bob Murase, Richard Pinkham, Nick Pouder, Rob Rempel, Clarissa
Rowe, Jeff Schloss, Larry Schwartz, Carl Steinitz, Mark Vian, Desheng Wang, Anne Weekes,
Dan Williams, Jennifer Zielinski, and Leslie Zucker. It is the insightful vision and dedicated
efforts of these individuals to whom we are all indebted for their instructional offerings
about how to live sustainably and more sanely in a healthy world where water is regarded,
protected, and cherished with the respect that it so deserves.
Robert L. France
September 2001
About the Editor

Robert France has published more than 100 papers on the ecology and conservation
biology of organisms from bacteria and algae to birds and whales, on research topics from
environmental pollution to theoretical biodiversity, and in locations ranging from the High
Arctic to the tropics (www.gsd.harvard.edu/info/directory/faculty/france/cv.htm).
He currently teaches courses at the Harvard Design School on the influence of landscape processes and development on aquatic systems, and on the ecopsychology of
humannature relationships. He is founder of the firm W.D.N.R.G. Limnetics, which
specializes in the restoration of degraded waterways.
Dr. France is senior editor at Green Frigate Books and also serves as series editor at
Lewis/CRC Press for Integrative Studies in Water Management and Land Development.
His books include: Designing Wetlands: Principles and Practices for Landscape Architects and
Land-use Planners, Landscape Architecture of Created Wetlands: CD-ROM of 17 Virtual Visual
Tours, Reflecting Heaven: Thoreau on Water, Profitably Soaked: Thoreaus Engagement with Water,
and Deep Immersion: The Experience of Water.
Contributors
Glenn Allen
Hargreaves Associates
Margot Young Cantwell

Environmental Design and Management, Ltd.
Halifax, Nova Scotia, Canada
Diana Balmori
Balmori Associates
New York, New York
Kelly A. Cave
Wayne County Department
of Environment
Watershed Management Division
Detroit, Michigan
Judi Barnes
Ontario Ministry of Environment
Toronto, Ontario, Canada
James S. Bays
CH2M Hill
Tampa, Florida
Bryan J. Bear
City of Overland Park
Overland Park, Kansas
Michael L. Clar
Ecosite, Inc.
Laurel, Maryland
Larry S. Coffman
Prince Georges County
Department of Environmental Resources
Largo, Maryland
Thomas S. Benjamin
Rizzo Associates
Framingham, Massachusetts
Timothy Collins
Carnegie Mellon University
College of Fine Arts
Pittsburgh, Pennsylvania
Catherine Berris
Catherine Berris Associates Inc.
Vancouver, British Columbia, Canada
Philip Craul
Harvard Design School
Michael W. Binford
Department of Geography
University of Florida
Gainesville, Florida
Jay D. Dorsey
Oxbow River and Stream Restoration
Delaware, Ohio
David Blau
EDAW, Inc.
San Francisco, California
John S. Felkner
University of Chicago
Department of Economics
Chicago, Illinois
Bruce K. Ferguson
University of Georgia
School of Environmental Design
Athens, Georgia
Frank Mitchell
University of New Hampshire
Cooperative Extension
Durham, New Hampshire
Michael Flaxman
Amir Mueller
ADM Landscape Architecture
Environmental Design
Raanana, Israel
Charles A. Flink
Greenways, Inc.
Durham, North Carolina
Robert K. Murase
Murase Associates
Portland, Oregon
Robert L. France
WDNRG Limnetics and Harvard Design
School
W. Erik Olson
Indian River County
Vero Beach, Florida
Wendi Goldsmith
The Bioengineering Group, Inc.
Salem, Massachusetts
Richard D. Pinkham
Rocky Mountain Institute
Snowmass, Colorado
Dennis A. Haag
Tetra Tech EM, Inc.
Lenexa, Kansas
Nicholas Pouder
Pouder Design Group
Bedford, New York
Stephen A. Hurst
Kansas Water Office
Topeka, Kansas
Robert Rempel
Ontario Ministry
of Natural Resources
Centre for Northern Ecosystem
Research
Thunder Bay, Ontario, Canada
Neil J. Hutchinson
Gartner Lee Ltd.
Markham, Ontario, Canada
Gail Krantzberg
International Joint Commission
Great Lakes Regional Office
Windsor, Ontario, Canada
Thomas Liptan
Bureau of Environmental Services
Portland, Oregon
James G. MacBroom
Milone & MacBroom, Inc.
Cheshire, Connecticut
Kaki Martin
Wallace Floyd Design Group
Boston, Massachusetts
Clarissa Rowe
Brown, Richardson and Rowe, Inc.
Boston, Massachusetts
Jeffrey A. Schloss
University of New Hampshire
Cooperative Extension
Durham, New Hampshire
Larry N. Schwartz
Camp Dresser and McKee, Inc.
Maitland, Florida
Carl Steinitz
Mark A. Vian
DeepStreams Consulting
Shokan, New York
Lee P. Wiseman
Camp Dresser and McKee, Inc.
Maitland, Florida
Desheng Wang
Carr Research Laboratory, Inc.
Natick, Massachusetts
Jennifer A. Zielinski
Center for Watershed Protection
Ellicott City, Maryland
Anne Weekes
University of Washington
Seattle, Washington
Leslie Zucker
Department of Natural Resources
Cornell University
Ithaca, New York
Daniel Williams
Daniel Williams Architect
Seattle, Washington
Contents
Series statement
Foreword
Preface
Background Perspectives of water management: Representative examples
from the recent literature..............................................................................................................1
Robert France
PART I
WATER SENSITIVE DESIGN
Responses by Robert France
Overview: New interpretations in stormwater management and
wetland park creation ...................................................................................................................9
I.1
Stormwater management and stormwater restoration...............................................11

Bruce K. Ferguson
Response Stormwater infiltration: Curing the disease rather than
treating the symptoms
I.2
Successful stormwater management ponds (Massachusetts) ...................................31

Desheng Wang
Response Centralized stormwater treatment: Improving
performance through engineering design
I.3
Open spaces and impervious surfaces: Model development principles

and benefits .........................................................................................................................49
Response Using computer scenarios to improve site design
I.4
Post-industrial watersheds: Retrofits and restorative redevelopment

(Pittsburgh, Pennsylvania) ...............................................................................................67
Richard D. Pinkham and Timothy Collins
Response Raising consciousness through interdisciplinary
design workshops
I.5
Low-impact development: An alternative stormwater

management technology ...................................................................................................97
Larry S. Coffman
Response Thinking big, acting small: Multi-tasking and the benefits
of dispersed micromanagement
I.6
Water gardens as stormwater infrastructure (Portland, Oregon) ..........................125

Thomas Liptan and Robert K. Murase
Response Letting it soak in
I.7
Retaining water: Technical support for capturing parking lot runoff

(Ithaca, New York) ...........................................................................................................155
Robert France and Philip Craul
Response To build an oxymoron: A green parking lot
I.8
A productive stormwater park (Farmington, Minnesota) .......................................175

Diana Balmori
Response Successfully marrying form and function in stormwater
management
I.9
A stormwater wetland becomes a nature park

(British Columbia, Canada) ...........................................................................................193
Catherine Berris
Response Naturalized design: Triumph of imagination and
innovation
I.10 Wetlands-based indirect potable reuse project (West Palm Beach, Florida) ......205
Larry N. Schwartz, Lee P. Wiseman, and W. Erik Olson
Response Treating wastewater with innovative technology
I.11 Restoring urban wetland pond systems (Boston, Massachusetts) ...................215
Clarissa Rowe
Response Project development through concerned citizenry
I.12 Water connections: Wetlands for science instruction (Wichita, Kansas) .............235
Robert France and Kaki Martin
Response Project development of interpretive wetlands
I.13 Constructed wetlands and stormwater management at the Northern
Water Feature (Sydney Olympic Park) ........................................................................247
Glenn Allen
Response Highly visible water: Recreating a landscape for public use
I.14 Principles and applications of wetland park creation .............................................263
James S. Bays
Response Designing wetlands for multiple benefits
I.15 Applications of low-impact development techniques (Maryland) .......................297

Michael L. Clar
Response Values of demonstration projects and case studies
of stormwater source management
I.16 Restoring and protecting a small, urban lake (Boston, Massachusetts) ..............317
Nicholas Pouder and Robert France
Response Buying time by bioengineering
I.17 Integrated ecology, geomorphology, and bioengineering
for watershed-friendly design.......................................................................................341
Wendi Goldsmith
Response Sustainability through interdisciplinarity
Discussion summary: Constraints, challenges, and opportunities
in implementing innovative stormwater management techniques ................................355
Discussion summary: Moving from single-purpose treatment wetlands
toward multifunction designed wetland parks...................................................................357
PART II
WATER SENSITIVE PLANNING
Responses by Robert France
Overview: New interpretations in the management of watersheds and
riparian buffers and corridors.................................................................................................359
II.1
Shoreline buffers: Protecting water quality and biological diversity

(New Hampshire).............................................................................................................361
Frank Mitchell
Response Buffer strips: More than green eyelashes?
II.2
River restoration planning (Connecticut)...................................................................379

James G. MacBroom
Response Water quality improvements are not enough
II.3
Greenways as green infrastructure in the new millennium ..................................395

Charles A. Flink
Response Corridors that integrate natural, societal, and social elements
II.4
Natural resource stewardship planning and design: Fresh Pond Reservation

(Massachusetts).................................................................................................................407
Thomas S. Benjamin
Response Protecting and restoring treasured landscapes:
Complexity and integration
II.5
Treating rivers as systems to meet multiple objectives...........................................431

Leslie Zucker, Anne Weekes, Mark Vian, and Jay D. Dorsey
Response Beyond the banks: Holistic planning of rivers as more
than the sum of their parts
II.6
What progress has been made in the Remedial Action Plan program
after ten years of effort? (Ontario, Canada) ...............................................................445
Gail Krantzberg and Judi Barnes
Response Measuring recovery of impaired waters
II.7
Watershed management plans: Bridging from science to policy

to operations (San Francisco, California)....................................................................459
David Blau
Response Sociology of implementing adaptive management
II.8
Watershed assessment planning process assessment

(Johnson County, Kansas) ..............................................................................................477
Dennis A. Haag, Stephen A. Hurst, and Bryan J. Bear
Response Managing suburban watersheds for multiple objectives
II.9
Urban watershed management (Detroit, Michigan) ................................................491

Kelly A. Cave
Response Looking beyond the end of the pipe
II.10 Modeling a soil moisture index using geographic information

systems in a developing country context (Thailand)...............................................513
John S. Felkner and Michael W. Binford
Response Incorporating scientific information into land-use planning
II.11 The design of regions: A watershed planning approach to sustainability.........541
Daniel Williams
Response Expanding planning vision in space and time
II.12 GIS watershed mapping: Developing and implementing a watershed
natural resources inventory (New Hampshire) .........................................................557
Jeffrey A. Schloss
Response Janus planning: Using computer tools to look backward
and forward simultaneously
II.13 The effect of spatial location in landwater interactions: A comparison
of two modeling approaches to support watershed planning
(Newfoundland, Canada) ...............................................................................................577
Response Linking land use to landscapes for water quality protection
II.14 Spatial investigation of applying Ontarios timber management guidelines:

GIS analysis for riparian areas of concern.................................................................601
Robert France, John S. Felkner, Michael Flaxman, and Robert Rempel
Response Size matters
II.15 Aquifer recharge management model: Evaluating the impacts of urban
development on groundwater resources (Galilee, Israel) .......................................615
Amir Mueller, Robert France, and Carl Steinitz
Response Planning by examining alternatives
II.16 Factors influencing sediment transport from logging roads near boreal
trout lakes (Ontario, Canada)........................................................................................635
Robert France
Response Empirically testing planning assumptions
II.17 Limnology, plumbing and planning: Evaluation of nutrient-based limits
to shoreline development in Precambrian Shield watersheds ..............................647
Neil J. Hutchinson
Response Landlake linkages and land-use limits
Discussion summary: Social and political issues in managing riparian
buffers and corridors.................................................................................................................683
Discussion summary: Multiple objectives in watershed management
through use of GIS analysis ....................................................................................................685
Postscript: Implementing water sensitive planning and design .....................................687
Index .............................................................................................................................................689
L1562-1Ch00Frame Page 1 Wednesday, April 3, 2002 12:40 PM
Background Perspectives of water

management: Representative examples
from the recent literature
Robert France
Water sensitive planning and design is not new. It lies at the historic base of civilization
from whence we developed and can be expected to play a profound role in shaping our
common future. As elaborated on in subsequent pages, the chapters comprising the present
book put forward the very best of modern water sensitive planning and design. It is
important to realize that though this new work offers imaginative and novel approaches
and solutions to water management, it very much stands on the collective efforts of many
involved in centuries of closely fostered and carefully studied water-culture relationships.
For example, the Harvard Library system lists over 2000 documents going back more than
400 years under searches for water planning and water design.
Giving credence to historic precedents in no way diminishes the contribution of those
whose work is included in these pages. Instead, by understanding the cumulative wisdom
and experience encapsulated within the state of the art of current water sensitive planning
and design, readers can have a frame of reference from which to appreciate the additional
benefits accruing from implementation of ideas advanced in the present compendium of
work. To this end, eight representative publications from the recent literature were selected
that demonstrate a snapshot of innovative thinking important for establishing what is in
many respects a new paradigm of water sensitive planning and design. Salient features
extracted from these publications concern the four major subject areas targeted in the
present book stormwater, wetlands, riparian buffers, and watersheds. This point-form
synopsis therefore outlines the foundation upon which the subsequent chapters in the
present book are further developed toward new, exciting, and profitable directions.
Stormwater
American Association of Landscape Architects. 1996. Integrating Stormwater into the Urban
Fabric. Portland, Oregon Conference Proceedings. 21 multi-authored chapters. ASLA Publishing. 99 pp.
Three-quarters of the job of water sensitive planning and design entails communicating and dealing with other people rather than producing construction drawings.
By focusing on end-of-the-pipe controls for stormwater management, we fail to
realize that it is the pipe itself that may be part of the problem.
A critical need exists for the utilization of natural land forms such as trees and
grasses to retain and filter stormwater.
The publication emphasizes the truism that we can shrink our streets and parking
lots (and therefore reduce impervious surfaces) without shrinking our economies.
Sprawling suburbs and vacant city cores represent slash-and-burn urbanism.
We must realize that floods are essential for healthy stream ecosystems.
1-56670-562-2/02/$0.00+$1.50
Handbook of water sensitive planning and design

It is important to look at the site history for understanding the effects of urbanization on hydrology.
In order to effectively manage river systems, we must realize that fish really live
in the trees; i.e., they are part of the riparian system.
Steam degradation can begin at levels of imperviousness as low as 10% watershed
cover.
Goals of urban water management should be balanced in respect to realistic expectations that are likely to occur with implementation of design improvements.
Landscape architects, urban planners and designers, and civil engineers need to
accept the precept that ecological functions are equally as important as other
elements of design such as aesthetics and hydrology.
Given that often the best designs are those that closely mimic nature, background
data are needed from natural sites.
Streams are most effectively managed from up in their watersheds than from
within their restricted reaches.
Effective stormwater management need not take up much space and can therefore
be retrofitted into dense urban fabrics.
Daylighting is increasingly becoming an important process in urban water restoration and urban neighborhood renewal; i.e., not only have we lost our buried
streams, we have often lost the very memory of those streams.
Development planning should consider concepts of watershed carrying capacity.
One of the most important reasons for protecting urban streams and wetlands is
for the education values they bring to city dwellers, whose concepts of water have
become much reduced.
Site planning measures need to be preventative instead of reactive.
Land-use planners need to understand that water management is as critical as
transportation issues in the urban environment.
Stormwater must come to be looked upon as a resource for, not an impediment
against, urban design, such as, for example, in the creation of rain gardens.
Development of green eco-roofs provide treatment and flood mitigation at the
source and are thus cost-effective.
Bay Area Stormwater Management Agencies Association. 1999. Start at the Source. Design
Guidance Manual for Stormwater Quality Protection. Tom Richman Associates. 165 pp.
The best opportunity to reduce urban runoff occurs during the planning and design
stage of project development.
Given site idiosyncrasies and local regulation vagaries, it is impossible to produce
a single set of specific design guidelines to be rigidly adhered to in all situations.
Maintenance and operation of control measures are as critical to stormwater management as are their proper selection and design.
Extent of impervious coverage provides a convenient barometer of environmental
disturbance.
A desperate need exists for a major paradigm shift to occur in urban stormwater
management from an end-of-the-pipe, conveyance mentality to a dispersed, startat-the-source form of micromanagement based on infiltration.
We must stress the concept that every site is in a watershed.
We need to think big in terms of beneficial effects while acting small in terms of
establishing site solutions and in accommodating runoff from storms.
Keep the design solutions as simple as possible.
Background perspectives of water management
Every opportunity should be taken to integrate stormwater solutions into the

overall site plan in order to provide recreational, aesthetic, and wildlife habitat
benefits; i.e., stormwater needs to be looked upon as a design resource, not a
planning hazard.
Starting at the source for managing stormwater makes demonstrable economic
sense.
Conventional zoning practices typically do not address issues of water sensitive
planning.
Clustered development preserves open space in addition to providing room for
inserting stormwater management practices.
Streets and parking lots are often the major development on the land and must be
carefully addressed in terms of minimizing runoff.
Generation of public awareness is integral to implementing nonconventional water
sensitive designs such as the use of porous pavement.
Attention to site design is critical to improving stormwater management, in particular the following elements: define development envelope and protected areas;
minimize directly connected impervious surfaces; maximize permeability wherever possible, and use drainage as a design element rather than hiding it away.
Design for the entire drainage system to cost-effectively manage flooding, control
streambank erosion, and protect water quality.
Water sensitive design employs both structural and nonstructural elements and
can be fit into all types of site conditions.
Wetlands
Azous, Amanda L. and Richard R. Horner (eds.), 2001. Wetlands and Urbanization. Implications for the Future. Lewis Publishers. 14 multi-authored chapters. 338 pp.
Wetlands are very susceptible to the effects of urbanization, with these effects being
felt on the scale of both the landscape as well as individual bodies of water.
The need for a landscape approach to managing wetlands as parts of watersheds
exists.
It is useful to develop a comprehensive research program based on strong scientific
acumen.
It is important to place the study system within a framework of previous research
conducted on other wetlands.
Undertaking an inventory of land-use patterns through the use of GIS analysis
produces notable benefits.
We must emphasize the cardinal importance of imperviousness in affecting wetland health.
We are required to understand wetland morphology and hydrology in order to
assess the influences of urbanization.
Comparisons of the functional ecology of wetlands in more pristine locations are
critical toward gauging development effects upon urban wetlands.
Consideration of soils and seasonality is important to assessment.
Size alone is an inadequate criterion for judging the health of wetlands.
Because wetlands are frequently characterized by relatively high biodiversity, even
when located in urban areas, they represent natural arks.
Cross-system empirical relationships, especially in relation to magnitude of deforestation, are useful for surveying the magnitude of urbanization effects.

Detailed water balance models are essential for understanding how wetlands are
affected by urbanization.
Production of a list of stormwater management guidelines provides a useful communication tool to measure alterations in ecosystem integrity.
France, Robert Lawrence. 2002. Designing Wetlands: Principles and Practices for Landscape
Architects and Land-Use Planners. W.W. Norton. In press.
Technology transfer to aid communication among disciplines is needed.
Assessments must be made of the cumulative impacts of many individual design
projects from a larger watershed perspective.
We must emphasize integrating concepts of wetlands as human amenities into the
design process; i.e., the form of created wetlands may be just as important as their
hydrological, chemical, or biological functions.
Principles of wetland design should be approached as general concepts rather than
construed as rigid instructions.
Structuring and maintaining biological integrity may be the most important role
of many wetlands.
The retention and removal of contaminants in wetlands depends on a complex
and interrelated system of chemical transformations.
A strong, positive relationship exists between the percentage of upstream wetlands
lost and the percentage increase in watershed peak flow discharges.
Wetlands offer a myriad of opportunities for formalized natural history education.
Careful attention to design, construction, and maintenance will ensure that created
wetlands will operate effectively.
Focusing on conceptual planning, site preparation and construction, planting, and
long-term monitoring is essential to guaranteeing success in the creation of wetlands.
On average, three-quarters of all analyses indicate that contaminant removal performances of 75% or greater are possible given adherence to appropriate treatment
distances and transport times.
Wetlands are created for a variety of different reasons, the choice of which will
influence the initial screening and selection of suitable vegetation.
The efficacy of treatment wetlands generally increases with their progressive size.
A myriad of opportunities exist for exciting and imaginative landscape artistry
that can augment wetland function.
The interlinked networking of cells within an individual wetland or the specific
associations of grouped wetland features within a larger combined waterwetland
system will strongly influence resulting performance.
Failed wetland creation/restoration projects abound due to opaque goal aspirations, poor engineering and ecological design, inadequate background data collection, absence of adaptive management, the erroneous assumption that project
termination occurs with planning, and a cavalier process of critical evaluation.
Recognition of the need for preserving undamaged wetlands as well as for restoring those that have already been degraded will be fostered by facilitating public
access and easy interpretation at created wetland parks.
An inventory of wetland resources is needed to assess their value to the community
and the ecosystem as well as to help establish precise goals and recognizable steps
for watershed management.
Screening adjudication of potential mitigation sites involves assessment of wetland
functions, project goals, and legal possibilities.
Riparian zones
Verry, Elon S., James W. Hornbeck, and C. Andrew Dolloff (eds.). 2000. Riparian Management in Forests of the Continental Eastern United States. Lewis Publishers. 20 multi-authored
chapters. 402 pp.
Riparian forests are in a constant state of dynamic equilibrium that integrates landuse alterations; in other words, they are defined as much by their functions as by
their structures.
It is important to consider development history on a site as a means for predicting
future responses.
Riparian forest management has a long-established tradition in Europe.
Because the multifaceted roles of riparian areas are so closely dependent on watershed position, it may be necessary to manage them differently in relation to
their steam order; i.e., no simple minimum buffer strip width is suitable for all
situations.
We need to consider floodplain geomorphology when managing riparian areas;
i.e., a landscape perspective must be applied to riparian management.
Riparian ecosystems are characterized by great biodiversity and high productivity
far in excess relative to the physical area that they comprise.
Riverine riparian areas operate as important wildlife corridors.
Timber production should always be a secondary benefit in riparian management;
instead, such areas should be managed for fish.
Valley segments are convenient management units.
Streams are closely linked to riparian forests through the role of the latter in water
yields, erosion, nutrient transport, water temperature, and organic matter supply.
Such ecotonal landwater linkages should form the basis for management practices.
Stream health can be closely influenced by groundwater processes and soil regimes.
Assessment of stream typology (based on morphology) is an important element
in riparian management.
Land clearance removes riparian sources of large woody debris needed to help
structure stream morphology and sustain aquatic health.
Road construction in riparian areas should be approached with caution and wisdom given justifiable concerns about erosion and sedimentation.
Management of the aquatic side of riparian zones should not be ignored.
Riparian zones are complex ecosystems necessitating complex (integrated) management solutions that focus on holistic options considering social, economic, and
political interests at the earliest planning stage.
Effective riparian management depends on closely considering human dimensions
as much as it does on dealing with biogeochemical processes.
The most successful management plans are those that engage the public in decision
making.
Planning is the most important best management practice (BMP).
We should develop management guidelines in relation to science, practicality, and
economics.
Effective management is predicated on effective monitoring.
Palone, Roxane and Albert H. Todd. 1998. Chesapeake Bay Riparian Handbook: A Guide for
Establishing and Maintaining Riparian Forest Buffers. USDA Forest Service. 890 pp.
Riparian forest buffers perform a suite of water-protective functions in urban,
suburban, agricultural, and forested landscapes.
Management of riparian forest buffers can be implemented most effectively when
approached on a spatial three-zone concept in relation to permitted development.
Understanding the physiographic and hydrological attributes of riparian forests
is necessary to effect wise management decisions.
In addition to roles in influencing water quantity and quality, and in supporting
shoreline and aquatic biodiversity, riparian forest buffer systems also offer recreational opportunities in an aesthetically pleasing environment.
The essential need to understand soil characteristics and forest ecology as part of
the process for planning the establishment of riparian buffer systems is essential.
There is a need to relate upland contaminant loading before the removal capabilities of forest buffers can be assessed.
Use of science-based criteria is essential for determining effective buffer strip
widths in relation to site factors and predicted environmental threats.
A whole suite of streamside restoration practices exist that can be implemented to
return riparian forests to their protective functions.
Stream systems rather than individual streams are the more appropriate scale for
designing and managing riparian buffer networks.
Forestry activities, road construction, and building development within riparian
buffer systems must obviously be undertaken with caution and wisdom to limit
deleterious aquatic effects.
Riparian forest buffers offer a range of economic, in addition to ecologic, benefits.
There is a need to educate the public and enforcement professionals about the
important roles played by forest buffers in all types of landscapes.
Watersheds
Lal, Rattan (ed.) 2000. Integrated Watershed Management in the Global Ecosystem. CRC Press.
23 multi-authored chapters. 395 pp.
There is need for a diverse interdisciplinary group considering social and economic, in addition to technical, issues.
Focuses on the interface of land-use planning with watershed management.
Interaction between water scarcity and food security in a global perspective, including tropical developing nations, exists.
The watershed is the basic land management unit for agriculture.
Soil erosion is the major threat to the quality of global water supplies as well to
cropland degradation.
The health of nations is determined by the health of their watersheds is examined.
Historical perspectives are important in order to help understand present problems.
Social justice (equity, gender, etc.), community participation, and institutional infrastructure when managing watersheds is critical.
We need good soil and hydrologic science to predict the extent and consequences
of soil erosion.
There are links among agricultural land-use, soil fertility, chemical nutrient applications, and aquatic eutrophication.
Greater reliance on indigenous wisdom in addition to scientific knowledge is

necessary.
Experimental manipulations are required to identify the most useful best management practices (BMPs).
Simulation models are an important tool for stimulating discussions through characterizing responses to land management interventions.
An organizational structure that incorporates both physical and social elements
for managing watershed agroecosystems must be developed.
Initiation of local-level resource monitoring by watershed management organizations is needed.
Holistic over reductionist approaches are required for effective management.
Successful watershed management is built on the two pillars of practical technical
innovation and participatory institutional innovation.
Locals should be empowered in generating bottom-up solutions.
Watersheds must be viewed from perspectives of both service and productivity
functions.
Concepts of sustainability must be infused into all stages of watershed management; i.e., the landscapelifescape interaction.
Scenario models linked to GIS analysis are a powerful means for evaluating actual
and alternative management practices.
Groundwater pollution in rural watersheds is significant.
Downstream economic impacts of soil erosion on a lake-by-lake basis exist in terms
of recreational boating and restorative dredging.
Global climate change influences watershed functioning and management is examined.
EPA. 1997. Top 10 Watershed Lessons Learned. EPA. 59 pp.
The best plans have clear visions, goals, and action items.
Watershed approaches are developed over time and need dynamic leaders with
strong interpersonal skills.
Implementation of watershed plans benefit from the presence of a dedicated coordinator to help build relationships.
Watershed management provides a means in which to integrate environmental,
economic, and social aspirations under an umbrella of sustainability.
Generating watershed management plans that stand a real likelihood of becoming
implemented rather than just sitting on the shelf is necessary.
It is important to engage all stakeholders, including industry, in a positive manner
working toward a common goal of improving livability and environmental integrity in the watershed.
Watershed groups should become familiar with a diversity of planning and communication tools such as guides, workshops, GIS mapping, etc.
A procedure for monitoring progress in the implementation of watershed planning
recommendations needs to be incorporated into management guidelines.
Local empowerment through education provides tangible results in increasing
awareness about watershed issues.
Watershed plans work best when they are designed to advance incrementally
forward in small, manageable components rather than as single, all-inclusive vague
concepts.
part I
Water sensitive design

Overview: New interpretations in stormwater management
and wetland park creation
Part I comprises 17 chapters and responses that deal with two of the most important issues
associated with water sensitive design: stormwater management and wetland park creation. Together, these chapters go far toward demonstrating the precept that just as water
is an integrator across physical landscapes, likewise it is an integrator across professional
disciplines. For in addition to landscape architecture, elements of civil engineering, urban
planning, environmental art, cultural geography, hydrology, soil science, biology, sociology, and geology are all present in these pages.
Seven chapters focus on managing stormwater on sites using a combination of decentralized, small-scale approaches that are representative of the emerging paradigm of lowimpact development. Landscape architect Bruce Ferguson provides an informative introduction to stormwater in urban and suburban settings and reviews concepts of porous
pavement as an effective management technique. Jennifer Zielinskis chapter emphasizes
the importance and wisdom of considering environmentally sensitive water management
at the very earliest possible stage in the site design process. This topic is elaborated on in
great detail by Larry Coffman, whose chapter summarizes his pioneering work in establishing low-impact development as a viable means for addressing stormwater management. Four of the chapters those by Richard Pinkham and Timothy Collins, by Robert
France and Philip Craul, by Michael Clar, and by landscape architects Thomas Liptan and
Robert Murase provide a diverse set of design examples of how stormwater can be
managed and utilized in an environmentally sensitive and aesthetically pleasing fashion,
true to the spirit of low-impact development.
Five chapters focus on managing stormwater through creating wetland parks. James
Bays provides a comprehensive survey and discussion of the various attributes needed
to be incorporated into designs in order to develop multifunctional wetland parks for
regulating the quantity and treating the quality of stormwater at the same time as providing a whole suite of additional environmental and cultural benefits. Four of the
chapters those by landscape architect Diana Balmori, by landscape architect Catherine
Berris, by Robert France and landscape architect Kaki Martin, and by landscape architect
10
Glenn Allen show how these design ideas can be imaginatively adapted to varied
situations of high visibility and ensuing public use.
Three chapters concern the restoration of degraded water bodies. Wendi Goldsmith
provides a discussion of how the emerging techniques of shoreline bioengineering can be
used in the broad context of both watersheds and disciplines. Landscape architect Clarissa
Rowe demonstrates how the methods of wetland design can be implemented into larger
management plans toward improving site conditions for urban ponds. In the study
described by landscape architect Nicholas Pouder and Robert France, both of these restoration tools are integrated in a demonstration project designed to raise pubic consciousness
about lake restoration.
The chapter by Desheng Wang describes methods needed to improve the design of
stormwater detention basins, long the favored practice of many civil engineers, but unfortunately rarely approached with such environmental sensitivity as illustrated by the
present examples. Finally, the chapter by Larry Schwartz presents an interesting example
of utilizing wastewater as a functional amenity in the design of a treatment wetland park.
chapter I.1
Stormwater management and

stormwater restoration
Bruce K. Ferguson
Abstract
A watershed maintains its natural health and its benefits to human beings by the accumulation, storage, and gradual flow of subsurface water. The fundamental disease of urban
watersheds is sealing by impervious cover, which deflects runoff across the surface and
carries pollutants into streams. Approaches that manage stormwater on the land surface
treat only the downstream symptoms: conveyance, detention, and stormwater wetlands
all fail to eliminate the fundamental urban problem of excess surface water volume. In
contrast, stormwater infiltration forces surface runoff back into the underlying soil, curing
the disease by restoring watershed process. Infiltration basins are well known in diverse
geographic regions and have been integrated into diverse site circumstances. Porous
pavements infiltrate urban runoff directly, where the rain falls; for many types of porous
pavement the long-term infiltration rate is sufficient to absorb and treat the rain that falls
during almost all storm events. Only a small capacity for daily infiltration is required to
create a substantial cumulative effect over time. Unlike any surface management approach,
infiltration is capable, within the limitations of specific sites, of solving all the problems
of urban runoff, because it calls on the power of the underlying landscape.
Introduction
From the viewpoint of the health of the environment, the most important aspect of
stormwater design is what the design causes water to do. Many other important aspects
of stormwater design are involved: design can integrate water with all the other concerns
of an urban site, and in this collection other fine practitioners discuss that kind of topic
(for example, Pinkham and Collins, Liptan and Murase, Berris, and Flink). My subject
here, however, is limited to the single question of where water goes in the environment,
and the kinds of processes it experiences there.
The volumes of water involved in this question are vast, and what happens to it is a
vital question in the welfare of people. In recent decades, designers have put quite a bit
of ingenuity into various alternative answers to this question, with some degree of success
if you look at them through sufficiently narrow glasses.
This chapter reviews some of those alternatives and then points to what we should
be doing a lot more of, given our scientific understanding of what makes landscapes
1-56670-562-2/02/$0.00+$1.50
11
12
Figure I.1.1 An apartment site in Long Beach, CA. (Color version available at www.gsd.harvard.
edu/watercolors. Password: lentic-lotic.)
healthy. In the course of this, examples of installations from around the country will be
presented, to illustrate that not only should we do these things, but we can do them.
The urban watershed problem

The discharge during a storm from typical urban culverts exemplifies the problem that
urban watersheds present to us. Even the largest culverts flow nearly full. With every rain
that falls on urban watersheds, floods bring oil, bacteria, and sediment from the watersheds impervious pavements and eroding stream banks. In some urban watersheds, the
floods get into sanitary sewers, adding overflows of raw sewage to the stream flow.
When the rain stops, little base flow remains in urban streams, because there is no
water left in the watershed. Groundwater levels are low. Many cities are left, paradoxically,
with local water shortages, and aquatic ecosystems are left without habitat.
The problem begins throughout urban watersheds, at sites similar to the one shown
in Figure I.1.1. This sites dense impervious surfaces accumulate oils, bacteria, and metals.
The same surfaces turn rainfall into surface runoff that carries the pollutants into streams.
Here, at the site where the rain falls, the problem has been created: surface stormwater
has been generated throughout the watershed, and pavements and channels are holding
the concentrated water and its concentrated pollutants. Now we have to manage what
we have created. This is the beginning of stormwater management.
It is possible to put the excess polluted water that comes off watersheds to constructive
use. A successful example is at a horticultural nursery in Thomson, GA, which is fortuitously located on a small headwater watershed. The nursery has paved most of the
watershed with plastic sheets and greenhouses. A pond captures the abundant runoff;
from the pond, pumps recycle the water for the nurserys irrigation. This on-site water
harvesting makes the use of water resources efficient and the future of this waterconsuming industry sustainable, but implementing this approach requires special site
conditions. Nationwide the occasions for doing it have been extremely limited. So we are
left with the generic problem of urban watersheds, and we need to find an answer.
It is possible to eliminate excess runoff at the start. For example, at an office site in
Atlanta, Robert Marvin dispersed 200 parking spaces across the wooded site, tucking a
Chapter I.1:
Stormwater management and stormwater restoration
13
Figure I.1.2 A detention basin at a commercial site in Georgia. (Color version available at www.gsd.
harvard.edu/watercolors. Password: lentic-lotic.)
space or two at a time between the preexisting trees, and carefully preserving the forest
floor adjacent to each pavement edge. The vegetated soil infiltrates and eliminates the
runoff from each bit of pavement; but spreading out the parking here meant that the 200
parking spaces used up a greater amount of land, with a greater length of paved traveling
lane connecting the dispersed spaces. The large amount of land this approach requires
inhibits its widespread use. So, for most sites, we are still left with the generic urban
problems of runoff and pollution, and we cannot escape the need to find an answer.
Some management alternatives that have been tried

Among the various approaches to stormwater management that have been developed
over the years, the oldest is conveyance, which is the use of channels of various kinds,
originally to move surface water away from urban sites to reduce on-site nuisance. Today,
the ability to move surface water safely from place to place remains vital on every site:
conveyance channels control the flow through any other type of system we put on a site,
and they discharge the overflows from large, rare storms; however, conveyance, where it
is used alone, passes all the urban problems on downstream, through successively larger
channels, without the benefits of filtration or recharge. So designers have explored supplements and alternatives to it.
For the last several decades, designers have been supplementing urban conveyances with
detention basins similar to the one shown in Figure I.1.2. The big culvert brings surface runoff
from this highly impervious commercial site into a reservoir. During storms the water rises
high in the basin. A relatively small orifice in the concrete structure limits the rate of outflow.
Figure I.1.3 shows a typical result. A pulse of surface runoff comes in at a high rate. The basin
stores the runoff for a while, while the outlet lets it out at a relatively low rate. Detention
storage suppresses the peak rate of storm flow, as it has been intended to do since this practice
started in the 1960s, but detention is only a relative, quantitative modification of conveyance;
the total volume of runoff still continues downstream during the storm event, stretched out
over time. The experience is that detention has failed to prevent urban flooding and erosion
and has never done anything for water quality, groundwater replenishment, or urban water
supplies (Ferguson, 1998, p. 164). So designers have considered further alternatives.
Rate of flow, cfs
14
Entering
basin
Leaving basin
Time, hours
Figure I.1.3 Typical hydrologic outcome of detention during a storm event.
Figure I.1.4 A stormwater wetland at a convention center near Orlando, FL. (Color version available
at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Relatively recently, designers have been supplementing conveyance more elaborately

with extended detention and stormwater wetlands. The wetland pond shown in
Figure I.1.4 is a particularly nice-looking example. It receives surface runoff from the roofs
and parking lots of a large convention center. The still water of the pool helps water
quality: suspended particles settle out, the wetland plants adsorb and biodegrade hydrocarbon pollutants, and the fountains in the background oxygenate the water. The design
as a whole has multiple functions for aesthetics and wildlife habitat. On the other hand,
the construction of the pool preempted the wetland and riparian functions that used to
exist in this low-lying spot, barred fish migration through the natural stream system, raised
the temperature of stream water, and shifted the stream systems trophic state. It also did
not eliminate the fundamental problem of the unnatural volume of surface water. Wetlands
such as this, as well as ordinary detention basins, modify the excess runoff before it passes
on downstream, but do not change the fact that it has been created. They manage the problem
of excess urban runoff, and perpetuate it. So we continue to consider still-further alternatives.
Chapter I.1:
15
1.5
Precipitation
Inches per day
1.0
P (in/day)
0.5
0.0
0
30
60
90
Days, March-May 1960
Figure I.1.5 Daily precipitation in Atlanta, GA, in the spring of 1960. (Data from Ferguson, B.K. and
Suckling, P.W., Changing rainfallrunoff relationships in the urbanizing Peachtree Creek watershed,
Atlanta, Georgia, Water Resour. Bull., 26, 313, 1990.)
What makes landscapes healthy

As a basis for dealing completely and fundamentally with the problem of urban watersheds, we must first establish how landscapes fundamentally work. Landscapes are open,
dynamic systems with inflows and outflows of resources (Ferguson, 1994, 1996).
Precipitation is the vast hydrologic inflow. Figure I.1.5 shows the daily precipitation
in Atlanta for a 90-day period in the spring of 1960. Please note some of the specific
numbers in this chart, because they will be referred to in the following figure. There is
one precipitation event with almost 1.5 in./day, several others with about 0.3 to 0.5 in./day,
and many very small events. This shows that rainfall comes to a landscape, over time, in
discrete, isolated events, spaced apart from each other by a few days or a few weeks with
no rain, and the events of different sizes come in quite random order.
Figure I.1.6 shows the stream flow coming out of Atlantas Peachtree Creek watershed
during the same 90-day period. This watershed is quite urban, with 30% impervious cover
in 1960 (Ferguson and Suckling, 1990), so during storms the stream outflows shown in
Figure I.1.6 are much bigger than they would have been without the impervious cover;
however, this chart applies to the same time period as the previous one and is graphed
in the same units, and you can see that the storm flows are smaller than the precipitation
that created them, even in this urban watershed: the big event is reduced to 0.7 in. of
outflow, and the smaller events are almost imperceptible. In place of the pattern of isolated
precipitation events, the watershed has created a constant, moderate base flow, which is
the base resource for perennial water supplies and aquatic ecosystems.
Figure I.1.7 again shows Peachtree Creeks stream flow, with daily evapotranspiration
added in the light tone. In the Peachtree Creek watershed, as in most of the landscapes
in the world, more than half of the annual rainfall discharges as evapotranspiration from soil
and plants, without passing through a stream. Evapotranspiration supports the terrestrial
16

1.5
Stream flow
Inches per day
1.0
Q (in/day)
0.5
0.0
0
30
60
90
Figure I.1.6 Stream discharge in Peachtree Creek, Atlanta, during the same period shown in Figure I.1.5.
(Data from Ferguson, B.K. and Suckling, P.W., Changing rainfallrunoff relationships in the urbanizing Peachtree Creek watershed, Atlanta, Georgia, Water Resour. Bull., 26, 313, 1990.)
1.5
Total discharge
1.0
Inches per day
Q (in/day)
Et (in/day)
Total
0.5
0.0
0
30
60
90
Figure I.1.7 Stream flow and evapotranspiration in the Peachtree Creek watershed, during the same
period shown in Figures I.1.5 and I.1.6 (evapotranspiration calculated by the monthly Thornthwaite
method and distributed daily in proportion to daily temperature).
ecosystem the same way stream base flow supports the aquatic ecosystem: all the terrestrial
vegetation lives by the flow of water from the soil, through the roots, and out through the
leaves. The heavy line at the top is the sum of stream flow and evapotranspiration; it is the
Chapter I.1:
17
Figure I.1.8 Subsurface moisture in a landscape. (Adapted from Leopold, L.B., Water, Rivers and
Creeks, University Science Books, Sausalito, CA, 1997.)
total discharge from the watershed. The continuity of that line shows that the watershed,
obeying the simple laws of physics, transforms the random, isolated, unorganized, potentially destructive pulses of inflow into a perennial discharge that is, relatively, a resource
to the ecosystem and to people.
Figure I.1.8 shows the underlying mechanism by which a landscape does that. The
figure is a section through a landscape showing the moisture under the ground surface.
On the left, the groundwater table can be seen gently reflecting the ground surface and
discharging at low points as springs and streams. On the right, the overlying unsaturated
soil has been lifted up to show that the water table is a continuous mass under the
landscape, flowing slowly and continuously, under its own weight, from its stored mass
in high areas toward the discharge points. From place to place in the earth, variations
exist in the details of this diagram with the relative permeabilities of underground layers
(Ferguson, 1985). But the accumulation, storage, and gradual flow of subsurface water
happen in essentially all landscapes (Ferguson, 1992). The infiltration of rainfall into the
subsurface of the Peachtree Creek watershed explains the reduction in that creeks storm
flow volume compared with the precipitation that produced it. The storage of that water
in subsurface pores explains the creation of stream base flow and continuous evapotranspiration.
The fact that landscapes evolve to work this way is not a matter of professional
preference or bureaucratic convention. Infiltration and storage happen in nature, all over
the world. A landscape, like any other open, dynamic system, absorbs its inflows, stores
them inside itself, turns them into resources, and uses them to maintain itself (Ferguson,
1994, 1996). The result of a watersheds working this way is that, compared with the
precipitation inflows it is given to work with, floods are moderate, erosion and sedimentation are in equilibrium, pollutants are degraded, the wetlands are sustained, and public
water supplies are secure. So nature works. It evolves to work. It has a capacity to work,
even in urban watersheds. It will work, greatly to our advantage, if we will only let it.
The understanding of this concept is a basis for us to move from treating the downstream symptoms of surface runoff to curing the disease of urban watersheds at the source.
Figure I.1.9 shows the disease: the contrast between the hydrologic processes of open,
vegetated soil and sealed, impervious cover. When rain falls on vegetated soil, it infiltrates.
Evapotranspiration sustains the local ecosystem. Further infiltrating water removes pollutants if there are any present, recharges groundwater, and restores stream base flow.
This process is the origin of a landscapes health and is of great benefit to people. In
contrast, when an impervious cover seals the soil surface, rainwater is deflected across
the surface. The surface runoff flushes pollutants directly into streams; flooding, erosion,
18
Vegetated soil
Rainfall
Impervious cover
Rainfall
Surface runoff:
Evapotranspiration:
Pollutants flush into streams
Sustains ecosystem
Flooding increases
Channels erode
Habitats expire
Infiltration is lost:
Ground water declines
Infiltration:
Streams and wetlands go dry in summer
Removes pollutants
Aquatic ecosystems die
Recharges ground water
Public water supplies decline
Maintains stream base flow
Figure I.1.9 The contrast between vegetated soil and urban impervious cover.
Figure I.1.10 A small infiltration basin near Orlando. (Color version available at www.gsd.harvard.
and habitat loss occur. Infiltration is lost, so groundwater declines; streams and wetlands
go dry in summer; aquatic ecosystems die; and public water supplies decline. If we are
going to cure the disease of urban watersheds, we have to minimize the process on the
right and maximize that on the left.
Infiltration basins
Figure I.1.10 shows an approach that aids this process. This is an example of stormwater
infiltration: the artificial forcing of surface water into the underlying soil. This infiltration
basin is located in an office park near Orlando, FL. A miniature watershed on the parking
lot pavement drains through a curb cut into a depression in the grassed island. This is
a dull-looking basin: you do not see the spectacular slash pine and palmetto vegetation
that theoretically used to grow on this site. Nevertheless, this basins process of infiltration
Chapter I.1:
19
Figure I.1.11 Average frequency of rainfall events of different sizes, in three cities (daily precipitation
data 19501989 from U.S. National Climatic Data Center, Asheville, NC).
is the restoration of hydrologic process in this landscape. Consequently, it is inherently

the most complete possible solution to the environmental problems of urban stormwater.
Figure I.1.11 shows an important point about designing for infiltration, and the performance of which infiltration is capable. The chart shows the number of rainfall events
that supply urban watersheds and that artificial infiltration would be given to exploit. On
the horizontal axis is the size of individual rainfall events in inches per day; the size
increases to the right. On the vertical axis is the number of times each event occurs in an
average year. The chart shows data for three cities: Atlanta, Boston, and Seattle. They all
tell us the same thing: the most common rainfall events by far are the smallest.
Designers have been trained conventionally to design conveyances and detention systems
for big flood events, such as the 10-year storm. Big floods are erosive and damaging when
they occur, and we need to plan for them. From an ecological viewpoint, however, they
have a problem: such events almost never happen. The 10-year, 24-hour storm is so rare
that it is off Figure I.1.11s chart to the right; its contribution to average annual inflow is
insignificant to the ecosystem and to people. The events that maintain the ecosystem are
the small ones, because they are frequent: they replenish the system many times per year.
Half of the average annual inflow comes in events of less than 1 in./day. Infiltrating only
a small volume of water from each small runoff event can largely restore cumulative
environmental function. The fact that only a small amount of water needs to be infiltrated
each day makes effective stormwater restoration surprisingly feasible. Even basins on slowly
permeable soils can have a substantial effect on water and pollution over the course of a year.
I estimate that 40,000 infiltration basins are in operation in the United States. They
are well known in diverse geographic regions and have been integrated into diverse site
circumstances.
20
Figure I.1.12 Infiltration swale at the Oregon Museum of Science and Industry in Portland. (Color
version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Figure I.1.13 Swale at a Merrill Lynch training office near Princeton, NJ. (Color version available at
www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Figure I.1.12 shows one of Robert Murases well-known projects in Portland, OR. This
is hydrologically the same idea as the grass basin in Florida, but here the parking bays
drain into swales thickly planted with native and adapted plants. Monitoring in the
Portland area has shown that the roots of these native plants help maintain an open soil
structure and cause more water to infiltrate than may be assumed in design calculations.
Weirs like the one in the foreground hold small increments of water in contact with soil
and vegetation to make sure they get treated and infiltrated, while the linear organization
of swales allows overflows from large storms to move down the swale system.
The swale shown in Figure I.1.13 does the same thing hydrologically, although it looks
more mechanical. This is an office site in New Jersey, designed by Roger Wells. Water from
Chapter I.1:
21
Figure I.1.14 Infiltration basin doubling as a playground at Village Homes, Davis, CA. (Photograph
by Roger D. Moore.) (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
the pavement surface flows across the grass slope, which acts as a kind of vegetated filter.
At the bottom of the slope, in the swale, concentrated runoff flows on and through the
stones. The stones slow the water down, prolonging its contact with soil and giving it a
chance to be treated and to infiltrate. The inlet in the background carries the overflows
from large storms.
The infiltration basin shown in Figure I.1.14 has been developed for multiple human
use. This is in the Village Homes community in Davis, CA, the work of Robert Thayer
and his colleagues. The sand-floored basin is the floor of a childrens playground. The
playground is located along a system of pedestrian ways that connect to the rest of the
community. There is little annual rainfall here, but when there is a substantial storm, runoff
comes down the system of drainage swales, paralleling the pedestrian ways, to points like
this throughout the community. Village Homes system of basins eliminates most of the
communitys surface runoff. Since it was installed in the late 1970s, it has very seldom
overflowed into the City of Davis storm sewer (Thayer and Westbrook, 1989).
The large infiltration basin shown in Figure I.1.15 has been combined with water
harvesting to support multiple ecological functions. This is on the campus of Hofstra
University on Long Island, and it receives runoff from an adjacent highway. The university
lined the floor of the basin with a plastic sheet to hold a permanent pool and planted it
with native wetland vegetation. Biological surveys here have found extraordinarily diverse
bird life, compared with the surrounding urban area. The university maintains the site as
part of its arboretum and has designated it a wildlife sanctuary. When runoff enters the
basin during storms, the water level rises; excess water infiltrates through the basins
unlined sandy sides. Due to the great capacity of this basin, there is no overflow swale;
this is the absolute end of the local surface drainage system, even for the largest storms.
Altogether, Long Island has more than 4,000 basins excavated into the sandy soil, recharging the underlying aquifer that is the exclusive water supply for 2.6 million persons (sum
of the populations of Nassau and Suffolk Counties, 1990, listed in the Web site of the
U.S. Census Bureau).
Figure I.1.16 shows stormwater infiltration lifted up to a massive scale of public works.
This is the 570-acre Rio Hondo Spreading Grounds, operated by the Los Angeles County
22
Figure I.1.15 Infiltration basin at Hofstra University, Long Island, NY. (Color version available at
Figure I.1.16 The Rio Hondo Spreading Grounds near Los Angeles. (Color version available at
Chapter I.1:
23
Figure I.1.17 Construction of a porous paver driveway in Athens, GA. (Color version available at
Department of Public Works. Throughout the Los Angeles region, for over 100 years,
numerous public and private agencies have operated thousands of basins large and small,
recharging water into aquifers in the regions alluvial fans and coastal plain. The Department of Public Works alone has 27 large facilities like this, through which it recharges
each year 40,000 acre-ft of urban runoff, mixed with even huger quantities of mountain
runoff, imported water, and reclaimed water, sustaining the coastal plain aquifer from
which 3.5 million persons take their water supply (derived from data in the Web sites of
the Los Angeles County Department of Public Works and the Water Replenishment District
of Southern California).
Porous pavements
Porous pavements allow us to take care of a large part of the urban runoff problem at the
source, where the rain falls, without having to get into downstream basins at all. For
example, Figure I.1.17 shows concrete paver units being placed for a porous driveway in
Georgia. The pavers rest on a sand setting bed with an aggregate base course. The pavers
configuration leaves drainage holes in 12% of the pavement area; the holes are filled with
fine gravel. Figure I.1.18 shows the long-term infiltration rate for pavers of this kind. The
rate declines in the first 4 to 6 years after installation, and then after 10 years it is nearly
stable at a long-term level of about 4 in./hour. The light horizontal lines show, for comparison, the intensity of the 10-year, 1-hour storm in three different cities. You can see that
the pavers long-term infiltration rate is sufficient to absorb and treat the rain that falls
even during intense storms, and even in Georgia where the 10-year storm approaches 3
in./hour.
In typical urban watersheds, pavements cover one third of the land; they produce
two-thirds of the runoff and almost all the petroleum-based pollution. In the United States,
we are paving or repaving over a quarter of a million acres per year. So paved areas that
are reclaimed for infiltration take advantage of a vast and cost-free resource for watershed
restoration at the seat of the problem of urban runoff and pollution.
Different types of porous materials are available to meet site-specific needs. For example, Figure I.1.19 shows porous concrete newly installed for a park road in Georgia. Porous
24

10
Surface infiltration rate (inches per hour)
Paver infiltration rate
8
7
6
Pavers with drainage holes
5
4
3
Atlanta 10-year storm
2
Boston 10-year storm
1
Seattle 10-year storm
0
0
10
Pavement age (years)
Figure I.1.18 Long-term infiltration rate in porous concrete pavers, in comparison with rainfall
intensity in three cities. (Infiltration rate data from Borgwardt, 1997; 1-h rainfall data from U.S.
Weather Bureau, 1955.)
Figure I.1.19 A porous concrete road in Webb Bridge Park, Alpharetta, GA. (Color version available
Chapter I.1:
25
Figure I.1.20 A parking lot in Medford, NJ, selectively combining impervious asphalt and porous
aggregate. (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
concrete bears the high traffic load for the parks main entry with a durably high infiltration
rate. This roads design illustrates an important detail for preserving the infiltration capacity
of porous pavements. In the photograph, the recently graded clay soil is eroding freely;
sediment is flowing visibly through the drainage swales. If that sediment were to get onto
the pavement, it would enter the pores and greatly reduce the infiltration rate before the
road has even been put into service. This site was properly configured by its designer,
John Gnoffo, to protect the pavement from that. Drainage is away from the edges of the
pavement in every direction, so sediment drains down eroding slopes without entering
the pavement. Curbs are omitted, so overflow drainage from the pavement, when it
happens, discharges freely into the swales, and potentially clogging surface debris is
washed or blown off the pavement. This installation also exemplifies the cost saving that
can come from the planned use of porous pavements. Because this pavement absorbs
some of the runoff during storms, a reduction was allowed in the size of downstream
conveyances and detention basins, producing a net cost saving when the porous road is
properly counted as both a pavement structure and a functional part of the drainage
system.
Figure I.1.20 illustrates the highly selective use of porous materials that is vital for
maximizing porous pavements cost advantages and for designating appropriate application of porous pavements anywhere. This is a parking lot in Medford, NJ. It was built
with porous surfaces to satisfy the environmental protection objectives established in the
famous study of the township by Ian McHarg and his colleagues (Juneja, 1974). Impermeable asphalt is used in the parking lots traveling lane, where many vehicles travel and
26
Developed
Return of volume
to base flow
Rate
of flow
Natural
Time during storm

Figure I.1.21 The outcome of stormwater infiltration during an individual storm event.
turn each day. Impermeable asphalt is of course a well-known material and is a highly
conservative response to the traffic load here. In contrast, simple stone aggregate is used
in the individual parking stalls where there is only one car per day, standing still most of
the time. The low traffic load freed the designers to select this material, which is highly
permeable and cheaper per yd2 than any other paving material, porous or nonporous.
The Medford designers analysis of the detailed usage levels within this parking lot
exemplifies the care, knowledge, and ingenuity that designers need to use in applying
infiltration as they do in implementing any other feature of any project they have responsibility for. Stormwater infiltration should not be done in some places: for example, on
steep unstable slopes, on grossly permeable gravel soils where groundwater pollution
would be possible, and over toxic deposits in old industrial areas. Outside those local spots,
carefully implemented stormwater infiltration succeeds environmentally and economically.
Infiltration outcomes in restoration

During an individual storm, infiltration returns runoff water from the surface to the soil
like the flow shown in Figure I.1.21. The water is no longer part of the surface flow during
the storm event. Instead it goes into the groundwater to discharge later as part of the longterm base flow.
Figure I.1.22 shows the outcome of infiltrating water from a series of storms, large
and small, over the course of many years. This is the result of a modeling study for a high
school site in Georgia (Ferguson et al., 1991). The average annual inflow of precipitation
on this site amounts to 270 acre-ft. of water. Each column shows the disposition of that
volume in the sites environment from some way of managing the stormwater. On the left
is the disposition before development, when the site was entirely vegetated. Evapotranspiration was large. Of the stream flow, almost all was in steady, continuous base flow;
direct runoff occurred only during large storms, and even then only in moderation. In the
middle is the disposition from the developed site using surface stormwater discharge
including culverts and detention basins. The impervious cover reduced evapotranspiration. Direct runoff dominated annual stream flow, claiming away water that used to go
into both evapotranspiration and base flow.
On the right is the disposition from the developed site using infiltration basins. This
plan captures the direct runoff from the impervious surfaces and transforms it, through
infiltration, into base flow. Base flow is actually greater than it was before development,
because the impervious surfaces divert water away from evapotranspiration. The residual
direct runoff comes from peripheral impervious surfaces that could not direct their runoff
into infiltration basins. According to studies like this, infiltration transforms, relatively
Chapter I.1:
27
300
250
Annual volume (af)
Evapotranspiration
Evapotranspiration
Evapotranspiration
200
150
Direct runoff
100
50
Direct runoff
Base flow
Direct runoff
Base flow
Base flow
Before development
Surface discharge
Infiltration
Figure I.1.22 The contrasting outcomes of surface discharge and infiltration over the course of an
average year, for a high school site in Georgia.
speaking, the erosive filth of surface runoff into the resources of clean groundwater and
steady base flow.
In conclusion, we can make a distinction between two fundamental parts of urban
watersheds and the processes that go with them: the surface and the subsurface. A designer
has the choice of directing urban stormwater toward one or the other. Design for stormwater on the surface of urban watersheds manages excess runoff by treating its downstream symptoms. Surface conveyance and detention ultimately continue the discharge
of runoff to streams. Because they lie on the surface of landscapes, they are hydrologically
isolated and artificially incomplete.
In contrast, infiltration into the subsurface of landscapes is restoration that cures the
disease. Infiltration, unlike any surface approach to stormwater management, is capable,
within the limitations of specific sites, of solving all the problems of urban runoff: peak
flow, base flow, groundwater recharge and quality, because it calls on the power of the
underlying landscape. The subsurface is a natural resource for filtration and storage
waiting to be taken advantage of on essentially every site, as nature has always used it
in the past.
Infiltration deserves to be the beginning of stormwater design, the fundamental tool
for solving urban watershed problems, the approach that is implemented to its fullest
feasible site-specific extent before anything else is attempted. Water belongs in the subsurface is a simple principle that is supported by the fundamental way landscapes maintain themselves, by the good it can do for human beings, and by decades of documented
experience in thousands of field installations. The only part of stormwater we need to
manage is the part we are not able, for some site-specific reason, to restore through
infiltration.
28
Literature cited
Borgwardt, S., Versickerungsfahige Pflastersysteme aus Beton: Voraussetzungen, Anforderungen, Einsatz
(Infiltratable Concrete Block Pavement Systems: Prerequisites, Requirements, Applications), Bundesverband Deutsche Beton- und Fertigteilindustrie, Bonn, 1997.
Cahill, T., A second look at porous pavement/underground recharge, Watershed Protection Tech., 1,
76, 1994.
Ferguson, B.K., Land environments of water resource management, J. Environ. Syst., 14, 291, 1985.
Ferguson, B.K., Landscape hydrology, a component of landscape ecology, J. Environ. Syst., 21, 3, 193,
1992.
Ferguson, B.K., Stormwater Infiltration, Boca Raton: Lewis Publishers, Boca Raton, FL, 1994a.
Ferguson, B.K., The concept of landscape health, J. Environ. Manage., 40, 129, 1994b.
Ferguson, B.K., The maintenance of landscape health in the midst of land use change, J. Environ.
Manage., 48, 387, 1996.
Ferguson, B.K., Introduction to Stormwater: Concept, Purpose, Design, John Wiley & Sons, New York,
1998.
Ferguson, B.K. and Suckling, P.W., Changing rainfallrunoff relationships in the urbanizing
Peachtree Creek watershed, Atlanta, Georgia, Water Resour. Bull., 26, 313, 1990.
Ferguson, B.K., Ellington, M.M., and Gonnsen, P.R., Evaluation and control of the long-term water
balance on an urban development site, p. 217 of Proc. 1991 Georgia Water Resour. Conf.,
Kathryn J. Hatcher, ed., University of Georgia Institute of Natural Resources, Athens, GA,
1991.
Juneja, N., Medford: Performance Requirements for the Maintenance of Social Values Represented by the
Natural Environment of Medford Township, NJ, Univ. Pennsylvania Dept. Landscape Architecture and Regional Planning, Philadelphia, 1974.
Leopold, L.B., Water, Rivers and Creeks, University Science Books, Sausalito, CA, 1997.
Los Angeles County Department of Public Works, General information,
http://www.deltanet.com/sgvw/lacdpw/general/.
Thayer, R.L., Jr. and Westbrook, T., Open drainage systems for residential communities: Case studies
from Californias Central Valley, p. 152 in Water, Proc. 1989 Annual Conf. Council of Educators
in Landscape Architecture, Landscape Architecture Foundation, Washington, 1989.
U.S. Census Bureau, American Factfinder, http://factfinder.census.gov.
U.S. Weather Bureau, Rainfall Intensity-Duration-Frequency Curves, Technical Paper No. 25, U.S.
Weather Bureau, Washington, 1955.
Water Replenishment District of Southern California, What Is the WRD?,
http://www.wrd.org/broch.htm.
Chapter I.1:
29
Response
Stormwater infiltration:
Curing the disease rather than treating the symptoms
The title of one of the most useful design guidance manuals for stormwater quality
protection is Start at the Source. This is very much the objective expressed by Bruce
Ferguson in this chapter. As Ferguson states, the fundamental disease of urbanized watersheds is the sealing of ground surfaces by impervious cover, thereby causing runoff to
move across those surfaces and carry untreated contaminants directly into receiving
waters. Other chapters in this book address means to ameliorate or negate the effects of
impervious surfaces, and several chapters present approaches that move toward starting
at the source, including reducing impervious cover at the start as covered by Larry
Coffman in Chapter I.5 and Michael Clar in Chapter I.15, setting limits to the amount of
impervious cover as described by Jennifer Zielinski in Chapter I.3, and building bioretention swales or stormwater gardens as presented by Richard Pinkham and Timothy Collins
in Chapter I.4, Thomas Liptan and Robert Murase in Chapter I.6, and Robert France and
Philip Craul in Chapter I.7. Fergusons chapter adds to the discussion of water sensitive
design by proposing porous surfaces as viable alternatives to traditional impervious ones.
Ferguson reminds us that we need to regard landscapes as dynamic, evolving entities
whose functions shift through time. What becomes very clear in this chapter is the cardinal
requirement to understand processes in natural landscapes as the models upon which to
pattern new water sensitive designs during development. Even simple understanding
alone may be insufficient in this respect, for as Ferguson argues, it is imperative that we
obtain advanced predictive knowledge about the patterns and dynamics of rainfall events
for any local before we can generate truly effective, site-specific water sensitive designs.
This is an extremely important lesson for design professionals and is a direction that
should be embraced. For it is the shift from understanding to prediction that represents
an intellectual maturity in any scientific investigation. This too underscores an important
message that can be taken from Fergusons chapter. If water sensitive design is to be fully
integrated into landscape architecture, the profession needs to become much more
informed about how the hydrological world works in a predictive, scientific fashion.
Ferguson introduces a variety of alternative pavements in his chapter: interlinking
blocks, concrete with pore spaces, stone aggregates, etc. Also touched upon are some
concerns about site limitations for applying this technology, as for example, surface slopes
and underlying soils. What is very much needed is a comprehensive examination of the
relative strengths and weaknesses of these various approaches, followed through time.
How well, for instance, does porous pavement perform in cold climates a decade after its
installation?
One of the most fascinating aspects of this chapter is Fergusons use of health metaphors in describing the ailments of urban watersheds. This is an exciting and rapidly
expanding field of environmental scholarship and ecological ethics, and one to which
those practicing water sensitive design could benefit from deeper exposure to. If, to run
with Fergusons analogy here, urban watersheds can be regarded as being diseased, can
we then look upon the symptoms of that disease, i.e., surface runoff, as a form of hemorrhaging?
chapter I.2
Successful stormwater management

ponds (Massachusetts)
Desheng Wang
Abstract
Stormwater management ponds have been one of the most widely used techniques in the
history of flood control. Because of the growing knowledge of stormwater quality and its
impact on natural resources such as water, wetlands, and wildlife habitat, federal, state,
and local government agencies are tightening their stormwater management regulations.
In this regulatory framework, stormwater management pond design naturally becomes
more sophisticated, requiring a better understanding of the complicated soil and water
conditions and different development needs of each project site. Therefore, designing a
successful stormwater management pond becomes more challenging for every land development engineer. This chapter presents a framework extracted from design and review
experience of over a hundred stormwater management ponds for industrial, commercial,
and residential developments in the past five years. The discussions include (a) establishing design goals, (b) acquiring critical site information, (c) understanding the components
of stormwater management ponds, and (d) proposing some thoughts for future design
and study. Several projects are discussed to illustrate the scheme of designing successful
stormwater management ponds, which includes
A site where little space is left for a pond
A 3:1 (H:V) slope of a pond that failed
Stopping the discharge of muddy water from the pond
Introduction
Civil engineers and hydrologists would naturally think about detention/retention ponds
or basins when they face stormwater problems. It is not surprising because stormwater
management ponds are one of the most widely used techniques in the history of flood
control. Stormwater pond design, however, is no longer only flood control oriented. In
the past 20 years, it has been realized that water quality impacts of stormwater discharges
on receiving waters are significant. The primary sources of area-wide (nonpoint source)
pollutants are surface runoff from agricultural and urban watersheds (Loganathan et al.,
1994). Many research and field studies were devoted to understand and enhance stormwater management systems, especially the pollution removal processes (Whipple et al.,
1-56670-562-2/02/$0.00+$1.50
31
32
1980; EPA, 1986; Schueler, 1987, 1995a,b; Urbonas and Stahre, 1993). This resulted in
guidelines for stormwater discharge permits in the National Pollutant Discharge Elimination System (NPDES) program (Code of Federal Regulations (CFR), 1994). A more
stringent Stormwater Phase II rule of NPDES was published on December 8, 1999, and
will become effective by early 2003. More information in this regard can be obtained at
the Web site http://www.epa.gov/owm/sw/phase2. The stormwater quality control trend is also
reflected in recent state and local environmental regulations and policies (MA DEP, 1997).
In current best management practices (BMPs), stormwater detention basins function
as pollutant-trapping units by enhancing the detention time for pollutant settling. This is
one of the most widely used methods due to pond reliability and effectiveness (Schueler,
1987). Whereas the method for designing detention ponds for flood control is relatively
simple and well developed, the method for designing stormwater quality detention ponds
is more complicated and relatively undeveloped (Wang and Carr, 1996). In addition, most
stormwater ponds will finally end up as a seasonal or a perennial pond. Designers should
know how the pond conditions will impact ecological systems, especially wildlife. These
impacts are little known to the engineering community and are not adequately reflected
in current stormwater management regulations.
This chapter will not address all stormwater pond design issues. Instead, it will try
to present a systematic approach to the design of successful stormwater management
ponds based on first-hand design experience and field observations of as-built ponds and
basins. Then, a simple sequence of designing successful stormwater ponds used by the
author will be discussed including planning, critical data acquisition, pond configuration,
and outflow control structure. A few prototype stormwater management pond designs
will be used to illustrate the discussion. More stormwater management design tips can
be found in other chapters (I.1, I.6, I.7, I.8, I.11, and I.14). Further research needs to
understand and enhance stormwater pond design are also discussed.
Functions of a stormwater pond

Humans create different kinds of ponds for different purposes: farm ponds, fish and
wildlife ponds, water supply ponds, recreation/swimming ponds, landscape ponds, and
stormwater/flood control ponds, which can be seasonal or perennial. Manmade pond
history can be traced back thousands of years. For a stormwater management engineer
facing this ancient topic in the new era, a question is constantly raised: how to design a
successful stormwater management or stormwater BMP pond? As mentioned previously,
stormwater management pond design has evolved from quantity oriented toward both
quantity and quality oriented in the past 20 years (Whipple and Hunter, 1980; Schueler,
1987; Wang and Carr, 1996). Before the discussion gets into how to, it will be helpful to
look a little harder and deeper into what a stormwater pond can do for our environmental
or ecological system.
Based on the aforementioned functions of manmade ponds, a stormwater pond may
provide three basic functions: (1) flood control; (2) water quality enhancement; (3) ecological and aesthetic values. There will rarely be any disputes on the first two functions. The
ecological and aesthetic functions of stormwater ponds can be controversial, however, and
difficult to define due to different land uses and owners lacking quantitative evaluation
methods. Today, this issue can no longer be overlooked because wildlife issues are receiving more attention in the permitting process. The pond aesthetics and recreational value are
likely reflected in the sale and purchase process of a house that is located close to a stormwater
pond. Based on the authors design and permitting experiences and field observations of
Chapter I.2:
Successful stormwater management ponds (Massachusetts)
33
as-built stormwater ponds, the following ecological and aesthetic functions are commonly
observed:
1. Most stormwater basins with permanent pools attract wildlife and waterfowl
including but not limited to wild ducks, frogs, turtles, spring peepers, and other
macro-invertebrates.
2. Stormwater ponds with proper landscaping are usually attractive and may add
value to commercial and industrial land.
3. The groundwater is recharged.
4. Stormwater ponds in residential subdivisions tend to trigger safety and mosquito
concerns from neighbors.
5. Even stormwater ponds with temporary water provide some wildlife habitat value
regardless of design purpose.
In general, therefore, a successful stormwater pond can provide optimal functions of
flood control, stormwater quality enhancement, and ecological and aesthetic values for
the environment. The following sections further discuss how to design a successful stormwater pond.
Designing a successful stormwater pond

Establishing design goals
As we know, there are different regulations from local, state, and federal governments for
todays stormwater management system design. The most practical goal for designing
successful stormwater ponds is to help a project meet these regulatory requirements. For
example, when an engineer designs a stormwater management system in Massachusetts,
he or she will have to meet nine stormwater management standards including (MA DEP,
1997):
1. No direct untreated discharge of stormwater
2. No increase in peak discharge rates (2-, 10-, and 100-year storms)
3. Compensation for loss of annual groundwater recharge due to development
through infiltration measures to maximum extent practicable
4. Eighty percent removal of annual total suspended solids for new development
5. Special treatment of discharge from hot spots (auto service, commercial parking,
etc.)
6. Special protection of critical resources (outstanding resource waters, shellfish beds,
swimming beaches, coldwater fisheries, and recharge areas for public water supplies)
7. To the maximum extent practicable, application of the above standards for redevelopment of previously developed sites
8. Erosion maintenance and sediment control during construction and land disturbance
9. Providing an operation and maintenance plan to ensure that system function will
be maintained as designed
To design a successful stormwater pond means meeting all regulatory requirements.
Therefore, a successful stormwater pond cannot stand alone without other auxiliary components including planning, pretreatment, and operation and maintenance. In real life,
lots of theoretically successfully designed ponds failed due to incorrect or missing site
34
Table I.2.1 Stormwater Pond Design Data

Design Goals
Design Data
Pond Factors
Flood control
Protected important resources in the

watershed (wetlands, rivers,
archaeological sites, wildlife habitat,
water supply resources)
Watershed topography
Up-to-date precipitation data
Existing and proposed land uses
Groundwater conditions and soil
infiltration rate
Pollutant sources in the watershed
(sediment, hydrocarbons, metals,
nutrients, bacteria, etc.)
Characteristics of pollutants
(concentration, particle size, and
distribution in runoff)
Wildlife and public safety
Recreation and landscape
Water budget (precipitation,
evapotranspiration, etc.)
Location
Volume
Outflow control
Emergency spillway
Energy dissipation aprons at
outfalls
Water quality
Ecological impacts
Pretreatment (catchbasin, water

quality inlets, and swales)
In-pond features (sediment
forebay, micropool, side slopes,
planting)
Flow distribution
Pondscape
Landscape
Food and water for wildlife
information. Some similar ponds in similarly sized watersheds can give quite different
performance results (JSB and TRS, 1997).
The author inspected two ponds in northeast Massachusetts that failed during the
early days of completion in the wintertime. The two ponds are located in different towns
and were designed by different engineers. Both ponds did not provide a proper emergency
spillway and were based on incorrect high groundwater and infiltration information. One
pond ended up using significant storage volume below the groundwater table for flood
control, which was overtopped during a 2-in. rain storm over frozen ground. The other
pond was built above the actual high groundwater table, however, there was little depth
below the outflow invert. Kids plugged the single-pipe outlet to create a skating rink,
which also created an overtopping during a earlier spring melt. Both cases caused thousands of dollars in retrofitting and litigation enforcement fees. These case histories pointed
to other important issues for successful stormwater ponds design: (1) design data acquisition and (2) maintenance-effective outflow control structures.
Critical data acquisition

As mentioned, a successful stormwater pond depends on a successful design after design
goals are clearly defined. A successful design is the bridge between design goals and the
materialized pond. On the other hand, the design data are the foundation of the design.
A good design theory without complete and quality design data is foolish and greatly
increases the chances of performance failing to meet design goals. When a project is
undergoing review processes, incomplete design data not only delay the review process,
but also make the project more expensive when actual data require changes in design. If
a design with erroneous critical data luckily had passed review, the project would fail
and be detrimental to downgradient resources. Therefore, design data acquisition is as
important as design itself. In order to present the design data with a clear and concise
relationship to design goals and pond features, the design data are divided into three
groups: flood control, water quality, and ecological impacts, as shown in Table I.2.1. More
detailed data information related to pond function design will be discussed.
Chapter I.2:
35
Pond function design

Flood control
In general, in order to achieve flood control purpose, a stormwater pond shall be able to
collect as much runoff as possible and as effectively as possible. The best place to locate
a stormwater pond in a watershed is where surface runoff naturally goes. This will
probably be the most effective way to collect runoff. In some cases, however, the best place
may be occupied by other important resources: wetlands, endangered wildlife species
habitat, or archaeological sites. Whenever a project comes across these issues, alternative
locations or a mitigation plan would be required. Facing this situation, it is worth asking
a question: Can a stormwater pond enhance the resource value (except for an archaeological site)? It is almost a no from a regulatory point of view at the current time in
Massachusetts, but it is a no that lacks a good understanding and appreciation of what
can be achieved in manmade systems. However, it is still a question for researchers to
pursue and for a project team to evaluate, which will provide a significant benefit in the
future design.
Land-use change is almost the most important factor triggering flood control in land
development and urbanization. Different land uses generate different runoff volumes and
rates. In the process of determining the total runoff, a representative runoff parameter is
usually calculated for a combination of different land uses. When applying TR-55 or TR20, the most popular methods in urban hydrology, the representative parameter is called
the curve number (CN). A composite CN value is commonly calculated for different land
use. When CN numbers for different land uses are significantly different, such as hydrological class A woods and sewered roadway, a simple weighted average CN number will
significantly undervalue both runoff volume and peak rate, which can cause significant
error in flood control calculation, even failure due to embankment overtopping. For such
cases, the watershed should be divided into more subcatchments with relatively uniform
CN numbers.
Groundwater conditions and the infiltration rate through pond side and bottom surface are critical in the determination of flood control volume and water budget. The latter
can also impact water quality and ecological functions. Groundwater table fluctuations in
the pond area can also cause slope failure, which will be further discussed in later case
studies.
Water quality
The water quality function of stormwater ponds is commonly evaluated by removal
efficiencies of pollutants. Different pollutants correspond to different removal rates or
efficiencies. There are different methods to predict pollutant removal rates. Some methods
are based on more sophisticated theory (Vittal and Raghav, 1997; Whipple and Hunter,
1980; Camp, 1964; Vanoni, 1975; Garde et al., 1990). Some are empirical (Wang and Carr,
1996; Meadows and Kollitz, 1995; Loganathan et al., 1994). Given so many parameters
involved in real-life cases, it is not necessary that a theoretical method is better than an
empirical method (JSB and TRS, 1997; Wang and Carr, 1996). More than a dozen parameters
affect pollutant removal rates of a pond (Wang and Carr, 1996). Understanding what each
parameter contributes to the pollutant removal can be helpful to simplify calculations and
provide better guidelines in stormwater pond design. The main mechanism of pollutant
removal in stormwater ponds is sedimentation, which has a more complete theoretical
description. The three most important factors in sediment removal processes in a stormwater pond are (a) sediment size and particle distribution, which can be represented by
settling velocity (); (b) pond detention time or residence time (td); (c) suspension or
reentrainment factor (/u*), where u* is shear velocity of flow in the pond. As a rule of
36
thumb, the removal rate of pollutant is proportional to and td, but inversely proportional
to u*. Then, the pond dimensions (width, depth, and length) can be determined based on
these parameters for a required removal efficiency. As we can see, however, the sediment
size and distribution would be very difficult data to collect for a designed site condition.
Therefore, some theoretical statistical data will be valuable. Even though sediment size
and distribution may change site to site, studies by different researchers revealed that
particle size in stormwater mostly fall in the range of 10 to 35 m (Urbonas and Stahre,
1993). If available, local data will help yield a more precise design. A significant effort is
currently ongoing to build a nationwide database for different stormwater BMPs (ASCE
et al., 1999). In real situations, the annual average pollutant removal is more important.
It will be very tedious and unrealistic for a design engineer to calculate the annual average
removal rate using theoretical methods and each rainfall amount, especially for small
projects. Therefore, some simple methods first flush (0.5" or 1") treatment volume,
graphics method, and empirical formula are more widely adopted in stormwater management policy and design guidelines (MA DEP, 1997; Schueler, 1987; Wang and Carr,
1996). Simple methods are usually accompanied by some general recommendations and
requirements, which result from more detailed studies (MA DEP, 1997; Schueler, 1987). Some
important considerations to enhance the pollutant removal rate include the following:
1. Detention time would be better between 24 h and 40 h. When in permeable soil,
the time can be shorter depending on the infiltration rate. The larger the infiltrate
rate, the smaller the detention time will be. Therefore, reliable infiltration rate
calculation methods using accurate in-site permeability test data are necessary to
achieve practically accurate groundwater recharge and water quality benefit due
to infiltration in stormwater ponds or basins (Wang, 1999).
2. Flow distribution shall be made as uniform as possible across the pond section,
which is perpendicular to the flow direction. This can be facilitated by shaping the
pond as an ellipsis, and installing flow distribution berms at the inlet and outlet,
a sediment forebay at the inlet, and a high marsh between the inlet and outlet.
3. Flow depth shall be no less than 1 ft for an annual mean storm event.
4. Maximize the distance between inlet(s) and outlet(s).
5. Minimize clean water through pond, such as roof runoff to maximize treatment
for dirty water, such as roadway runoff.
Ecological and safety impacts

As observed in the field, a stormwater pond will likely become some kind of habitat for
wildlife and waterfowl. The water quality in the pond itself can be important to wildlife
coming to use the pond, which somehow is contradictory to the stormwater quality treatment
function of the pond design. To minimize this conflict, the following are recommended:
1. Stormwater runoff shall undergo pretreatment as much as possible before being
discharged to the pond. Pretreatment includes deep sump catchbasins, oil/grit
separators or water quality inlet, vegetation strips, and grass swales.
2. Pretreatment shall be installed offline unless detailed design requires online installation.
3. Pollutant tolerant vegetation shall be chosen for pondscape, such as cattail, sedges,
and bulrushes, which can enhance in-pond water quality.
4. A safety leveled bench (5 to 10 ft) can be installed at a flow depth of 3 ft to minimize
public hazard.
5. A minimum 3:1 slope shall be used.
Chapter I.2:
37
Figure I.2.1a Schematic detention pond design features. (From Schueler, T., Controlling Urban Runoff:
A Practical Manual for Planning and Designing Urban BMPs, Dept. of Environmental Programs,
Metropolitan Washington Council of Governments, Washington D.C., 1987. With permission.)
6. A 1- to 2-ft freeboard should be allowed for the maximum design storm (100-year)
depending on the pond size. The larger the pond, the larger the freeboard should be.
7. Whenever the groundwater table will be penetrated, a slope stability analysis
should be conducted to determine if the proposed slope is stable and slope stabilization is needed. For such a condition, a perimeter drain installed above the high
groundwater table can increase slope stability.
A schematic stormwater pond is presented in Figure I.2.1a. Another public concern
about stormwater ponds is mosquito breeding. In general, stormwater ponds do not have
a high mosquito nuisance level compared with wastewater treatment ponds due to relatively low nutrient loads and more frequent replenishment. In case mosquitoes are breeding in the pond, many methods are available to control them, including biological control
and chemical control. Biological controls are preferred because the biological controls
specifically target mosquito larvae and are harmless to humans, unlike many chemicals
even at standard doses (McLean, 1995).
In addition, a maintenance-effective outflow control structure is also important to
assure the design function of the pond, especially the flood control function, which should
have the following features:
Multilevel discharge points, which can provide long enough detention time (>24 h)
for storms less than a 2-year event
A protection component, which can prevent debris and vandalism from interfering
with the structure design function
An overflow path for larger storm events (larger than a 50-year storm)
38
Figure I.2.1b A maintenance-effective outflow control for detention ponds.
Figure I.2.1b presents an outflow control structure based on the above criteria that has
been proved effective in reality.
Roof detention
When space and soil are very limited for creating stormwater pond and subsurface infiltration trenches, the author has used a roof detention method for stormwater management
to compensate for a smaller in-ground stormwater pond or subsurface detention. The
mechanisms include ballast roof and water dams around roof drains, which can provide
more storage and prolong the time of concentration for roof runoff and reduce the peak
discharge rate. This method can only be applied to a relatively flat roof condition. The
time of detention should be controlled within 6 h. The load of the water due to detention
should be communicated to the architect and structure engineer for a corresponding roof
supporting structure design. Based on design experience in Massachusetts, roof load due
to detained water does not significantly increase cost in other components. One reason is
that the roof has to be designed for significant snow load. The other advantage of roof
detention is to reduce the number and pipe size of roof drains due to the reduced roof
peak discharge rate. When applying this method to areas without snow load, costs may
rise. Roof detention may be a new concept, however, in real situations, ballast roofs have
Chapter I.2:
39
Figure I.2.2 Drainage System for Case I combination of roof, subsurface, and open pond detention, Mansfield, MA.
been used to increase the roof stability for buildings along coasts subject to strong wind.
At the same time, the ballast roof can slow down the roof runoff and provide flood control
benefit. See Figure I.2.3a for a real-life ballast roof.
Project illustration
In order to demonstrate the importance of design data and design concept, three projects
are briefly introduced and the highlights of each project are discussed.
40
Case I: A site where little space is left for a pond

A manufacturer needs to expand its factory on the same lot with limited space abutting
bordering vegetated wetland in a industrial park in Mansfield, MA. The site conditions
and project goals are as follows:
Total space available:
Depth to high groundwater:
On-site soil:
On-site utility lines:
Building addition:
Space for pond:
Effective depth for flood control:
Effective flood storage:
Flood storage need:
3.34 acres
7 ft
Silty with very low permeability (5.19 106 ft/s)
Subsurface electric, phone, and gas lines
2.36 acres (70% of available space)
190 ft 47 ft
1.5 ft due to the limits to inflow and outflow inverts
6,000 ft3
16,000 ft3
As can be seen, 70% of the site will be building. Not enough space is available to
create a large enough pond to achieve flood control. Low permeability does not allow the
effective use of subsurface infiltration trenches. To overcome the difficulties, a small pond
was created for stormwater quality and landscape purposes, which worked successfully
together with roof detention and a small amount of subsurface detention to serve the
overall stormwater management requirement. See Figure I.2.2 for details. Roof detention
is achieved by a series of 4-in. PVC pipes fixed on the roof around each roof drain. Onein. orifices are drilled every 10 ft along the pipe bottom. The roof detention time was
designed as 4 h, which compares to 5 min without the pipe check dams. See details of the
water dam of roof detention in Figure I.2.3b.
Case II: A 3:1 (H:V) slope of a pond that failed

A detention pond (180 ft by 60 ft) in a residential subdivision in Marlboro, MA, seemed
designed and constructed according to all MADEP criteria: 3:1 pond length-to-width ratio;
3: 1 side slope; sediment forebay at inlet and micropool at outlet; high marsh (cattail)
between sediment forebay and micropool; maintenance-effective outflow control structures; and energy dissipation aprons at inlet and outlet. Water coming out of the detention
pond looked crystal clean (see pictures in Figure I.2.4a). The pond was built in the summer
of 1997 by excavating in a 30-ft drop on a 10% slope. The maximum excavation was about
12 ft. The downgradient wetland is at an elevation similar to the bottom of the sediment
forebay and micropool. In November 1997, part of the upgradient side of the pond
slumped, which was simply fixed in the summer of 1998. By the late fall of 1998, the slope
failed again. Then, the town engineer and conservation commission got worried and
demanded a professional fix. The owner finally contacted the author in the spring of 1999.
After a site inspection, the following was observed:
The failure was a shallow, noncircular slide.

Groundwater broke out at the upper limit of the failure.
Soils in slope were a silty, sandy loam.
A drainage channel was located on top of the slope running toward the corner
with the failed slope.
Based on the information, it was concluded that the failure was caused by a large
groundwater fluctuation (about 5 ft). When groundwater raises, it increases the pore
Chapter I.2:
41
Figure I.2.3a Ballast roof, Milton, MA.
Figure I.2.3b Water dam roof detention, Mansfield, MA.
Figure I.2.4a Distant view of a detention pond with upper left slope slumped, Marlboro, MA.
pressure and reduces the soil strength against the gravity. The mitigation plan is as follows:
a perimeter drain consisting of a 6-in. crushed stone blanket placed on the failed slope
from the limit of failure to about 2 ft above the normal pond level and a 4-in. perforated
pipe installed at the bottom of the trench. The crushed stone and drain pipe are underlain
by filter fabric. Along the bottom of the drain, some 15-in. to 24-in. key stones are placed
to increase the stability (see Figure I.2.4d).
42
Figure I.2.4b A close look of the outflow control structure in Figure I.2.4a.
Figure I.2.4c A close look of the slumped slope.
Figure I.2.4d Slope in Figure I.2.4c after restoration.
Case III: Stop the discharge of muddy water from the pond
On April 12, 1995, Carr Research Laboratory (CRL) received a call from a land developer
in Merrimac, MA. A detention pond in his subdivision under construction was discharging
muddy water to downgradient wetland and a neighbors pond. The DEP issued an
enforcement order to stop all construction activities unless the problem was solved. The
CRL staff went to inspect the site the same afternoon:
Chapter I.2:
43
Figure I.2.5a A detention pond as observed, Merrimac, MA, April 1995.
The soils in the watershed of the detention basin consisted of silty, sandy loam.
The cause of muddy water was due to runoff from the disturbed lot.
The detention basin is about 270 ft 108 ft.
Two inlets to the pond were observed: an 18-in. RCP and a 24-in. RCP.
Only the 18-in. pipe was discharging water into the pond.
Only a 12-in. PVC outlet with an invert at the pond bottom was observed, which
was connected to an 18-in. level spreader covered with pea stone to the southwest
of the pond.
A sediment deposition delta was observed above pond level below the 18-in. inlet.
Riprap aprons were at the inlets and outlet.
No trash rack was observed at the outlet.
No emergency spillway was observed for the pond.
The inlets and outlet were installed across the width of the pond, which shortcircuits the flow.
Neither sediment forebay nor micropool was created in the pond bottom.
The average water depth in the pond was less than 0.6 ft.
Withered cattails spread in the bottom of the pond.
An overtopping of the pond embankment had taken place and caused erosion and
siltation in an area 80-ft long by 70-ft wide and about 0.5 ft deep in the downgradient wetland.
Not enough sediment source control existed in the subdivision. All temporary
disturbed lots were barren.
The overtopping was caused by a blockage of the outlet, which children had done to
create a skating rink. Because there was no designed emergency spillway, the water rose
to a low spot of the earth embankment. Figure I.2.5a shows the pond condition as
observed. The pond was obviously not designed according to stormwater BMP criteria.
Not enough stormwater treatment volume was designed to remove enough silt and clay
sediment in the incoming water before discharging it to the downgradient. Based on
observed information, a retrofit plan was created and approved by the MA DEP and the
Town Conservation Commission. The highlights of the retrofit plan of the pond are as
follows:
44
Figure I.2.5b Pond in Figure I.2.5a after retrofit, September 1995.
Remove sediment deposition at the inlet.

Install a 1-ft-tall stone flow distributor at the inlets of the pond.
Excavate a micropool at the outlet and install a hay bale dam around it to increase
the normal pool level for more water treatment volume.
Install a trash rack at the outlet to prevent debris or humans from blocking it.
Restore the eroded embankment and install an emergency spillway.
Install a series of check dams along the slope at the end of the outfall of the outflow
pipe.
See Figure I.2.5b for the retrofitted pond conditions. The results turned out to be
satisfactory.
Summary and conclusions

This chapter illustrates a design philosophy for stormwater ponds. A stormwater pond is
only a component in one stormwater management system. A successful pond design can
only be achieved through successful stormwater management system planning, acquisition of important data, clear pond function definition, and a sound pond structure design,
which provides the best pond function in the most cost-effective way.
The chapter also introduced a new concept: roof detention and its advantages and
limits. The use of roof detention can be critical for some site conditions to make an effective
stormwater management system practical.
Almost all stormwater ponds provide some wildlife habitat value to the ecological
system. It is not clear how the pond water quality will impact the wildlife community
using the pond, which is critical in evaluating future design functions of the stormwater
ponds. More field data and studies would be needed in this area. In the meantime, it is
recommended that stormwater from pollutant-laden areas should be pretreated as much
as possible before being discharged to stormwater ponds, which is important to the
wildlife community.
Chapter I.2:
45
Acknowledgments
The author is thankful to Dr. Jerome Carr, Carr Research Laboratory, Inc. for his review
and useful comments and for many opportunities provided to test design strategies, and
Pamela Carlson for her meticulous help in the preparation of figures. Special thanks to
my wife, Amy, who made this work possible, for her understanding and support during
many nights and weekends.
Literature cited
ASCE and EPA, Users Guide National Stormwater Best Management Practices (BMP) Database,
Version 1.0, prepared by Urban Water Resources Research Council (UWRRC) of ASCE,
Wright Water Engineering, Inc., Urban Drainage and Flood Control District, URS Griner
Woodward Clyde, and Office of Water, U.S. Environmental Protection Agency, Washington,
DC, 1999.
Camp, T.R., Sedimentation and design of settling tanks, Trans. ASCE, 111, 895, 1964.
Cheng, N.S., Simplified settling velocity formula for sediment, J. Hydr. Eng., 123, 149, 1997.
Code of Federal Regulations (CFR), The National Pollutant Discharge Elimination System, 40 CFR 122,
U.S. Government Printing Office, Washington, DC, 1994.
Garde, R.J., Ranga Raju, K.G., and Sujudi, A.W.R., Design of settling basins, J. Hydr. Eng., 18, 81, 1990.
JSB and TRS, A tale of two regional wet extended detention ponds, Technical Note 98, Watershed
Protection Tech., 2, 529, 1997.
Loganathan, G.V., Watkins, E.W., and Kibler, D.F., Sizing stormwater detention basins for pollutant
removal, J. Environ. Eng., 120, 138, 1994.
MA DEP, Stormwater Management Stormwater Policy Handbook, prepared by MA Department of
Environmental Protection and MA Office of Coastal Zone Management, March 1997.
McLean, L., Mosquitoes in constructed wetlands A management bugaboo, Watershed Protection
Tech., 1, Center for Watershed Protection, 203, 1995.
Meadows, M.E. and Kollitz, M.M., Sediment pond trapping efficiency curves, Water Resources
Engineering: Proceedings of the First International Conference, San Antonio, Texas, Vol. 2,
12581262, August 1418, 1995.
Reed, S.C., Natural Systems for Wastewater Treatment Manual of Practice FD-16, Water Pollution
Control Federation, 1990.
Schueler, T., Controlling Urban Runoff: A Practical Manual for Planning and Designing Urban BMPs,
Dept. of Environmental Programs, Metropolitan Washington Council of Governments, Washington DC, 1987.
Schueler, T., Design of Stormwater Wetland Systems: Guidelines for Creating Diverse and Effective Stormwater Wetland Systems in the Mid-Atlantic Region, Anacostia Restoration Team, Dept. of Environmental Programs, Metropolitan Washington Council of Governments, Washington DC,
1992.
Schueler, T.R., Performance of a stormwater pond/wetland system in Colorado, Watershed Protection
Tech., 1, Center for Watershed Protection, 68, 1995a.
Schueler, T.R., Pollutant dynamics of pond muck, Watershed Protection Tech., 1, Center for Watershed
Protection, 39, 1995b.
UDFCD, Douglas County Storm Drainage Design and Technical Criteria, prepared by WRC Engineers,
Inc., under a contract to Urban Drainage and Flood Control District, Denver, CO, Jan. 1986.
Urbonas, B. and Stahre, P., Stormwater Best Management Practices and Detention for Water Quality,
Drainage, and CSO Management, PTR Prentice Hall, Englewood Cliffs, NJ, 1993.
U.S. Environmental Protection Agency (EPA), Methodology for Analysis of Detention Basins for Control
of Urban Runoff Quality, Nonpoint Source Branch, Office of Water, Washington, DC, EPA440-5-87-001, 1986.
Vanoni, V.A. (ed.), Sedimentation Engineering, Manuals and Rep. on Eng. Pract., ASCE, New York, 54,
p. 582, 1975.
46
Vittal, N. and Raghav, M.S., Design of single-chamber settling basins, J. Hyd. Eng., ASCE, 123, 469,
1997.
Wang, D.S., A simple mathematical model for infiltration BMP design, J. Hydrol. Sci. Tech., November,
117, 1999.
Wang, D.S. and Carr, B.J., Pollutant removal rates for stormwater detention ponds, Proc. of 1996 AIH
Annual Meeting, Boston, AMBP 1221, 1996.
Whipple, William, Jr. and Hunter, Joseph V., Detention basin settleability of urban runoff pollution,
Final Technical Completion Report for Project A-058-NJ, Water Resources Resource Research
Institute, Rutgers University, New Brunswick, NJ, 1980.
Chapter I.2:
47
Response
Centralized stormwater treatment:
Improving performance through engineering design
Detention ponds continue to be one of the most widely used techniques in stormwater
management. As Desheng Wang describes, due to a tightening of stormwater regulations,
detention pond design has become more sophisticated, requiring better understanding of
complicated soil and water conditions and different development needs of particular sites.
In particular, the most notable effect has been a shift from designing such ponds solely
for purposes of flood control, to more sophisticated water sensitive designs that improve
stormwater quality while simultaneously producing ecological benefits such as wildlife
attraction and groundwater recharge.
In this chapter, Wang outlines a systematic approach to the design of successful
stormwater ponds that includes planning, critical data acquisition, consideration of pond
configuration, and attention to outflow control structures. He also stresses the need for
auxiliary components such as watershed planning beyond the specifics of the site, pretreatment to reduce the contamination of inflowing stormwater (and thereby reduce the
opportunities for bioaccumulation by resident wildlife), and a strategic plan for operations
and maintenance. By establishing clear design goals, and by fitting the pond into the
topographic idiosyncrasies of the site rather than being imposed against them, it is possible
for such stormwater management controls to have a strong positive role in boosting nearby
property values while improving environmental conditions.
chapter I.3
Open spaces and impervious surfaces:

Model development principles
and benefits
Abstract
Recent research has demonstrated that subwatershed impervious cover has a strong influence on the quality of streams and lakes, and that impairment occurs at levels as low as
5 to 15%. Most current suburban development patterns, or the status quo, create impervious cover that equals or exceeds this level. This chapter reviews the potential of better
site design techniques to minimize the impervious cover created by new development,
conserve open space, and more effectively manage stormwater at development sites. A
series of 22 model development principles developed by a consortium of planning, road,
banking, engineering, development, and public safety organizations are described along
with an overview of the research data on economic, market, legal, safety and social benefits
of each. The environmental benefits of applying the principles will be demonstrated in a
series of redesign comparisons of residential development projects. The redesign comparisons document changes in runoff rates and nutrient export between the status quo and
subdivision designs incorporating the principles. The comparisons consistently indicate
that application of the model development principles significantly reduces impervious
cover, runoff, nutrient loads, and construction cost.
Introduction
Though they may not realize it, site planners have an excellent opportunity to reduce
stormwater runoff and pollutant export simply by changing the way they lay out new
residential subdivisions. Planners that employ open space design techniques can collectively reduce the amount of impervious cover, increase the amount of natural land conserved, and improve the performance of stormwater treatment practices at new residential
developments.
Simply put, open space designs concentrate density on one portion of a site in order
to conserve open space elsewhere by relaxing lot sizes, frontages, road sections, and other
subdivision geometry. Although site designs that employ these techniques go by many
different names, such as clustering or conservation design, they all incorporate some or
all of the following better site design techniques:
1-56670-562-2/02/$0.00+$1.50
49
50

Using narrower, shorter streets and rights-of-way
Applying smaller lots and setbacks, and narrow frontages to preserve significant
open space
Reducing the amount of site area devoted to residential lawns
Spreading stormwater runoff over pervious surfaces
Using open channels rather than curb and gutters
Protecting stream buffers
Enhancing the performance of septic systems, when applicable
In this chapter, I describe some of the benefits of employing better site design techniques as they apply to residential subdivisions. The analysis utilizes a simple spreadsheet
computer model to compare actual residential sites constructed in the 1990s using conventional design techniques with the same sites redesigned utilizing better site design
techniques. For each development scenario, site characteristics such as total impervious
and vegetative cover, infrastructure quantities, and type of stormwater management practice are estimated.
The Simplified Urban Nutrient Output Model (SUNOM) was used to perform a
comparative analysis for two subdivisions. The first is a large-lot subdivision known as
Duck Crossing, and the second is a medium-density subdivision known as Stonehill
Estates. In each case, the model was used to simulate five different development scenarios:
Predeveloped conditions
Conventional design without stormwater practices
Conventional design with stormwater practices
Open space design without stormwater practices
Open space design with stormwater practices
This chapter compares the hydrology, nutrient export, and development cost for these
sites under both conventional and open space design, and with and without stormwater
treatment. The article also summarizes other research on the benefits of open space design
and discusses the implications it can have for the watershed manager.
Duck Crossing a low-density residential subdivision

Duck Crossing is a large-lot residential development located in Wicomico County on
Marylands Eastern Shore. Prior to development, the low-gradient coastal plain site contained a mix of tidal and nontidal wetlands, natural forest, and meadow (Figure I.3.1). Its
sandy soils were highly permeable (hydrologic soil group A). Three existing homes were
located on the parcel, which relied on septic systems for on-site sewage disposal. The
existing septic systems discharged a considerable nutrient load to shallow groundwater.
A conventional large-lot subdivision of eight single-family homes was constructed on
the 24-acre site in the early 1990s. The subdivision is reasonably typical of rural residential
development along the Chesapeake Bay waterfront during this era (Figure I.3.2). Each
new lot ranged from three to five acres in size and was set back several hundred feet from
an access road. The access road was 30 ft wide and terminated in a large-diameter cul-desac. Sidewalks were located on both sides of the street. Each lot was served by a conventional septic system with a primary and reserve field of about 10,000 ft2. Stormwater
management consisted of curb and gutters that conveyed runoff into a storm drain system
that, in turn, discharged to a small dry pond (designed for the water quality volume, only).
Chapter I.3:
Open spaces and impervious surfaces
51
Figure I.3.1 Predevelopment conditions at the Duck Crossing site. (From Zielinski, J.A., The benefits
of better site design in residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.)
Figure I.3.2 The low-density conventional subdivision built at Duck Crossing (eight lots). (From
Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed Protection
Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.)
The entire site was privately owned, with the exception of the tidal marsh, which was
protected under state and federal wetland laws and represented the only common open
space on the site. As a result of construction, the existing meadow was entirely converted
to lawn, and the impervious cover for the site increased to slightly over 8%.
52
Figure I.3.3 The open space subdivision that could have been built at Duck Crossing (eight lots).
(From Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed
Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.)
Open space design for Duck Crossing

The critical ingredient of the open space redesign was a reduction in lot size from several
acres to about 30,000 ft2. This enabled about 74% of the site to be protected and managed
as common open space, which included most of the existing forest, wetlands and meadow
(Figure I.3.3). Consequently, only 19% of the site was managed as turf, nearly all of which
was located on the private lots.
The open space redesign at Duck Crossing also incorporated a narrower access road
(20 ft wide) along with shorter, shared driveways that served six of the eight lots. The
road turnaround was designed as a loop rather than a cul-de-sac bulb. Also, a wood chip
trail system was provided through the open space instead of sidewalks along the road.
Each home site was carefully located away from sensitive natural areas and the 100-year
flood plain. Taken together, these better site design techniques reduced impervious cover
for the site by about a third compared to the conventional design (from 8 to 5%).
The redesigned stormwater conveyance system utilized dry swales rather than a curb
and gutter system and featured the use of bioretention areas in the roadway loop to treat
stormwater quality. This combination of stormwater practices provided greater pollutant
removal through filtering and infiltration.
One of the most important objectives in the redesign strategy was to improve the
location and performance of the septic systems that dispose of wastewater at the site.
Home sites were oriented to be near soils that were most suitable for septic system
treatment. In addition, six homes shared three common septic fields located within open
space rather than on individual private lots. Last, given the permeability of the soils,
advanced recirculating sand filters were installed to provide better nutrient removal than
could be achieved by conventional septic systems.
Comparative hydrology for Duck Crossing

Given its low impervious cover and permeable soils, the water balance at Duck Crossing was
dominated by infiltration, even after development. The comparative hydrology under the
Chapter I.3:
53
Table I.3.1 Annual Water Budget of Duck Crossing
Runoff (in./year)
Infiltration (in./year)
No practices
Practices
No practices
Practices
Predeveloped
Conventional
Design
Open Space
Design
2.3
18.2
4.8
4.8
15.3
15.3
3.9
3.7
17.0
17.2
Source: Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed
Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.
Developed
Areas
Runoff
15%
Natural
Areas
Runoff
2%
Developed
Areas
Infiltration
9%
Septic
Systems
70%
Natural
Areas
Infiltration
4%
Figure I.3.4 Nitrogen load distribution from the conventional design of Duck Crossing, without
stormwater practices. (From Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000.
With permission.)
five development scenarios is presented in Table I.3.1. As might be expected, the conventional
design yielded the greatest volume of surface runoff and the least amount of infiltration. The
open space design produced about 25% less annual surface runoff and 12% more infiltration
than the conventional design but did not come close to replicating predevelopment conditions. The use of stormwater practices did not materially change the water balance under
either the conventional or open space design at Duck Crossing (see Table I.3.1).
Comparative nutrient output at Duck Crossing

Nutrient export at Duck Crossing was dominated more by subsurface water movement
than by surface runoff. Indeed, stormwater runoff seldom comprised more than 15% of
the annual nitrogen or phosphorus load from this lightly developed site. The SUNOM
model indicated that the major source of nutrients was subsurface discharges from septic
systems, which typically accounted for 60 to 80% of the total load in every development
scenario (see Figure I.3.4).
The open space design sharply reduced nutrient export, primarily because recirculating sand filters were used in the shared septic systems and helped to reduce (but not
eliminate) subsurface nutrient discharge. The other elements of the open space design
(reduced impervious cover, reduced lawn cover, and multiple stormwater practices) also
helped to reduce nutrient export, but by a much smaller amount. The comparative nutrient
export from each Duck Crossing development scenario is detailed in Figure I.3.5.
54
Figure I.3.5 Annual nitrogen and phosphorus loads for each development scenario at Duck Crossing. (From Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed
Comparative cost of development

The cost to build infrastructure for the open space design was estimated to be 25% less
than the conventional design at Duck Crossing, due primarily to the necessity for less
road paving, sidewalks, and curbs and gutters. Even when higher costs were factored in
for the more sophisticated stormwater and on-site wastewater treatment used in the open
space design, the total cost was still 12% lower than the conventional design. In addition,
the open space design had seven fewer acres that needed to be cleared and graded, or
served by erosion and sediment controls, compared with the conventional design (these
costs are not currently evaluated by the SUNOM model). Overall, the SUNOM model
estimated that the conventional design at Duck Crossing had a total infrastructure cost of
$143,600, compared with $126,400 for the open space design.
Summary
The comparative results for the Duck Crossing redesign analysis are summarized in
Figure I.3.6. The open space design increased natural area conservation and reduced
impervious cover, stormwater runoff, nutrient export, and infrastructure costs compared
with the conventional subdivision design.
Stonehill Estates a medium-density residential subdivision

Stonehill Estates, located near Fredericksburg, VA, is situated in the rolling terrain of the
Piedmont. The undeveloped parcel was 45 acres in size, nearly all of which was mature
Chapter I.3:
55
Figure I.3.6 Percentage change in key site conditions from a conventional design to an open space
design, both with stormwater practices. (From Zielinski, J.A., The benefits of better site design in
residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott
City, MD, 2000. With permission.)
Figure I.3.7 Predevelopment conditions at the Stonehill Estates site. (From Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed Protection Techniques, Center for
Watershed Protection, Ellicott City, MD, 2000. With permission.)
hardwood forest (Figure I.3.7). An intermittent stream bisected the site, discharging into
a perennial stream near the southern edge of the parcel. Roughly 3.6 acres of forested
wetlands were found along the stream corridors, and an extensive floodplain was located
56
along the perennial stream. Soils at the site were primarily silt loams and were moderately
permeable (hydrologic soil groups C and D).
The site was highly attractive for development, given the excellent access provided
by two existing roads, both of which had public water and sewer lines that could be easily
tapped to serve the new subdivision. The conventional design was zoned for three dwelling units per acre. After unbuildable lands were excluded, the parcel yielded a total of
108 house lots, each of which was about 9,000 ft2 in size (Figure I.3.8). The subdivision
design typifies medium-density residential subdivisions developed in the last two decades
in the Mid-Atlantic region, where lots sizes were uniform in size and shape and homes
were set back a generous and fixed distance from the street. The design utilized a mix of
wide and moderate street sections (34 ft and 26 ft), and included six large-diameter culde-sacs for turnarounds. Sidewalks were generally installed on both sides of the street.
The stormwater management system for the conventional design represents the typical
pipe and pond approach utilized in many medium-density residential subdivisions.
Street runoff was conveyed by curbs and gutters into a storm drain system that discharged
into the intermittent stream channel and then traveled downstream to a dry extended
detention pond. The pond was primarily designed to control flooding but also provided
some limited removal of stormwater pollutants.
Interestingly, about 25% of the site was reserved as open space in the conventional
design at Stonehill Estates. Nearly all of these lands were unbuildable because of environmental and site constraints (e.g., floodplains, steep slopes, wetlands, and stormwater
facilities), and the resulting open space was highly fragmented. Even so, about a fourth
of the forested wetlands were impacted by two roads crossing over the intermittent stream.
Almost 90% of the original forest cover was cleared as a result of the conventional design
and was replaced by lawns and impervious cover. Overall, about 60% of the site was
converted to lawns, and another 27% was converted to impervious cover.
Open space design for Stonehill Estates

In the redesign analysis, Stonehill Estates was designed to incorporate many of the open
space design techniques advocated by Arendt (1994). The resulting design retained the
same number of lots as the conventional design but had a much different layout
(Figure I.3.9). The average lot size declined from about 9,000 ft2 in the conventional design
to 6,300 ft2 in the open space design. This reduced lot size allowed about 44% of the site
to be protected as open space, most of which was managed as a single unit that included
an extensive natural buffer along the perennial and intermittent stream corridor.
The basic open space layout was augmented by several other better site design practices, including narrower streets, shorter driveways, and fewer sidewalks. Loop roads
were used as an alternative to cul-de-sacs. In some portions of the site, irregularly shaped
lots and shared driveways were used to reduce overall road length. Each individual lot
was located adjacent to open space, so that the more compact open space lots would not
feel as crowded. As a result of these techniques, the open space design for Stonehill Estates
reduced impervious cover from 27 to 20%. In addition, lawn cover declined from 60 to
30% of the total site area.
The innovative stormwater collection system utilized dry swales instead of storm
drains in gently sloping portions of the site. The dry swales and several bioretention areas
located in loop turnarounds were used to initially treat stormwater quality. Each of these
practices then discharged to a small micropool detention pond, with an embankment
created by the single road crossing over the intermittent stream.
Chapter I.3:
Figure I.3.8 The conventional subdivision design that was built at Stonehill Estates (108 lots). (From Zielinski, J.A., The benefits of better site design in
residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.)

57
58
Figure 1.3.9 The open space subdivision that could have been built at Stonehill Estates (108 lots). (From Zielinski, J.A., The benefits of better site design
in residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.)
Chapter I.3:
59
Table I.3.2 Comparative Hydrology of Stonehill Estates
Runoff (in./year)
Infiltration (in./year)
No practices
Practices
No practices
Practices
Predeveloped
Conventional
Design
Open Space
Design
2.1
4.9
10.6
10.6
3.1
3.1
8.8
8.0
4.0
4.8
Source: Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed
Protection Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.
Comparative hydrology
Prior to its development, the highly wooded site produced very little surface runoff but,
because of relatively tight soils, generated only a modest amount of infiltration. After the
site was converted into the conventional subdivision, however, surface runoff increased
by a factor of 5, and infiltration was reduced by about 40% (Table I.3.2). In contrast, the
open space design worked to reduce stormwater runoff and increase stormwater infiltration compared to the conventional design, although it did not come close to replicating
the original hydrology of the forested site (Table I.3.2).
Comparative nutrient output

As might be expected, the conversion of the forest into a conventional subdivision greatly
increased nutrient export from the site; the model indicated that annual phosphorus and
nitrogen export would increase by a factor of 7 and 9, respectively, after development (see
Figure I.3.10). Unlike Duck Crossing, nutrient export at Stonehill Estates was dominated
by stormwater runoff after development. The SUNOM model indicated that stormwater
runoff contributed about 94% of the annual nutrient export from the site, with subsurface
water movement adding only 6% to the total export. Nutrient loads were not greatly
reduced by the dry extended detention pond installed at the conventional subdivision;
the model indicated that nutrient export from the conventional design would still be six
to seven times greater than the predevelopment condition even with this stormwater
treatment practice.
In contrast, the open space design resulted in greater nutrient reduction (Figure I.3.10).
For example, the open space design scenario without stormwater practices produced a
lower nutrient load than the conventional design scenario with stormwater practices. This
was primarily due to lower impervious cover associated with the open space design.
When the open space design was combined with more sophisticated stormwater practices
(i.e., bioretention, dry swales, and wet ponds), nutrient export was half that of the conventional design. It is interesting to note, however, that even when the most innovative
site design and stormwater techniques were applied to the site, nutrient export was still
three to four times greater than that produced by the forest prior to development.
Infrastructure costs
The total cost to build infrastructure at Stonehill Estates was about 20% less for the open
space design than for the conventional design. Considerable savings were realized in the
form of less road paving and shorter lengths of sidewalks, water and sewer lines, and
curbs and gutters. The cost difference between the open space and conventional designs
would have been greater if not for the fact that higher costs were incurred for the more
sophisticated stormwater practices used in the open space design. It was estimated that
60
Figure I.3.10 Annual nitrogen and phosphorus loads for each Stonehill Estates development scenario. (From Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed
the infrastructure cost for the conventional design was $1.54 million, compared with $1.24
million for the open space design.
Summary
The comparative results for the Stonehill Estates redesign analysis are summarized in
Figure I.3.11. The open space design reduced impervious cover, natural area conversion,
stormwater runoff, nutrient export, and development costs compared with the conventional subdivision design.
Other redesign research

Several other researchers have employed redesign comparisons to demonstrate the benefits of open space subdivisions, over a wide range of base lot sizes. The results are shown
in Table I.3.3. It should be recognized that each study used slightly different models and
assumptions, and as such, strict comparisons should be avoided. The redesign comparisons clearly show that open space designs can sharply reduce impervious cover and
stormwater runoff while accommodating the same number of dwelling units, at least to
base lot sizes of an eighth of an acre. The reductions in impervious cover and runoff range
from 7 to 65%. The ability of open space design to reduce impervious cover starts to
diminish for residential zones that exceed densities of four dwelling units per acre.
Chapter I.3:
61
Figure I.3.11 Percentage change in key site conditions from a conventional design to an open space
design, both with stormwater practices. (From Zielinski, J.A., The benefits of better site design in
residential subdivisions, in Watershed Protection Techniques, Center for Watershed Protection, Ellicott
City, MD, 2000. With permission.)
Table I.3.3 Redesign Analyses Comparing Impervious Cover and Stormwater Runoff
from Conventional and Open Space Subdivisions
Residential
Subdivision
Remlik Hall1
Tharpe Knoll2
Chapel Run2
Pleasant Hill2
Prairie
Crossing3
Buckingham
Greene2
Belle-Hall4
Original Zoning
for Subdivision
Impervious Cover at the Site

Conventional Open Space
Net
Reduction in
Design
Design
Change Stormwater Runoff
(%)
(%)
(%)
(%)
5-acre lots
1-acre lots
-acre lots
-acre lots
- to -acre lots
5.4
13
29
26
20
3.7
7
17
11
18
31
46
41
58
20
20
44
31
54
66
-acre lots
23
21
High density
35
20
43
31
Maurer, 1996; 2DE DNREC, 1997; 3Dreher, 1994; 4SCCCL, 1995

Source: Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed Protection
Techniques, Center for Watershed Protection, Ellicott City, MD, 2000. With permission.
These studies reinforce the conclusion that open space designs are usually less expensive to build than conventional subdivisions. The projected construction cost savings
associated with open space designs ranged from 40 to 66% (Table I.3.4). Most of the cost
savings were due to the reduced need for road building and stormwater conveyance. In
another study, Liptan and Brown (1996) report that open space design produced infrastructure construction costs savings of $800 per home in a California subdivision.
Numerous economic studies have shown that well-designed and marketed open space
designs are very desirable to homebuyers and very profitable for developers. Strong
evidence indicates that open space subdivisions sell faster, produce better cash flow, yield
a higher return on investment, and appreciate faster than their traditional counterparts
62
Table I.3.4 Projected Construction Cost Savings for Open Space Design from Redesign Analyses
Residential
Subdivision
1
Construction
Savings (%)
Remlik Hall
52
Tharpe Knoll2
Chapel Run2
Pleasant Hill2
Buckingham Greene2
56
64
43
63
Notes
Includes costs for engineering, road construction, and
obtaining water and sewer permits
Includes roads and stormwater management
Includes roads, stormwater management, and reforestation
Includes roads, stormwater management, and reforestation
Includes roads and stormwater management
Maurer, 1996; 2DE DNREC, 1997

(Arendt, 1994; Ewing, 1996; NAHB, 1997; ULI, 1988; CWP, 1998a; and Porter et al., 1988).
Although open space designs are often perceived as applying only to upscale and affluent
consumers, several successful open space subdivisions have been built for moderate- to
lower-income buyers. Both ULI (1988) and Ewing (1996) report that open space designs
can be an effective tool to promote affordable housing within local communities.
The relatively high demand for open space designs reflects two important economic
trends. The first trend is that the tastes and preferences of many new homebuyers are
gradually changing. Recent market surveys indicate that homebuyers increasingly desire
natural areas, smaller lawns, better pedestrian access, wildlife habitat, and open space in
the communities in which they choose to live. The second trend is that open space
developments that can provide these amenities seldom comprise more than 5% of the new
housing offered in most communities. Consequently, there appears to be a large and
relatively untapped potential demand for more open space developments. Other compelling benefits of open space design are detailed in CWP (1998a) and Schueler (1995).
Evaluating the quality of individual open space developments

In the real world, site designers must satisfy a wide range of economic objectives, and
water quality or resource protection is usually not on the top of the list. It is certainly
possible to design a lousy open space design, and communities should expect a wide
range in the quality of open space designs they review. How can a community objectively
evaluate the quality of individual open space design proposals, and differentiate poor or
mediocre projects from the good and outstanding ones?
Nerenberg and Freil (1999) have recently developed a simple rating system to evaluate
the quality of individual open space design proposals. The rating system, known as the
Conservation Development Evaluation System (CeDES), was developed in consultation
with a host of planning agencies and organizations. The CeDES employs 10 core criteria
to test how well a proposed open space design reduces impervious cover, minimizes
grading, prevents soil loss, reduces and treats stormwater, manages open space, protects
sensitive areas, and conserves trees or native vegetation. Each of the 10 core criteria has
a quantitative benchmark for comparison. An example of one benchmark that rates the
quantity and quality of open space is provided in Table I.3.5. A full description of the
CeDES rating can be found in Conservation Fund (1999).
Based on the total score achieved under the 10 core criteria, an open space design
project can earn anywhere from zero oak leaves up to four oak leaves. The more oak
leaves earned, the better the quality of the proposed project. Based on initial testing, the
CeDES seems to do a good job of sorting the poor projects from the outstanding ones.
Chapter I.3:
63
Table I.3.5 Sample Evaluation Criteria for the Quantity and Quality of Open Space Development
(Conservation Fund, 1999)
Points
Achieved
by the
Development
Percent of Open Space Achieved for Different Residential Zones

Less Than
From
From
From
More Than
0.5 Unit
0.5 to 1 Unit
1 to 2 Units
2 to 4 Units
4 Units
per Acre
per Acre
per Acre
per Acre
per Acre
(%)
(%)
(%)
(%)
(%)
2
1
0
+1
+2
0 to 9
10 to 14
15 to 24
25 to 30
More than 30
Less than 15
15 to 24
25 to 34
35 to 40
More than 40
15 to 24
25 to 34
35 to 49
50 to 55
More than 55
25 to 34
35 to 49
50 to 59
60 to 70
More than 70
Less than 40
Less than 50
Less than 60
Less than 70
More than 80
Note: The total open space achieved by the site is computed using the following formula:
[A(0.2) + B(0.2) + C(0.5) + D] * 100/E
A = open space acres in managed landscape
B = open space acres in annual crops
C = open space acres in perennial crops
D = open space acres in native vegetation
E = total undeveloped acres in open space
The CeDES is intended for use as a tool for local development review, but it can also be used
as a marketing tool to let homebuyers know how green their new subdivision actually is.
Implications for the watershed manager

The redesign comparisons have several implications for the watershed manager. First,
they offer compelling quantitative evidence that open space design can sharply reduce
stormwater and nutrient export from new development and, as such, can serve as an
effective tool for watershed protection. It is interesting to note that open space design
alone produced nutrient reductions roughly equivalent to those achieved by structural
stormwater practices. In other words, nutrient export from open space designs without
stormwater treatment was comparable to the conventional designs with stormwater treatment. When open space designs were combined with effective stormwater treatment,
nutrient loads were sharply reduced but were still greater than predevelopment conditions.
A second, more troubling implication is that it may well be impossible to achieve a
strict goal of no increase in nutrient load for new development, even when the best site
design and most sophisticated stormwater practices are applied. A handful of communities
have adopted stormwater criteria that mandate that no net increase in phosphorus load
occur as a result of development, but as the redesign comparisons in this article show,
such criteria are not likely to be actually achieved. Thus, if nutrient loads are capped in
a watershed, managers may need to remove pollutants at existing developments with
stormwater retrofits in order to offset increases in nutrient loads produced by new development.
The redesign research also has some implications for watershed-based zoning. Quite
simply, a shift from conventional to open space design can reduce the impervious cover
of many residential zoning categories by as much as 30 to 40%. In some watersheds, an
aggressive shift to open space design in new residential zones is an essential strategy to
meet an impervious cover cap for protecting sensitive or impacted streams.
64
Another notable finding is that large-lot subdivisions have the potential to generate
the same unit area nutrient export as higher-density subdivisions. The high nutrient
loading from large-lot developments in unsewered areas is attributed to subsurface discharges from septic systems. From a nutrient management standpoint, it may be more
cost-effective to regulate septic system performance than stormwater performance in very
low-density residential subdivisions located on permeable soils.
Last, watershed managers have only a few tools at their disposal that offer developers
a real chance to save money. The economic evidence clearly suggests that open space
design is such a tool and has the potential to either reduce the cost of development or at
least offset the cost of other watershed protection measures. Despite its economic and
environmental benefits, however, open space design is neither a development option in
many communities nor widely used by most developers even when available. Many
communities will need to fundamentally change their local development rules in order
to make open space design an attractive development option.
Site planning roundtables that involve the local players that shape new residential
development, described later in this issue, are an effective way to bring this change about.
The ultimate goal is to make open space design a by-right form of development, so that
its design, review, and approval are just as easy and certain as a conventional subdivision.
The day may come when a special exception or permit is needed to build a conventional
subdivision.
Literature cited
Arendt, Randall, Designing Open Space Subdivisions: A Practical Step-by-Step Approach, Natural Lands
Trust, Media, PA, 1994.
Center for Watershed Protection (CWP), Better Site Design: A Handbook for Changing Development
Rules in Your Community, Ellicott City, MD, 1998a.
CWP, Nutrient Loading from Conventional and Innovative Site Development, prepared for the Chesapeake
Research Consortium, Ellicott City, MD, 1998b.
Conservation Fund, Pilot Conservation Development Evaluation System, Great Lakes Office, available at www.conservationfund.org/conservation/sustain/gloindex.html, 1999.
Delaware Dept. of Natural Resources and Environmental Conservation (DE DNREC), Conservation
Design for Stormwater Management, Dover, DE, 1997.
Dreher, D. and Price, T., Reducing the Impact of Urban Runoff: The Advantages of Alternative Site Design
Approaches, Northeastern Illinois Planning Commission, Chicago, IL, 1994.
Ewing, R., Best Development Practices: Doing the Right Thing and Making Money at the Same Time,
American Planning Association, Chicago, IL, 1996.
Liptan, T. and Brown, C., A Cost Comparison of Conventional and Water Quality-based Stormwater
Designs, City of Portland, Portland, OR, 1996.
Maurer, G., A Better Way to Grow: For More Livable Communities and a Healthier Chesapeake Bay,
Chesapeake Bay Foundation. Annapolis, MD, 1996.
National Association of Homebuilders, Cost Effective Site Planning, Washington, DC, 1986.
Nerenberg, S. and Freil, K., The conservation development evaluation system (CeDES): Evaluating
environmentally friendly developments, Land Development, Fall 22, 1999.
Porter, D., Phillips, P., and Lassar, T., Flexible Zoning: How It Works, Urban Land Institute, Washington,
DC, 1988.
Schueler, T., Site Planning for Stream Protection, Center for Watershed Protection. Ellicott City, MD,
1995.
Urban Land Institute (ULI). Density by Design, J. Wetling and L. Bookout, Eds., Urban Land Institute,
Washington, DC, 1988.
Zielinski, J.A., The benefits of better site design in residential subdivisions, in Watershed Protection
Techniques, Center for Watershed Protection, Ellicott City, MD, 2000.
Chapter I.3:
65
Response
Using computer scenarios to improve site design
Studies have shown that water can be impaired at levels as low as 5 to 15% subwatershed
impervious cover, amounts most suburban development patterns exceed. As a result, we
need to explore better site designs to minimize pavement from new development, while
at the same time conserving more open space and enabling more effective stormwater
management. One way to further this exploration of viable development alternatives is
to undertake iterative analyses of conceptual redesigns through computer simulations or
models.
Jennifer Zielinskis method is to examine a series of model development principles
developed by a consortium of planning, transportation, financial, engineering, real estate,
wastewater, and public safety experts and to see what the effect of varying these principles
might have on changes in runoff rates and consequent nutrient export. These redevelopment principles are similar to those discussed elsewhere in this book, as for example, lowimpact development as in Chapter I.5 by Larry Coffman and various open space
approaches as in Chapter II.3 by Charles Flink.
The lessons presented in this chapter for water sensitive designers are of major importance. We need to shift our focus from reacting to damages already produced, to instead
proactively planning to circumvent those damages from occurring in the first place. The
use of conceptual redesigns is an optimal way to generate such predictive understanding.
In other words, as Zielinski states, improvements should begin right at the initial planning
stage for new site developments.
The key to all this is a move toward embracing the principles of cluster development,
what Zielinski refers to as open space designs. By concentrating or clustering development to one portion of the site through relaxing lot sizes, frontages, road sections, and
other subdivision geometry, open spaces are left with the positive benefits of reduced
overall impervious surfaces at the same time as increased areas for recreation activities or
wildlife habitat with a corresponding improvement in the quality of life for the residents.
With the additions of other low-impact development techniques such as narrow setbacks,
riparian buffers, and removal of curbs and gutters, an added benefit is a reduction in
development costs by as much as one-half.
Zielinski reports that the benefits ensuing from undertaking such open space planning
in terms of stormwater runoff reductions can also be substantial. Indeed, her modeling
exercises raise the very interesting possibility that, through the inclusion of open spaces
and clustered development alone, runoff can be reduced by 30 to 40%; in other words,
roughly equivalent to that occurring through adoption of structural stormwater BMPs
(best management practices). Thus, a very compelling case can be made that the earlier
water sensitive design can be applied to newly planned developments, the more effective
(and less expensive) can be our solutions for protecting water bodies from the consequences of those developments.
chapter I.4
Post-industrial watersheds:
Retrofits and restorative redevelopment
(Pittsburgh, Pennsylvania)
Richard D. Pinkham and Timothy Collins
Abstract
Although concepts and techniques to manage urban runoff in ways that maintain ecosystem health are being effectively demonstrated in many developing watersheds, efforts to
change stormwater management and restore ecosystem function in older, highly developed watersheds are still in their infancy. This chapter highlights the restorative redevelopment concept first proposed in the Pittsburgh region. When development plans for
a brownfield site in Pittsburgh called for filling in and culverting a local stream, an
interdisciplinary, artist-led team at Carnegie Mellon Universitys STUDIO for Creative
Inquiry developed and implemented a program of public education and dialogue, stream
corridor evaluation, and planning for alternatives. A greenway concept for Nine Mile Run
was developed with the support and participation of multiple agencies and local stakeholders. Greenway proponents also addressed watershed-scale problems. Nine Mile Run
suffers from combined and sanitary sewer overflows, erosive pulses of stormwater, contributions of pollutants carried in urban runoff, and reduced base flow problems traced
to high levels of impervious surfaces in its watershed and crumbling sewer and stormwater infrastructure. The STUDIO and Rocky Mountain Institute hosted an interdisciplinary charrette to select and illustrate measures for retrofitting streetscapes, parks, and
properties in the upper watershed to infiltrate, detain, and treat runoff at the site scale.
The charrette also developed policy action plans to encourage the widespread micromeasure implementation necessary to produce cumulative benefit for Nine Mile Run. A key
theme of the restorative redevelopment approach envisioned for Nine Mile Run is to
integrate infrastructure, ecosystem, and community issues. The charrette results exemplify
seven principles of restorative redevelopment.
Introduction: The opportunity of older urban watersheds

The watersheds of Americas older urban areas suffer a multitude of problems: their high
proportions of impervious surfaces produce polluted runoff, erosive storm flows, and
reduced stream base flows. Frequently, crumbling or inadequate infrastructure results in
sewer overflows that compromise ecosystem and human health. Financial resources are
1-56670-562-2/02/$0.00+$1.50
67
68
typically very limited, and the social fabric of these places is often frayed and faded.
Traditional solutions such as stormwater detention basins, increased conveyance capacity,
and wastewater plant expansions require space that is often not available, impose high
monetary costs, and treat only the hydrologic results of urbanization rather than the
diverse causes of the problems.
On the other hand, urban watersheds present remarkable opportunities if we look
carefully and creatively for new solutions. This requires consideration of both the green
infrastructure the open spaces, vegetation, and soils remaining and recoverable even
in highly urban areas and the conventional pipes and plants of the gray infrastructure
as an integrated system. Although we cannot expect to restore urban watersheds to
predevelopment conditions, the rehabilitation of ecosystem services can assist in the
resolution of wet weather problems and expose millions to the values and benefits of
functional hydrological and biologic systems.
Many techniques have emerged in the new development context to prevent degradation of streams, lakes, and other water bodies. There are scores of ways to reduce impervious surfaces and incorporate measures that infiltrate, retain, and filter runoff. And in
recent years the stormwater field has moved from isolated application of best management
practices (BMPs) to comprehensive, integrative schemes for protecting watersheds as they
develop. These include model development principles (see Chapter I.3 by Zielinski) and
low-impact development (see Chapter I.5 by Coffman). These approaches focus on
preservation of natural function, which requires going beyond the conventional stormwater detention emphasis on maintaining predevelopment peak flows, to include maintenance of groundwater recharge and the use of soil and vegetation to neutralize pollutants
(see Chapter I.1 by Ferguson; also Chapter I.7 by France and Craul). The BMPs emphasized
by these approaches are typically micromeasures permeable pavements, dry wells,
rain barrels, vegetated roof covers, subsurface recharge beds, bioretention cells, infiltration
trenches, vegetated swales, and other techniques that are installed near buildings and
pavements, on individual lots, in public rights of way, and otherwise distributed throughout the built environment. They are implemented as close to where the rainfalls as is
feasible, rather than some downstream point below conveyance systems that have concentrated runoff.
Can these concepts and techniques be applied to problems in already-built urban
landscapes? Scattered applications are occurring. Toronto, Ontario, has encouraged disconnection of residential roof downspouts from combined sewers by doing providing free
rain barrels or free labor to divert downspouts to vegetated areas (Recycle Your Rain,
1998). Portland, OR, has trialed a variety of retrofits, including vegetated roof covers,
vegetated swales, and water gardens (see Chapter I.6 by Liptan and Murase). The Low
Impact Development Center is developing retrofits for the U.S. Navy Yard in Washington,
D.C. and is preparing a U.S. EPA-funded guidance document on the use of low-impact
development techniques to control wet weather flows and combined sewer overflows in
urban areas (Low Impact Development Center homepage, 2001). In southern California,
TreePeople developed a model retrofit program (Condon and Moriarty, 1999), has implemented a comprehensive residential retrofit as well as infiltration systems and cisterns at
two schools, and is working with regional agencies on impervious surface reductions and
an ambitious retrofit project for a 2,700-acre watershed as an alternative to a $42 million
storm drain (T.R.E.E.S. Project Overview, 2001; Lipkis, 2001).
Further study will be necessary to demonstrate the effectiveness of microscale retrofit
and redevelopment techniques in restoring more natural function to urban watersheds;
but already, many cities and developers are motivated to try these measures by the
staggering costs of conventional management and infrastructure investments. The U.S.
Environmental Protection Agency (1997) estimates national combined sewer overflow
Chapter I.4:
69
remediation costs at $45 billion and sanitary sewer overflow remediation costs at up to
$87 billion (Parsons Engineering Science et al., 2000). The agency estimates the costs for
municipalities to come into compliance with the Phase I stormwater regulations (excluding,
importantly, many O&M costs and all costs borne by private parties) to be $7.4 billion
nationwide over 20 years (U.S. EPA, 1997). The EPA (1999) also estimates public and private
costs for implementing Phase II stormwater regulations at $848 to $981 million annually.
What is also clear is that urban watersheds face a host of other problems, and their
residents desire and deserve social, economic, and environmental progress. Yet most cities
view wet weather management as a technical matter, best managed by specialists in
technical disciplines concerned with conveyance infrastructure. The question we want to
ask is: Cannot investments in sewer and stormwater infrastructure be made in ways that
address additional urban goals? The opportunity in older urban watersheds is to expand
the sewer overflow and stormwater management agendas to redefine the wet weather
problem and make infrastructure improvements that will provide multiple benefits and
energize multiple constituencies. To do so, we must link infrastructure rehabilitation with
the restoration of ecosystems and a broader cultural restoration. In turn, the resources for,
and likely implementation rates of, sewer and stormwater measures will be increased.
Exploring this uncharted territory is the story of this chapter.
We know that wet weather management takes place in the context of continual redevelopment and retrofitting of urban properties. Pavements are relaid; buildings are renovated and reconstructed; transportation systems are reorganized; and utilities are maintained and replaced. These changes are opportunities for reevaluating, correcting, healing,
and educating. We can find ways to make redevelopment serve ecosystem restoration. We
can also find ways to make infrastructure and ecosystem rehabilitation projects serve
broader economic and social purposes. This give-and-take, this expanding of agendas, is
the heart of what we call restorative redevelopment, which we define as redevelopment
and retrofit projects that improve the value and livability of the city while effectively
restoring natural processes and functions. The concept emerged from a watershed project
in metropolitan Pittsburgh, Pennsylvania (Ferguson, Pinkham, and Collins, 1999).
Envisioning restorative redevelopment: Pittsburghs Nine Mile Run

The watershed: Industrial legacies and a redevelopment opportunity
Nine Mile Run is a watershed and stream that joins the Monongahela River nine miles
from the Point, the spit by downtown Pittsburgh where the Monongahela and Allegheny
rivers join to form the Ohio River (Figure I.4.1). The run (a regional term for creek or
stream) drains five municipalities, flowing through wooded Frick Park and then into an
urban brownfield that dominates the bottom of the watershed.
The brownfield is a mountain of slag the detritus from steel making as much as
20 stories high surrounding and overshadowing the creek. It covers Nine Mile Runs
former flood plain, identified by Frederick Law Olmsted, Jr., for a new city park in 1910,
and instead purchased by the steel industry. Roughly 240 acres were covered by slag
dumped over a 50-year period. In 1993, Pittsburgh City Planning developed a conceptual
plan for houses and open space on this mound of industrial by-product. An economic
development team working with the city decided that the water in Nine Mile Run was
dirty enough and the benefits to development attractive enough to propose that the stream
be buried once and for all. Culverting would result in a 20% increase in housing on the
site. Thus, the initial proposal called for a flat development site that would bury what
was left of the stream under 150 ft of slag a deed even the steel industry could not
accomplish.
70
Figure I.4.1 The Nine Mile Run watershed in the Pittsburgh region. (From Ferguson, B.K., Pinkham,
R., and Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment,
Rocky Mountain Institute, Snowmass, CO, 1999. With permission.)
In response to this unique brownfield site linking the Frick Park and Monongahela
River ecosystems, and the proposal to develop it, a group of artists in the STUDIO for
Creative Inquiry at Carnegie Mellon University began a community-based research program about post-industrial public space. The thesis was simple: instead of destroying the
stream, why not let it define the development? The research agenda would focus on
developing knowledge to inform public values, targeting the lack of care that had defined
and denigrated this stream for over 100 years. Talking points included the potential for
these lands to instigate a renewed civic dialogue, the role of artists and academia in this
program, and the application of restoration ecology in the redevelopment of post-industrial public space. The contention from the beginning was that urban brownfields, the
artificial great estates handed down from the industrial era, are an important public space
opportunity, particularly for rust belt cities like Pittsburgh. The Nine Mile Run Greenway
Project team, directed by three artists, included a diverse group of professionals from
academia and industry. The project had a partnership with local government through
Pittsburgh City Planning but retained its autonomy through external funding from the
Heinz Endowments and others.
Arguably, the issues that drove the private development were different than the issues
that might drive a public investment in a post-industrial open space. The first question
to address was: Was the dump really a dump? Did the stream have any residual or
emergent values? Was the riparian link between the Monongahela and Frick Park of any
import? The project team developed a program to understand community interest, then
followed that up with an ecological study to confirm the nature of the opportunities and
constraints. The effort focused on:
History and public policy. Purchased a year before the first zoning laws would have
protected it, the Nine Mile Run dump site was a key component of the steel
industry. For each ton of steel, two to three tons of slag were produced. Its transport
Chapter I.4:
71
and disposal cut into profits, making disposal at nearby Nine Mile Run essential.
A complicit city government looked away as the original grandfathered parcel was
filled and slag began to creep toward communities and parklands.
The stream. Erosive urban runoff has downcut the stream channel and revealed
adjacent sewer systems. The aging, cracked, and disjointed sewers leak sewage
into the stream in wet weather. In dry weather, the hydraulics reverse and much
of the stream disappears into the sewers. The results are alarmingly high pathogen
counts (in a city park) and a diminished aquatic ecology. Nonetheless, fish are
found in a number of pools and beavers are regularly seen each spring.
The slag slopes. The development interests wanted to bulldoze the slopes; the project
team thought that they could be successfully vegetated in place. Porosity, alkalinity,
fines on the surface, dark color, and south-facing slopes makes for a challenging
environment for plants and seedlings. Yet the north-facing slopes and the oldest
south-facing slopes are populated with Box Elder, Sycamore, Bigtooth Aspen, Black
Locust and Slippery Elm, Pin Oak, Red Maple, and a curious find the Hop Tree,
a Pennsylvania endangered species not usually found in this area.
Sustainable open space. Pittsburgh is only now planning to resuscitate its four largest
city parks. The meaning, form, and function of post-industrial parklands were not on
the primary agenda of the brownfield site developer. The challenge in Nine Mile Run
was to see the post-industrial landscape as a contemporary mirror to Frick Park and
as a demonstration site for ecological restoration. The goal was to seek a public private
model of development that would result in a new era of sustainable open space.
With an understanding of the existing development program, the team knew it had
to develop specific programmatic goals to engage the public in an alternative development
dialogue focusing on the potential form of this post-industrial public space and its function
within a social and ecological context. The first year of the project, dubbed Ample
Opportunity: The Community Dialogue, was designed to create opportunities for on-site
experience and community dialogue; expand the understanding of the issues and the
discourse on public space; and increase understanding of the environmental and public space
issues of a range of local constituencies (Ample Opportunity: A Community Dialogue, 1998).
The second year, Ample Opportunity: The Ecology of a Brownfield, built on the work of the
first year and followed through with expert studies to define the range of opportunities. The
team developed an exhaustive biological and landscape ecology study and a set of development alternatives (Nine Mile Run Rivers Conservation Plan, 1998). This study provided a
baseline upon which further work on the ecosystems and the infrastructure that affects them
could be judged, and it set out the content for design alternatives in year three. In addition
to outlining alternatives, the goals of the second year included grounding the professional
effort in a rigorous public discourse, addressing the watershed-wide issues impacting Nine
Mile Run, and constraining the site development and greenway opportunity.
In the final year, in Ample Opportunity: The Brownfield Transformation, the experts
recommended alternatives were developed into a set of effective communication tools to
inform community decision making. The team produced a series of images, texts, and
experiences directed toward the goal of achieving a community consensus conceptual
design for the site, the greenway, and Frick Park. Concurrently, the STUDIO and its
advisory board developed an institutional and economic plan to place the alternatives
and the final community consensus process in a realistic light, and provide the program
with the inertia and support to move forward independently of the STUDIOs involvement. (Comprehensive materials for the Nine Mile Run Greenway Project and related,
evolving activities are available at http://slaggarden.cfa.cmu.edu.)
72
The overarching goal of the Nine Mile Run Greenway Project was to enable an
equitable public dialogue about brownfields, nature, and public space. The team saw its
work in terms of a community consensus process and a public policy discussion about
the form and function of post-industrial public space, a discourse that was missing and
continues to be an anomaly in the current program of local brownfield development. To
do this, the project leaders sought support from the specific municipal agencies managing
the current development program. At the same time, to retain its objectivity as a separate
autonomous entity, the STUDIO obtained independent funding from the Heinz Endowments. As the team began to clarify its role, members realized they were revealing a
complex aesthetic based on discourse, restoration ecology, and sustainable landscape
systems. It became clear that Nine Mile Run was a better site for experimentation and
modeling new approaches to discourse and dialogue than originally expected. Nine Mile
Run provided both the context and the subject for experimentation. Issues that arose from
efforts to reclaim, restore, or heal the site enveloped diverse disciplines and areas of
knowledge engineering, ecology, social and political democracy, and the arts. The
project showed that examining sites like this in the context of their watersheds and
communities has a curious potential to transform culture in unexpected ways.
The ecosystems and infrastructure charrette

In the second year of the Nine Mile Run Greenway Project, it became clear to its organizers
that long-term restoration of the aquatic and riparian ecosystem in the lower watershed
(the brownfield site, the potential greenway, and Frick Park) would require actions to
address the source of the excess stormwater flows: the highly built-up upper watershed
(Figure I.4.2). Also at this time, regulators heightened concerns over sewer overflows in
the Pittsburgh region by threatening fines and other legal actions if local communities did
not move quickly toward compliance with relevant laws. In 1998, the STUDIO and Rocky
Mountain Institute convened a three-day stormwater management charrette with 60 local
and national landscape architects, engineers, architects, artists, planners, policy analysts,
and local citizens.
The interdisciplinary teams at the charrette developed conceptual designs for sample
sites to concretely illustrate the restorative development concept. They also produced
recommendations for institutional arrangements and policy initiatives that could drive
adoption of restorative redevelopment measures in the watershed. The charrettes results
show how implementation of progressive stormwater management measures at the scale
of individual properties and neighborhoods could benefit both the upper and lower
watershed by linking infrastructure rehabilitation, ecosystem restoration, and community
development.
Four sample designs

The sample designs for four sites in the upper portions of Nine Mile Run reuse, restore,
and revitalize their sites by resolving existing site-specific issues, adapting techniques of
construction and stormwater management to Pittsburghs fine-textured soil, frequent
frosts, steep hillsides, and unstable geology. They infiltrate or detain the runoff from a
2-year, 24-hour storm on-site, within a construction budget of $2/gallon of hydraulic
capacity. These parameters are consistent with standards and conventional project costs
established in Allegheny County in recent years.
A variety of on-site measures is available for removing stormwater from sewers and
restoring beneficial natural processes. The following general approaches were employed
in many of the sample designs:
Chapter I.4:
73
Figure I.4.2 The Nine Mile Run watershed, showing impervious streets and buildings, topography,
and the stormwater charrette sites. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating
Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version available at www.gsd.harvard.edu/watercolors.
Password: lentic-lotic.)
Capturing roof runoff in tanks or cisterns for irrigation or indoor graywater use
Disconnecting pavement and roof drainage from sewer lines and directing it to adjacent
vegetated soil or to infiltration basins
Engineering infiltration basins water gardens, dry wells, and subsurface recharge beds to collect runoff and percolate it into the soil
Planting trees to intercept a portion of rainwater
Rehabilitating soils to increase infiltration rates and pollutant-neutralizing microbial
activity
Reconfiguring driveways, parking lots, and streets to turn more of a site over to
pervious, vegetated soil
Using porous pavements special varieties of asphalt, concrete, masonry, and other
materials with open pores that allow water to pass through
Routing runoff through vegetated surface channels swales to slow its velocity,
remove pollutants, and infiltrate it into the soil
Restoring (daylighting) historic streams by excavating culverts and creating naturalized open channels (Pinkham, 2000)
Each design integrates several stormwater management strategies into the built environment of its site. Additionally, each exemplifies restorative redevelopment by integrating
the physical strategies into the social and economic life of the site and its neighborhood.
74
Hunter Park
Hunter Park is located near the headwaters of the Nine Mile Run watershed. It is a
neighborhood park in the community of Wilkinsburg. This is a low-income area; neighborhood streets and sidewalks are in poor condition. The parks ball field, wading pool,
basketball courts, and small playground are in disrepair, although all are heavily used in
season (photographs of original conditions at the four sites are available in Ferguson,
Pinkham, and Collins, 1999).
The bulk of the park is in pervious turf, although it hides the culverted remains of a
natural stream. In contrast, the surrounding residential blocks are mostly impervious with
densely built houses, streets, and sidewalks. Most of the runoff from the impervious
surfaces drains into sewers, contributing to downstream flood pulses, sewer overflows,
water pollution, and reduced base flow.
Despite the currently neglected condition of the Hunter Park area, it has a vigorous
past that symbolizes the industrial development of the region and the character of its
people. Before settlement of the area, the site was a V-shaped headwater stream valley. In
the 19th century a coal mine filled and flattened the site with yards and spoil piles; the
stream was diverted around the periphery. The mining industry brought in a working
population and built company housing nearby. In the early part of the 20th century, the
industrially created land forms served as a baseball field for the Negro League. Some of
the best baseball players in the country played as semipros at Hunter Field. Beginning in
the 1950s, a series of developments gradually transformed the site into a general-purpose
recreational park.
The site is in a valley with a drainage area of 59 acres, of which impervious roads and
rooftops cover approximately 9 acres, or 16% (Figure I.4.3). (The percent impervious area
figure for this and other catchments in this article does not include driveways, sidewalks,
and most parking lots.) Most drainage inlets are clogged with sediment; some drainage
pipes are broken. Some grass swales in the park improve water quality to a degree but are
undersized even for the small amount of water they carry. Concentrated runoff from nearby
impervious surfaces has eroded some of the parks drainage swales and steep side slopes.
The proposed design. The proposed design is a convergence of history, hydrology,
recreation, and neighborhood revitalization, wedding the sites social history to its hydrologic future. Water is brought through a sequence of historicalrecreational spaces and
celebrated at the end. The hydrologic strategies are given form by the parks natural and
cultural history; in turn, the forms illuminate the parks environmental and historic features.
The design uses complementary strategies for various portions of the catchment
(Figure I.4.4). At the upper end of the park, a woodland bioretention area consists of
sand and soil mixtures planted with native plants. It includes a pretreatment area to
dissipate the energy of inflowing runoff and to collect coarse sediment. Then a constructed
wetland treats water at the upstream end of the ball field. It is planted with emergent and
scrub-shrub plants in a complex microtopography. It filters pollutants, reduces peak flow
rates, and stabilizes the flow of water into the grass swales below.
Swales take overflow drainage from the wetlands, and runoff from the fields and
surrounding slopes, around the ball field and through the lower part of the park. The
swales have grass and other vegetation, which help remove pollutants from runoff. For
further infiltration and filtering, they are enhanced with beds made of sand and topsoil
1 to 2 ft deep and 10 to 15 ft wide (Figure I.4.5).
At the bottom of the park, the area where coal mine shanty houses once stood is made
into a public square for the neighborhood. The once-culverted stream is reopened (daylighted) through the square to convey stormwater in restored stream habitat as an amenity and focal point for the park. The square includes a stage for public plays and festivals,
Chapter I.4:
75
Figure I.4.3 The Hunter Park site, catchment, and surrounding area. (From Ferguson, B.K., Pinkham,
Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version available at
76
Figure I.4.4 The proposed plan for rehabilitation of Hunter Park. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The Nine
Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Chapter I.4:
77
Figure I.4.5 Construction of Hunter Parks swales for stormwater infiltration, storage, and water
quality improvement. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater:
The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO,
1999. With permission.)
adding a cultural role to the recreational park. One type of festival could annually celebrate
the watershed with stream cleanups. The stream is expected to carry 45 cfs during the
two-year storm. The streams meanders are dimensioned for natural dynamic equilibrium with the flow. Bioengineering (the use of living plants in combination with nonliving
materials to stabilize streams and slopes) is used to protect the banks during 2-year and
10-year storms.
Around the edges of the park, street pavements are narrowed to reduce impervious
cover and allow infiltration while adding more parking spaces on permeable edges. The
pervious parking stalls are made of concrete pavers with grass, over a gravel bed
(Figure I.4.6). The open-celled paver surface and the deep gravel storage basin beneath it
adapt the pavement construction to the regions frequent frosts and fine-textured, slowly
permeable soil.
In the residential areas all around and above the park, roof leaders, street gutters, and
drainage inlets are disconnected from the storm sewer system. Their drainage is diverted
into swales and across vegetated slopes in and around the park. Excess runoff remaining
in the streets is conveyed to the parks wetlands and swales for treatment.
Stone traces through the park mark lines of old mining features. Street trees are
added for air and water quality improvement. The combination of strategies preserves
and celebrates the natural and cultural history of the area. The improved access to the
park promotes its use. Opportunities to learn about the hydrologic strategies are available
through interpretative signs and guided tours.
Edgewood Crossroads
Edgewood Crossroads is located near the center of the Nine Mile Run watershed. It is the
public center of the Borough of Edgewood, where a historic train station fronts on busy
Swissvale Avenue. The old train station is the only one that the famous 19th-century
architect Frank Furness designed on this side of the Allegheny Mountains. Across the
street are old storefront commercial buildings, a church, and a school. Nearby are Edgewoods town hall, public library, community swimming pool, and numerous well-kept
old residences.
Residential streets converge from several directions. Public buses stop at the street
intersection; the old railroad bed is slated to become the route of a regional busway.
78
Figure I.4.6 Modification of a Hunter Park street to reduce impervious surface and increase infiltration, while increasing the tree canopy and the availability of parking. Note the gravel-filled storage/recharge bed under the parking stall and sidewalk. (From Ferguson, B.K., Pinkham, R., and
Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky
Mountain Institute, Snowmass, CO, 1999. With permission.)
Numerous pedestrians, especially children, move between their homes and community
facilities along the sidewalks, across the street intersection, and through the railroad
underpass.
Edgewood has dozens of civic groups, and the social closeness almost of a village. In
the minds of the local Edgewood people, the cluster of streets, structures, and open spaces
around the old train station is the unified center of their community. Protecting and
enhancing the sense of community is the central task of any urban design here.
The impervious streets, roofs, and sidewalks of the site and its catchment (Figure I.4.7)
generate runoff that ponds up in the street intersection, disrupting pedestrian and vehicular
traffic. Eventually, it flushes into storm sewers, carrying oils and other pollutants, while
denying recharge of groundwater. As in many parts of the Pittsburgh region, some roof
leaders here are connected to sanitary sewers, contributing to sewer overflows downstream.
The proposed design. The design for this site is based on relationships between water
systems, urban design, and social values (Figure I.4.8). It integrates the following community issues: reinforcing the social and physical sense of community, preserving public open
spaces, reinforcing pedestrian access, eliminating street flooding, and bringing Edgewood
into compliance with federal water quality standards by separating storm drainage from
the sanitary sewer system.
In the small community park facing the train station, a plaza is developed to be the
gateway to a public greenway in the new transit corridor. Here, a prominent stormwater
restoration facility integrates stormwater solutions and public education with urban
design (Figure I.4.9). The center is a depressed bowl, with a porous block bottom that
retains and infiltrates stormwater. During rainfall, about 30 days per year, the depression
diverts flood waters off the street and the surrounding plaza; it fills and then slowly drains
over a 1- or 2-day period through an infiltration bed beneath the plaza. On dry days, the
plaza and the bowl are for communal gathering and play; the wall around the bowl is for
Chapter I.4:
Figure I.4.7 The Edgewood Crossroads site, catchment, and surrounding area. (From Ferguson, B.K., Pinkham, R., and Collins,
T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO,
1999. With permission.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)

79
80
Figure I.4.8 The proposed plan for Edgewood Crossroads restoration and redevelopment. (From
Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for
Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color
Figure I.4.9 The demonstration infiltration basin at Edgewood Crossroads in the proposed public
plaza by the train station. The basin provides a community gathering place. During large storms,
it captures runoff, which percolates into the subsoil within one to two days. (From Ferguson, B.K.,
Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative
Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.)
Chapter I.4:
81
Figure I.4.10 Retrofit of a typical Edgewood residence to disconnect the roof leaders from the
sanitary sewers and recharge the roof runoff into the ground using an infiltration trench. (From
Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.)
sitting. Permeable unit pavers continue from the plaza across the street intersection, to
strengthen pedestrian connections.
The street intersection receives stormwater runoff from a catchment around and uphill
from the crossroads. The 73 acres of the catchment are occupied by the community center
area, several schools and churches, and numerous, relatively large, Victorian single-family
homes. The proposed design manages all stormwater within the catchment by incorporating infiltration measures into lots and landscapes upslope from the site itself.
Open lawns and play fields in the catchments small parks and institutional grounds
are open spaces that can serve the dual purposes of recreation and runoff control. Groundwater recharge beds could be constructed under these areas, while maintaining their
surface uses for sports and parks. For example, the retrofit of a playing field to maximize
infiltration would include aggregate beneath the turf. If a bed of gravel 18 in. deep with 40%
storage volume were provided over the entire 6.2 acres of reasonably available area, it could
infiltrate the entire volume of a 2-year storm collected from an area of 18 acres. An alternative
construction of preformed infiltrator chambers could provide the same capacity.
An additional strategy diverts the runoff of residential roofs into on-lot infiltration
basins to significantly reduce stormwater inflows to the sewers (Figure I.4.10). Infiltration
and recharge features can be shaped to each individual lot. For example, a large residence
has half of its 2,500 ft2 roof area draining to the front and rear yards, respectively. For each
half of the roof, a bed of aggregate or infiltrator chambers with a storage capacity of 208 ft3
(1,560 gal) would infiltrate all the runoff from all rain events up to and including the
2-year storm. The bed or trench must be properly spaced away from the house to avoid
leaking of water into the basement.
Porous pavements at the parking lots of churches and other public places infiltrate
additional stormwater (Figure I.4.11). Among the alternatives in porous pavement construction suitable for local soils, frosts, and traffic loads are masonry pavers with open
joints, a bituminous mix with open-graded aggregate, or gravel with a layer 4 in. deep of
#89 fines over a layer 24 in. deep of uniform-graded aggregate 2 in. in diameter.
Finally, increasing the urban forest throughout the catchment reduces runoff, moderates urban climate, improves air quality, and reduces noise. A dense vegetative structure,
such as trees, shrubs, and native ground covers, absorbs more rainwater than a turf slope
and is more resistant to erosion during intense storms.
82
Figure I.4.11 Proposal for porous pavement to be retrofitted into parking areas of Edgewood Crossroads catchment. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The
Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999.
With permission.)
Sterrett School
Sterrett School is located near the headwaters of Fern Hollow, which is Nine Mile Runs
largest tributary. It is a middle school in the midst of Pittsburghs South Point Breeze
neighborhood, adjacent to Frick Park. Sidewalks connect homes, shops, and the school
grounds safely and conveniently; pedestrians use them actively. The school shares a city
block with eight homes (Figure I.4.12).
In the low area between the houses, a combined sewer line follows the course of the
original streambed of what was once called Salamander Creek. Drainage inlets between
the houses sometimes back up and flood. Local culvert overflows contribute to flooding
in the basements of homes. The sites impervious roofs, streets and sidewalks, driveways
and parking lots dump runoff into the combined sewer, contributing to polluting overflows downstream in Fern Hollow and Nine Mile Run.
The proposed design. The proposal integrates a variety of stormwater measures with
educational programming and neighborhood improvements (Figure I.4.13). For the school
building itself, the plan calls for cisterns to collect roof runoff. The 16,000 ft2 roof generates
a lot of runoff 3,300 ft3 (25,000 gal) of water during the 2-year storm. Diverting this
large volume into the cisterns by disconnecting the schools downspouts from combined
sewers has a big effect on downstream overflows and pollution. One possible form of
cistern is a transparent water wall that would let students monitor the water level in
relation to rainfall and water use (Figure I.4.14). Some water can also be permanently
stored in the buildings attic as a thermal battery to moderate indoor temperatures.
From the cisterns, the water from the 2-year storm and all smaller storms during the
year is put to productive use on the school grounds, irrigating the schools gardens,
greenhouse, and ball field. Water is also put to indoor graywater uses such as flushing
toilets and urinals. Also, a greenhouse utilizing water collected from the schools roof
provides a teaching tool for explaining the water cycle and the role of the school and the
neighborhood in the watershed.
Runoff from a storm larger than the 2-year storm, or a rapid succession of smaller
storms, will exceed the capacity of the cisterns. Overflow from the cisterns will flow
Chapter I.4:
83
Figure I.4.12 The Sterrett School site and surrounding area. (From Ferguson, B.K., Pinkham, R., and
Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky
Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version available at www.gsd.
through the school grounds along an Art Creek. The creek follows the path of water
with a mosaic of tiles, embedding childrens poetry and images of animals and leaves.
As water flows farther away from the building, the artificial tiles give way to a
meandering water course of earth and plants, a water garden that brings water into
contact with the ground (Figure I.4.15). The drainage then flows through a vegetated swale
with a gravel infiltration bed, ephemeral ponds, and community gardens for residents.
Where Edgerton Avenue crosses the path of the old stream, it has barred the movement
of water. Water that would have flowed naturally into the Fern Hollow ravine flows instead
in combined sewer lines and is not available to the ravine ecosystem. Closing off this street
would cause little traffic disruption, and regrading to remove the street embankment
would prevent local basement flooding and restore natural drainage. The water would
enter the ravine via a boulder cascade under overhanging willows. A pedestrian bridge
would maintain access along the old street alignment.
84
Figure I.4.13 Redevelopment and restoration plan for Sterrett School and nearby residences. (From
Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color
Figure I.4.14 Cisterns for collecting roof waters in the form of a transparent water wall alongside
the Sterrett school building. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating
Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.)
Chapter I.4:
85
Figure I.4.15 The water garden water course at the Sterrett School. (From Ferguson, B.K., Pinkham,
Rocky Mountain Institute, Snowmass, CO, 1999. With permission.)
Porous materials replace impervious pavements in the schools parking lot, playground, and sidewalks. The runoff will be further reduced by tree plantings, where the
canopy intercepts rainwater during small, frequent storms at the same time it moderates
air quality and temperature.
The design reduces the runoff from the site to the level of that from a naturally wooded
site. The principal effect comes from disconnecting stormwater drainage from the combined sewer system: diverting the roof runoff into cisterns and vegetated swales, and
allowing water to infiltrate as it flows over grass slopes and in broad open swales, instead
of buried culverts. These measures, together, disconnect the drainage from 90% of the site
area, reducing the 2-year runoff into the culvert to only 40% of its previous volume.
Further runoff reduction comes from specially constructed storage facilities. The
design for the infiltration swale illustrates how a quantity of storage is created. A gravelfilled infiltration trench running under the 500-ft length of the swale provides 2,800 ft3
(21,000 gallons) of storage, based on a depth of 4 ft, a width of 4 ft, and a storage ratio
(volume of stored water per total volume of gravel-filled trench) of 0.35. On the swale
surface, during large storms, ponding up to 6 in. deep and 12 ft across in a gently banked
channel provides an additional 2,000 ft3 (15,000 gal) of storage. The total storage in the
swale is thus 4,800 ft3 (36,000 gal). Because of the gravel and sandstone substrate, infiltration is probably feasible, but if necessary an underdrain system could be added to release
the stored water slowly after the peak of the storm.
The combination of disconnects and stormwater storage reduces the runoff to only
26% of its existing amount, an amount equal to that from a naturally wooded site. These
calculations do not take into account additional reductions due to mulching, pervious
pavements, tree plantings, or bioretention cells (vegetated infiltration basins that capture
sheet flow from parking areas, sidewalks, and paved play areas). These are all extra
restoration capacities.
Regent Square Gateway

The Gateway site is located low in the Nine Mile Run watershed, where the main Nine
Mile Run culvert discharges for the first time into an open channel. Here, the borders of
86
Edgewood, Swissvale, and the City of Pittsburgh converge, by a neighborhood named

Regent Square. At the site, Braddock Avenues on-ramp enters I-376 (the Parkway East),
and the abandoned alignment of Old Braddock Avenue abuts an underutilized commercial
building.
This seemingly neglected place is in fact an extraordinary focus for the Nine Mile Run
watershed and its people, for here the culverted stream first comes into full view in an
open channel, and here the historic plan for Frick Park has always foreseen a major eastern
public gateway. This is the junction between the upper, developed, urban portion of the
watershed and the lower, open, natural part in Frick Park. This highly visible place is the
physical confluence of the watersheds stream flows, municipal jurisdictions, and
potentially community and watershed consciousness.
Local runoff comes through the site from a 64-acre catchment, densely built up with
residences (Figure I.4.16). Impervious rooftops and streets comprise 41% of the total area
of the catchment. During intense storms, a large part of the local runoff currently bypasses
inlets due to the steep slope of Braddock Avenue. Where this runoff reaches the bottom
of the site, it has eroded the edges of the Nine Mile Run channel.
The proposed design. The restoration plan locates small ephemeral ponds (500 to
3,000 ft2 in area) in the neighborhood above the site wherever there is adequate open space,
sufficient drainage area, and appropriate soil for infiltration (Figure I.4.17). They filter the
runoff that occurs during small storms and the first flush of large storms. Stormwater
detention during large storms is not their purpose; high peak flows are allowed to pass
through without additional ponding, so they will not combine with relatively long, slow
peak flows on the main stream. These ponds will treat runoff during every storm, replenish
groundwater to support stream base flows, and allow only the excess water from occasional large storms to enter the stream directly.
The site plan accommodates multiple uses (Figure I.4.18). The large upper story of the
building is suitable for profitable reuse as a single-occupant retail store. Runoff from the
upper, retail parking lot flows onto grass filter strips to enhance water quality
(Figure I.4.19). Excess runoff passes through inlets and perforated pipes into banks of
aggregate filter material. The embankments are capped with topsoil for rooting of trees
and other vegetation. An underdrain collects the filtered excess and discharges it like
spring flow to trickle into the Nine Mile Run channel. In this parking lot and throughout
the site, plantings of trees and shrubs intercept a portion of rainfall.
The lower levels of the old building can be reused to serve visitors to Frick Park and
the greenway. The citys Parks Department could use indoor space for watershed education and research. Private retailers could offer bicycle rentals and food services. At the
trail head, the buildings facilities could provide trail guide information, trash disposal,
and public restroom facilities.
Some runoff from upslope areas will continue to wash over the site during occasional
large storms even after construction of small infiltration basins in the neighborhood above.
Surface runoff entering the site from Braddock Avenue is diverted into the formerly
neglected area between Old Braddock Avenue and the parkway ramp. Here, a series of
terraced basins reclaim the area to filter the runoff (Figure I.4.20). The sculptural earth
forms symbolize the fluvial processes with which they unite to mitigate the runoff from
the urban watershed. Low-flow and first-flush runoff infiltrates into the basins and is
filtered. Underdrains collect the filtered water and add it to the spring flow for discharge
to the stream channel. Larger flows spill gradually over the surface from one terrace to
the next, discharging through a spillway before high-peak flows arrive on the main stream.
At the entry to Frick Park, Old Braddock Avenues remnant trolley tracks disappear
under a highway embankment, marking the end of a former era, while Nine Mile Run
Chapter I.4:
87
Figure I.4.16 The Regent Square Gateway site, catchment, and surrounding area. (From Ferguson,
B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative
Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version
available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
88
Figure I.4.17 Typical shallow infiltration basin proposed for the neighborhood above the Regent
Square Gateway site. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater:
The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO,
1999. With permission.)
emerges from its culvert, marking the beginning of a new. Filtered spring flow trickles
across the area from the sites underdrained filter embankments, symbolizing the restoration of the watersheds soil and streams through revitalizing retrofit and redevelopment.
Here also is the outfall of the main Nine Mile Run culvert. The headwall of the culvert is
retrofitted with scuppers for the trickling spring flow and a spillway for the terraced basin
overflow, the surrounding area is redesigned to allow safe low-water access and highwater viewing of the stream, and the channel bed is fitted with sculptural elements that
will dissipate the energy of high, wet weather flows emerging from the culvert.
Policies for restorative redevelopment

The charrette identified four overarching policy objectives to support restorative redevelopment. They are as follows:
Establish a permanent coordinating body with the authority and financial security to
plan, maintain, and manage the watersheds interrelated infrastructure, natural
processes, and urban land uses. The organization must transcend municipal
boundaries. It should unite the responsibilities of infrastructure management and
ecosystem protection.
Manage the watersheds sewer and stormwater infrastructure for efficiency, reduced costs,
and reinforcement of natural processes and community vitality. Site-level measures
such as those identified by the charrette design teams should be considered part
of the infrastructure of the watershed and integrated with conventional infrastructure remediation in order to focus limited community resources on effective systems that produce multiple benefits.
Restore the watersheds hydrologic and ecological processes in a manner that utilizes and
supports infrastructure rehabilitation and community redevelopment. This includes rehabilitating urban runoff by reconnecting storm drainage with the natural
capacities of the watershed. It also includes restoring natural stream, wetland, and
forest habitats in critical areas.
Chapter I.4:
Figure I.4.18 Plan for the rehabilitation and redevelopment of Regent Square Gateway. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating
Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With permission.) (Color version

89
90
Figure I.4.19 Embankment to filter parking lot runoff at Regent Square Gateway: 1) Grass filter strip
and inlet; 2) Perforated pipes; 3) Underdrain. (From Ferguson, B.K., Pinkham, R., and Collins, T.,
Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment, Rocky Mountain
Institute, Snowmass, CO, 1999. With permission.)
Figure I.4.20 The terraced, underdrained basins, and rehabilitation of parking lots at the Regent Square
Gateway site. (From Ferguson, B.K., Pinkham, R., and Collins, T., Re-Evaluating Stormwater: The Nine
Mile Run Model for Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. With
permission.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Enable, support, and require economic revitalization that reinforces infrastructure management and watershed restoration. Communities, agencies, and developers can
structure redevelopment programs and projects in ways that support sewer rehabilitation and restoration of beneficial natural processes.
Chapter I.4:
91
For each of these objectives, the charrettes policy team prepared detailed rationales
and action plans (Ferguson, Pinkham, and Collins, 1999; Re-Evaluating Stormwater
Technical Appendix, 1999).
Status
At this writing, further planning for redevelopment of the charrette sites is proceeding,
albeit slowly. The Sterrett School site is not in need of immediate retrofit or redevelopment.
Edgewood Crossroads continues to be bound up in debates over the proposed regional
mass transit system through the site. Wilkinsburg has officially designated Hunter Park
and its surrounding neighborhood as a redevelopment zone and is seeking funding for
projects there. Features of the Regent Square Gateway plan have been incorporated into
a Pittsburgh Parks master plan. The commercial property there remains for sale.
Meanwhile, the restoration and policy concepts put forward through the charrette and
the overall Nine Mile Run Greenway Project are gaining currency in the watershed. Edgewood and the Pennsylvania Environmental Council have secured funding for a stormwater
infiltration retrofit demonstration project at a municipal parking lot not far from the Edgewood Crossroads site. Water sensitive designs for several sites within Frick Park and the steel
slag redevelopment site were endorsed by watershed citizens and the STUDIO for Creative
Inquirys advisory board in 1999. In 2000, the U.S. Army Corps studied and agreed to these
designs, with minor modifications, in preparation for a $7.7 million Corps aquatic ecosystem
restoration program in the park and greenway, funded by the City of Pittsburgh and the
federal government under Section 206 of the Water Resources Development Act of 1996. Also
in 2000, as the STUDIO stepped back from its driving role in Nine Mile Run restoration,
interested citizens incorporated a nonprofit Nine Mile Run Watershed Association. Initial
funding for the associations director has been secured.
In 1999, the four watershed municipalities entered into a joint agreement to inspect
and monitor sewers for illegal connections, develop sewer overflow remediation plans,
and implement those plans. This work is proceeding well, and new institutional forms
for improved watershed management are under study, including a scheme to change the
joint drinking water authority serving the watershed and several other nearby communities into an integrated water, sewer, and stormwater authority.
The projects also have helped catalyze water management initiatives and understandings beyond the watershed boundaries. Ecological restoration is a major component of
the Pittsburgh Parks master plan for three major parks in addition to Frick Park. A
nationally funded wet weather management demonstration program, 3 Rivers Wet
Weather, Inc., has notified Allegheny County communities that it will entertain funding
requests for projects incorporating the sorts of techniques illustrated by the charrette
designs. ALCOSAN, the regional sanitation authority, is investigating stream daylighting
opportunities in conjunction with its efforts to identify and remove stream flows from
combined sewers. With funding from the Heinz Endowments, the Three Rivers 2nd
Nature initiative by 3 Rivers Wet Weather, Inc., ALCOSAN, and the STUDIO for Creative
Inquiry is now underway to inventory and assess the ecosystem and community values
of Allegheny Countys rivers by conducting water quality monitoring, riverbank biotic
assessments, a study of public access issues, and a review of tributary streams for restoration and daylighting potential. This effort will likely add to the impetus for restorative
redevelopment in the uplands surrounding the rivers.
Patterns of restorative redevelopment

The results of the charrette exemplify certain principles, or patterns, of restorative redevelopment. These patterns of design and process are key concepts for restoring and
92
revitalizing old urban watersheds everywhere (Ferguson et al., 1999). Projects that follow
these patterns can have an important effect on sewer overflows; add incrementally to a
watersheds long-term, broad-based reduction in impervious surfaces and generation of
stormwater; and contribute significantly to the economic and social health of a community.
Patterns of design
Make components multifunctional
Everything that is done in a retrofit or redevelopment project should produce multiple,
mutually reinforcing benefits. When a component is multifunctional, it attracts advocates
promoting each of its several functions; it attracts broad community and political support.
For instance, stormwater has traditionally been moved off city roofs and streets through
a single-purpose system of underground pipes. Instead, it can be kept on the surface,
recreating a creek that was lost, or infiltrated into the soil to recharge the groundwater
and nourish vegetation in either case providing ecosystem benefits such as habitat for
wildlife and stream base flow support, human benefits in experiencing the beauty and
wonder of natural systems, and financial benefits in reduced municipal costs of maintaining hidden infrastructure.
Whenever an important component of a project appears to be an undesirable cost,
seek ways to shape it so that it acquires additional desirable benefits. The project and
maintenance budget is thereby enlarged as the cost becomes absorbed into the provision
of other necessary functions. Multiple functions as various as water quality improvement,
employment, housing, separation of storm drainage from sanitary sewers, parking
improvements, noise reduction, pedestrian safety, temperature moderation, and social
equity can and should be found in the design of every building, street, sidewalk, park,
water course, drainage system, residential yard, and institutional landscape.
Use every square inch

Cities are crowded places. The solution to a watershed-wide problem has to be on-site,
on every site, because there is nowhere else to go. There is no room to add conventional
detention basins or treatment plants. Successful restoration and revitalization depends on
utilizing every square inch of a retrofit or redevelopment project for positive, multiple
functions. Every component is in the midst of community life and must have positive
community benefit in addition to technical function. The cumulative public benefits are
enormous. There must be a constant search for restoration and revitalization opportunities
on additional sites.
Use freely available natural processes

Freely available natural processes are capable of working for the great benefit of watershed
restoration. Vegetated soil absorbs rainwater, and the chemical and microbial processes
of the soil capture and degrade most pollutants that may be present. The infiltrated water
recharges groundwater tables and restores flows to streams. These processes reduce peak
flows and erosion, eliminate sewer overflows, prevent and mitigate pollution, and sustain
watershed ecosystems. Taking advantage of them enacts the idea of green infrastructure,
which broadens the conception of a stormwater infrastructure to include the capacities of
soil and vegetation to absorb water and filter pollutants. This is a smarter, cheaper
approach to infrastructure because it puts nature to work and reduces the work humans
must do, in contrast to the more active systems of pipes and facilities for conveyance and
mechanically dependent treatment.
Chapter I.4:
93
Use disconnections and reconnections

Sewer overflows are usually the biggest pollutant sources in the watersheds where they
exist, such as Nine Mile Run. In such places, the drainage from impervious surfaces should
be disconnected from sanitary sewers at every opportunity, no matter how small. To
disconnect rooftop drainage, each downspout can be detached from sewers and routed
to dry wells, water gardens, and cisterns. To disconnect pavement runoff, the drainage
from driveways and walkways can be pitched away from street gutters and onto vegetated
soil; large parking areas can be broken up with infiltration islands or served by underground storage/recharge beds; street drainage inlets can be detached from combined
sewers, and their stormwater diverted into vegetated swales. These techniques can also
remove water from storm sewers to improve water quality and restore watershed function.
Drainage that is disconnected from sewers in these ways is reconnected with its
natural path in contact with soil and vegetation. The reconnection with natural processes
reduces the volume of surface runoff, filters the pollutants, replenishes the groundwater,
and maintains stream base flows. The volume of stormwater, which once seemed a hazard
and a nuisance, is turned into a resource and a productive public benefit.
Patterns of process
Cooperate among disciplines
In the process of conceiving and implementing retrofit and redevelopment projects, members of different professions have insight into different problems and opportunities of
watersheds and communities, and different types of skills for analyzing and developing
them. All professionals need to be members of the project team. History, society, economy,
quality of life, art, engineering, and ecology do in fact interact in retrofit and redevelopment projects, because all these processes share the same urban environment. Taking them
all into account, as an interdisciplinary team, produces a sound multifunctional result.
The choice of individual participants may be important to a projects success. Individuals
must be open to the unanticipated insights of members of other disciplines and willing
to work with them in design.
Find out what is possible

Diverse, flexible, economical techniques for treating and storing stormwater within urban
retrofit and redevelopment projects are being proven in applications throughout the
United States. Many techniques useful in the new development context can be adapted
and retrofitted into already-built sites or applied upon redevelopment. Developers, public
officials, and citizens need to be aware of the alternatives that are available.
Engage the community

Each city and its respective communities have a unique social and political history, style
of governance, method of public discourse, and capacity for action. We must carefully
define the local application of potential solutions and seek locally integrated forms of
innovation. The citizens who must live with the consequences of retrofit and redevelopment projects should have a substantive role in forming them. Collaborative, communitybased efforts are key to developing sustainable approaches to issues as broad as sewer
overflows, ecosystem restoration, and community development. If functions and benefits
in these areas are to be coordinated and maximized, everyone must be involved in the
search for solutions. In that process we build cohesive cultural forces invested in longterm success.
94
Conclusions
The Nine Mile Run stormwater charrette brought together interdisciplinary experts in
restorative design and policy from various parts of the country with Pittsburgh natives
profoundly experienced in unique local conditions. Their work served as modeling experiments that tested the question: Are these kinds of ideas feasible in the specific conditions
of the Pittsburgh region? The results demonstrate that numerous techniques, old and new,
can be applied in the Pittsburgh region, specifically in the old urban neighborhoods in
ways that are economical, effective, and supportive of economic vitality and quality of life.
As our older cities were built, the cumulative impacts of transforming the landscape
mounted, and municipalities replaced natural systems with cost-intensive, conveyancebased infrastructure. Now, when much of the older infrastructure fails to perform to
todays or even yesterdays standards, we have an opportunity to reconsider the form and
function of the urban landscape and ultimately integrate gray infrastructure and green
infrastructure into a seamlessly operating whole. By adopting the mindset and techniques
of restorative redevelopment that is, by linking stormwater management with urban
economic and social development agendas through physical measures that provide multiple benefits designers and other advocates of water sensitive development can turn
old urban watersheds and their waterways from problems into environmental and community assets.
Acknowledgments
The authors thank Bruce Ferguson, who provided considerable assistance with the Nine
Mile Run ecosystems and infrastructure charrette and was the lead author of the ReEvaluating Stormwater report (Ferguson, Pinkham, and Collins, 1999). Portions of this paper
are based on that report. The 60-plus participants in the charrette and many staff of the
authors organizations the STUDIO for Creative Inquiry and Rocky Mountain
Institute are also co-authors of the restorative redevelopment approach. We thank
TreePeople, of Los Angeles, CA, for the inspiration provided by its Second Nature
charrette and T.R.E.E.S. program. We gratefully acknowledge the generous funding provided to our work in Nine Mile Run by The Heinz Endowments and the Pennsylvania
Department of Conservation and Natural Resources. We also thank the following key
partners in these efforts: Allegheny County Sanitary Authority (ALCOSAN), Allegheny
County Health Department, Borough of Wilkinsburg, Borough of Edgewood, Borough of
Swissvale, City of Pittsburgh, Pittsburgh Sewer and Water, and 3 Rivers Wet Weather, Inc.
Literature cited
Ample Opportunity: A Community Dialogue, Carnegie Mellon University, Pittsburgh, PA, 1998.
Condon, P. and Moriarty, S., Eds., Second Nature: Adapting LAs Landscape for Sustainable Living,
TreePeople, Los Angeles, 1999.
Ferguson, B.K., Lipkis, A., Pinkham, R., Condon, P.M., and Collins, T., The future of old urban
watersheds, p. 260-262 in 1999 Annual Meeting Proc. Amer. Soc. Landscape Architects, American
Society of Landscape Architects, Washington, DC, 1999.
Restorative Redevelopment, Rocky Mountain Institute, Snowmass, CO, 1999. Available at
www.rmi.org as a downloadable file in the library section or for purchase in the bookstore section.
Lipkis, A., President, TreePeople, Los Angeles, Personal communications, 2001.
Low Impact Development Center homepage, http://lowimpactdevelopment.org/mainhome.html,
downloaded April 25, 2001.
Chapter I.4:
95
Nine Mile Run Rivers Conservation Plan, Carnegie Mellon University, prepared for the City of Pittsburgh; Carnegie Museum of Natural History; Pennsylvania State University; and Commonwealth of Pennsylvania, Dept. Conservation and Natural Resources, Rivers Conservation
Plan Program, 1998.
Parsons Engineering Science, Metcalf and Eddy, and Limno-Tech, Sanitary Sewer Overflow (SSO)
Needs Report, prepared for the U.S. EPA, Washington, DC, 2000.
Pinkham, R., Daylighting: New Life for Buried Streams, Rocky Mountain Institute, Snowmass, CO, 2000.
Recycle Your Rain, Flyer from Toronto, Ontario, City Works Services, 1998.
Re-Evaluating Stormwater: The Nine Mile Run Model for Restorative Redevelopment Technical Appendix,
Rocky Mountain Institute, Snowmass, CO, 1999. Available at www.rmi.org as a downloadable file in the library section.
T.R.E.E.S. Project Overview, Web page maintained by TreePeople, http://www.treepeople.org/
trees/, 2001.
U.S. EPA, 1996 Clean Water Needs Survey Report to Congress, Washington, DC, Publ. 832/R-97-003,
1997.
U.S. EPA, Economic Analysis of the Final Phase II Stormwater Rule, Washington, DC, Publ. 833-R-99-002,
1999.
96
Response
Raising consciousness through interdisciplinary design workshops
It is important to direct attentions of water sensitive designers not just to issues concerning
new developments but also to the restorative redevelopment of older urban regions
affected by crumbling or inadequate water infrastructure. As Richard Pinkham and Timothy
Collins correctly point out, restoration involves more than technological solutions if it is
to succeed; it also must address social and economic concerns in the watershed. When
accomplished effectively, redevelopment and retrofit projects can revitalize communities
by improving the overall livability of such areas.
The design approach taken in this case study is one of placing small, stormwater
treatment and management measures in a dense urban framework. These measures range
from disconnecting storm drains, establishing cisterns and water gardens, reconfiguring
driveways, adding porous pavements, and daylighting streams. Together, these approaches
are those represented in the new paradigm of low-impact development as discussed by
Larry Coffman in Chapter I.5 and applied by Michael Clar in Chapter I.15, Thomas Liptan
and Robert Murase in Chapter I.6, and Robert France and Philip Craul in Chapter I.7. The
conceptual design project described in this chapter by Pinkham and Collins is of particular
interest in its linking of water management with brownfield redevelopment, in its recommendations for a review of existing institutional arrangements and policy initiatives to
increase the chances for success, and in its straightforward message of turning stormwater
from being regarded as a nuisance or hazard to being looked upon as a resource (see also
the case study described by Glenn Allen in Chapter I.13).
What is perhaps most interesting about this study is the use by Pinkham and Collins
of an intense 3-day charrette design workshop with 60 local and regional landscape
architects, engineers, architects, artists, planners, policy analysts, and neighborhood citizens in which to raise public consciousness and education about the opportunities for
restorative redevelopment. In the end, it is precisely such collaborative, community-based
efforts that are the key to developing sustainable water sensitive designs.
chapter I.5
Low-impact development: An alternative

stormwater management technology
Larry S. Coffman
Abstract
Low-impact development, or LID, is the general term used to describe an alternative
innovative comprehensive suite of lot-level land development principles and practices
designed to create a more hydrologically functional urban landscape to better maintain
or restore an ecosystems hydrologic regime in a watershed. This new approach combines
a variety of conservation strategies, minimization measures, strategic timing techniques,
integrated smallscale site-level management practices, and pollution prevention measures
to achieve desired stormwater management or ecosystem protection goals. When these
various strategies are integrated into the site design, they create a distributed decentralized
management approach. Through the combined cumulative beneficial impacts of all the
possible integrated LID site design and management techniques, it is now technically
feasible to develop a site with little impact on hydrology or water quality. The basic goal
of LID is to engineer a site with as many small-scale retention, detention, prevention, and
treatment techniques as needed to achieve the hydrologic functional equivalent to predevelopment conditions.
Background
The need for more effective economically sustainable stormwater management technology
has never been greater. With the wide array of very complex and challenging ecosystem and
human health protection goals and regulatory requirements to be addressed by stormwater
programs, many practitioners are beginning to question the efficacy of conventional stormwater management technology to meet these challenges. Communities are struggling with
the economic reality of funding stormwater infrastructure maintenance; inspection, enforcement, and public outreach necessary to support an ever-expanding and aging infrastructure
and continued growth. Even more challenging are the exceptionally high costs of retrofitting
existing urban development using conventional stormwater management end-of-pipe
practices to protect the integrity of receiving waters and living resources.
To assist local governments in their efforts to develop more effective economically
and environmentally sustainable stormwater management programs, Prince Georges
County, MD, Department of Environmental Resources (PGDER), with the support of the
1-56670-562-2/02/$0.00+$1.50
97
98
U.S. Environmental Protection Agency (EPA), developed a two-volume set of national

guidance manuals on the LID approach (PGCDER, 1999). The EPA provided grant funding
to assist PGDER in their efforts to develop national guidance manuals to make this
technology available to other local governments. This new approach is a significant step
toward advancing the state-of-the-art of stormwater management and will provide valuable and useful tools for local governments in their efforts to control urban runoff for both
new development and redevelopment.
Prince Georges County received the EPAs 1998 first-place National Excellence Award
for Municipal Stormwater Management Programs for its pioneering work on LID technology and local LID manual (PGCDER, 1997). Many other efforts are currently underway
across the nation to further advance LID technologies such as improving the sensitivity
of current hydrology and hydraulic analytical models and developing of new microscale
control approaches and practices for highway design, urban retrofit applications, and
numerous monitoring efforts.
Some practitioners have found LIDs site-oriented microscale control approach to be
controversial, because it often conflicts with building codes, challenges conventional
stormwater management paradigms, and is perceived by some to accommodate urban
sprawl. However, many have found LIDs distributed source control technology to be an
economical common sense management approach that can be used to achieve superior
environmental protection for new development and provide extremely useful new tools
to retrofit existing development.
This chapter only briefly outlines the LID approach and its basic theories, philosophy,
control principles, and practices. A more detailed explanation on the planning, design,
and application of LID technologies is provided in the two-volume national LID manuals
(PGCDER, 1999). These two volumes are a good introduction to LID; however, it should
be noted that this technology is rapidly evolving, adding new techniques to the lot control
principles and practices. Across the country and around the world, an amazing number
of new lot-level control techniques are being developed such as multifunctional landscapes
features; restoring soil functions; the use of bioretention plant/soil filtration; providing
lot-level storage, runoff capture, and use; and modifying timing and pollution prevention.
All these techniques go beyond the scope of the original manuals.
For more information on how to obtain copies of the national LID guidance manuals,
call Prince Georges Countys Department of Environmental Resources at (301) 883-5834.
It is hoped that the LID national manuals will help to stimulate debate on the state of
current stormwater management, watershed protection and restoration technology, and
its future direction.
LID in general
LID is a powerful technology that allows development to take place in a manner that can
preserve water-related ecological functions/relationships and maintain development
potential. LID achieves stormwater management and ecosystem protection goals through
the cumulative effects of a wide array of techniques. LID uses new site design planning
principles, microscale management practices, and pollution prevention to create environmentally sensitive landscapes that allow the developed area to remain a functioning part
of the ecosystems instead of being dysfunctional and apart from the ecosystem.
LID maintains or restores the hydrologic regime and manages stormwater by fundamentally changing conventional site design to create a hydrologically functional landscape
that mimics natural ecological hydrologic functions. LID provides numerous tools to
maintain the predevelopment volume relationship between rainfall/runoff, recharge,
interflow, and evaporation. This is accomplished in five basic steps for new development:
Chapter I.5:
Low-impact development: An alternative stormwater management technology
99
1. Apply conventional conservation planning techniques to define the building envelope. These would include master zoning and environmental features such as
streams, wetlands, forests, agricultural/historical preservation, trails, and open
spaces.
2. Apply impact minimization strategies to the extent practicable (or allowable) by
reducing imperviousness, reducing use of pipes, saving recharge areas, and minimizing clearing and grading.
3. Maintain predevelopment time of concentration by strategically routing flows to
maintain travel time throughout the site.
4. Apply distributed integrated management practices to treat, detain, retain, and
infiltrate runoff to restore predevelopment conditions. These practices would include use of multifunctional open swales, bioswales, infiltration practices, bioretention (rain gardens) water capture and use (rain barrels), and depression storage
in conservation areas.
5. Provide effective public education and socioeconomic incentives to ensure property owners use effective pollution prevention measures and maintain on-site
management practices.
LID is a new and creative way of thinking about site design to make every site
landscape, roadway, and building feature (green space, landscaping, grading, streetscapes,
roads, parking lots, roofs, etc.) multifunctional, multibeneficial, and optimized to manage,
treat, or use runoff to maintain/restore hydrologic functions.
The effective use of LID site design techniques can significantly reduce the cost of
providing stormwater management. Savings are achieved by eliminating the use of stormwater management ponds; reducing pipes, inlet structures, curbs, and gutters; and resulting in less roadway paving, less grading, and less clearing. Where LID techniques are
applicable and depending on the type of development and site constraints, stormwater
and site development design, construction, and maintenance costs can be reduced by 25
to 30% compared to conventional approaches.
The creation of LIDs lot-level management principles and practices have led to the
development of new tools to retrofit existing urban development. Microscale decentralized
management practices to recharge, filter, retain, and detain runoff can be easily integrated
into the existing green space, parking lots, building design, landscaping, and streetscapes.
These integrated management practices, or IMPs, can be constructed as part of the routine
maintenance and repair of urban infrastructure, requiring less capital outlay for retrofit
compared to conventional large-scale, highly capitalized centralized approaches. LID
microscale techniques have been shown to reduce the cost of retrofitting existing urban
development. Reducing urban retrofit costs will increase the ability of cities to implement
effective retrofit programs to reduce the frequency and improve the quality of combined
sewer overflows (CSOs) and improve the quality of urban runoff to protect receiving
waters.
Why develop an alternative stormwater technology?

Those jurisdictions that have used conventional pipe and pond technology (centralized
BMP treatment) over the past 20 years have gained a tremendous amount of experience
and insight into the economic and environmental sustainability of a massive stormwater
infrastructure. Essentially, practitioners have learned that there are serious economic,
environmental, public safety, political, and practical limitations associated with many of
the conventional BMPs. LID was designed to address and reduce many of these limitations
and burdens, some of which are discussed in the following sections.
100
Maintenance burdens of a growing, aging infrastructure

As experience is gained with the current management technology, many highly urbanized
jurisdictions are beginning to question the efficacy of traditional structural approaches to
meet complex environmental objectives. They are also finding it harder to fund the inspection, enforcement, and maintenance of programs necessary for the massive stormwater
management infrastructure created by conventional approaches.
Some larger, highly urbanized jurisdictions now have the responsibility for the maintenance, inspection, and enforcement of thousands of BMPs (ponds, infiltration practices,
and filters), thousands of miles of pipes and gutters, and tens of thousands of structures
(inlets, manholes, and catch basins). This infrastructure, like all urban infrastructures, is
growing and aging at the same time. Many of the oldest BMPs have reached their expected
service life and are failing. Most jurisdictions have reached the point where they can no
longer afford to adequately pay for the upkeep of their stormwater BMP infrastructure.
For example, in Prince Georges County (population 800,000) the annual stormwater
maintenance budget is approximately $6.5 million and rising every year by about $250,000.
Generally, most jurisdictions cannot afford to have proactive maintenance programs
for the current suite of conventional BMPs. Maintenance occurs when there are complaints
or a total system failure that results in detectable damage. Many infiltration or underground treatment systems are never inspected or maintained. Survey and studies of these
devices show a failure rate of about 50% after 5 years of operation. In most cases, neither
the property owner nor the local jurisdiction can afford to maintain these BMPs and
therefore do not. For those jurisdictions that do not have a dedicated funding source for
their stormwater programs, the problem of affordability is only compounded, because
they cannot successfully compete for resources against police, education, and fire services.
One might ask why we would continue to build treatment systems that we simply
cannot afford to maintain? Also, just how effective can current BMPs be in meeting our
protection goals if they are rarely or never maintained?
Environmental concerns
Many studies currently demonstrate or strongly suggest that, for example, stormwater
management ponds can create their own unique set of environmental impacts. These
include problems associated with fish blockages, thermal pollution, groundwater contamination, bioaccumulation of toxics, export of nutrients, sediments and toxics during high
flows, and increases in stream erosion. Perhaps the biggest failure of the centralized pond
approach is that it does not, and cannot, replicate predevelopment hydrology. Ponds can
only be designed to reproduce peak discharges. They do not reestablish the predevelopment rainfall/runoff/recharge volume relationships or maintain the natural frequency of
surface discharge. The changes in hydrology resulting from the pipe and pond approach
accelerates stream channel erosion, changing stream morphology and adversely affecting
aquatic habitat structure such as pools, riffles, and shading needed to protect the biological
integrity of aquatic biota. Because ponds have their own set of impacts, it is questionable
if this control technology can be helpful in maintaining or restoring the ecological integrity
of a receiving stream and its biota.
Furthermore, because current management practices only mitigate or lessen the effects
of urban development, there is concern about the cumulative impacts of the widespread
use of conventional mitigation practices. With the use of conventional management, continued growth will allow increases in pollutant loads and fundamental alterations in a
watersheds hydrologic regime. At best, conventional approaches only slow down the rate
Chapter I.5:
Low-impact development: An alternative stormwater management technology 101
of change but allow an overall net increase in adverse environmental impacts including
pollutant loads and hydrodynamic modifications.
Political problems
There are many public complaints and concerns generated by our current predominant
use of management ponds. These complaints deal with issues such as public safety
(drowning and mosquito-borne diseases), lack of maintenance (aesthetic), maintenance
costs and property owner legal liabilities, i.e., insurance costs. Justifying the continued
use of technology that is viewed (real or perceived) by the public as a liability is becoming
problematic as political pressure rises to find more acceptable, sustainable, and safer
solutions.
Practical problems
A dilemma for local governments is that they are confronted with many protection and
restoration goals. They must respond to a wide variety of state and federal regulations
and address unique local needs associated with the adverse impacts of urban runoff. Local
governments have the difficult task of developing complex multiobjective stormwater
management programs. They require multidisciplinary and integrated approaches with
the need for as many tools as necessary to meet the desired objective.
New regulatory programs, such as NPDES Phase II and/or TMDLs, now focus on
specific targeted issues of concern or compliance requirements. Can conventional technologies meet new goals in a cost-effective and sustainable manner? LID provides many new
cost-effective principles and practices that can be added to existing technologies to help
tailor a program that meets the economic and environmental needs of each community.
Philosophical platforms and organizing principles of LID

Better technology or more restrictive land-use policies
Generally, land-use decisions and development are not typically organized around a set
of strong environmental or conservation principles. For the most part, development occurs
and is organized around typical land-use principles based on economic needs, individual
property rights, public policy, and politics. There are exceptions where environmental
laws (wetland and endangered species) can delay development until an acceptable mitigation option is worked out by the regulators or courts. In the end, there is a high degree
of certainty that development will occur (perhaps conditioned) but will not be stopped
based solely on environmental constraints.
Furthermore, most local governments need continued development or redevelopment
to maintain an adequate tax base and are almost always supportive of economic development projects. The problem with relying too heavily on land-use controls (conservation
measures alone) to protect natural resources and receiving waters is that they depend on
firm and continued political support. The practical reality is that political support is
ephemeral, especially in the face of economic development pressures.
Given that development cannot be stopped and that current technology does not
reduce new development impacts to predevelopment levels, the impacts of urbanization
will continue to increase (perhaps at a slower rate). Development of an appropriate
technology that will ensure no net increase in pollutant loads or change in the ecosystem
hydrology is needed. The development and use of better technology is key to protecting
102
our receiving waters and ecosystems from continued and rapidly increasing urbanization.
Better technology does not mean just better, more efficient BMPs. Technology is defined
in the broadest sense as a comprehensive spectrum of planning and design techniques that
balance the appropriate level of conservation, minimization, and control techniques to ensure
land uses will not impact receiving waters and economic development can continue.
The advantage of a better technology approach is that, once established, it is slow to
change and it is application is less susceptible to changes in political points of views or
economic development pressures. Generally, technology is apolitical and can be supported
by growth, no growth, and conservation proponents.
Are we protecting a watershed or ecosystem?

How you answer this question will have a dramatic effect on your protection priorities,
technology, and management strategies. The ultimate watershed protection goal as stated
in the Clean Water Act is fishable and swimmable waters; however, in our attempts to
achieve these goals it has become necessary to understand what an ecosystem is and how
it works. The main focus of the Clean Water Act is to protect the aquatic living resources
and protect human health. The quality of the receiving waters and the integrity of the
aquatic biota are a reflection of and response to what occurs on the land. To improve the
integrity of the water and aquatic biota, we must understand the vital environmental
processes and ecological functional interrelationship between the uplands and receiving
waters.
Watershed-based approach
When we define a watershed, we first place it in the context of topographic boundaries
lines defined by geographical and hydrological features. We further define a watershed
by other easily measurable tangible features such as acres of land cover types, stream
buffer, forests, wetlands, slopes, soil types, etc. Watersheds are compared, evaluated, and
ranked by the percentages of these various features such as stream buffer, forest cover,
open space, and impervious cover. We begin to develop concepts of what is good or bad
based on what is believed to be an appropriate mix of features within each watershed.
We may even place limits on impervious surfaces, believing that less is better. We may
encourage more open space, believing that more is better. We tend to think of the watershed features in much the same way as we do adequate public facilities necessary to
achieve acceptable service or use levels.
When we develop a watershed protection plan, one goal is to protect enough of the
watershed features to meet our perceived service levels and uses. Ultimately, the mix and
amount of watershed features are determined not by environmental needs but by social,
economic, political, cultural, and legal values or constraints. A watershed plan is very
much tied to the general land-use or zoning plan. The ecological effects of the plan are
considered as part of the development scenarios, with the main objective to minimize
environmental impacts by preserving environmental features to the extent practical or
politically allowable.
Implementation of watershed plans may be difficult when watershed boundaries cross
political boundaries. This is because watershed plans rely heavily on a subjective human
construct of what is an acceptable mix of appropriate watershed features and relative to
what is politically acceptable. Watershed plans do not start out with the goal of achieving
fishable or swimmable waters; they merely strive to reduce impacts. What a watershed
plan generally achieves is to define where development is most likely to take place and
where environmental resources might best be protected. The degree to which a plan is
ultimately followed depends on economic and political factors.
Chapter I.5:
In the end, once a watershed plan is implemented, we may or may not achieve the
fishable and swimmable goal or be left with enough space to maintain vital ecological
functions or ensure the long-term sustainability of the receiving waters biological integrity. We are unable to achieve our goals through watershed planning alone, therefore, we
almost always to have supplement these plans with technological solutions, i.e., BMPs.
LID was developed to increase the technological tools necessary to complement watershed
plans and better meet environmental objectives.
Ecosystem-based approach
An ecosystem is hard to define in the context of boundaries or measurable features. An
ecosystems boundaries are fuzzy, fluid, and overlapping and differ based on species.
Much of ecosystem protection research and debate has centered on trying to define and
understand how much space must be conserved to maintain a viable species within an
ecosystem. For most species, the space requirement has yet to be clearly defined. Therefore,
it is difficult, if not impossible, to know or define an ecosystem in the same physical context
as a watershed.
Ecosystem management has been defined by the U.S. Fish and Wildlife Service as
protecting and restoring the natural functions, structure and species composition of an
ecosystem that are interrelated (National Park Service, 1994). From an ecosystem management perspective, to achieve the goal of fishable and swimmable, it is necessary to
protect species, their habitats, and the natural functions that support them. Much attention
has been given to saving species and their habitats. State and federal laws provide legal
mechanisms for habitat protection. What has not received as much attention in ecosystem
management is maintaining the natural functions. Given that it may not be possible to
protect enough habitat to maintain natural functions (due to political and economic constraints), one option would be to develop technology that mimics these natural functions
within the built environment.
Ecosystem functions include the flow, cycling, and processing of nutrients, energy,
and materials (water, sediments, chemicals, and organic materials) through the ecosystem.
From the perspective of achieving the fishable and swimmable goal, the hydrologic regime
is of primary concern. The flow regime (water cycle) of water from uplands to receiving
waters is the primary driving force and mechanism for the transport of materials, energy,
and flow of nutrients through the ecosystem. Hydrology is perhaps the most important
factor that shapes and defines an ecosystem and the nature of and relationship of the
upland to the receiving waters. After all, it is the hydrology that ultimately shapes and
defines a desert or a wetland. Significant changes to the nature of the upland will change
the ecosystem hydrology and affect receiving waters.
We know much more about ecological functions and processes than we do about the
habitat requirements of species. We have monitored, modeled, and studied in detail and
know the most about the hydrologic regime, functions, fate, and transport of materials.
Numerous models can accurately predict the hydrologic regime for surface and groundwater flow and water quality impacts given the types of soils, slopes, and land cover.
From an ecosystem management perspective, maintaining or restoring the ecosystems
natural hydrology regime would be one of the most important goals. The theory of LID
is that reproducing the natural hydrologic regime is the one best thing that can be done
to protect the integrity of the receiving waters.
The goal of LID from an ecological perspective is to achieve a state of hydrologic
homeostasis. This is the state of dynamic self-regulation where the ecosystem is able to
maintain essential hydrologic processes and functions to sustain its viability and to avoid
changes that would destroy it. LID allows for the recreation of a more natural rainfall/runoff/recharge volume relationships and processes that avoid the extremes in, or flashy
104
nature of, the hydrologic regime typical of conventional site designs. If it is technologically
possible to reproduce the hydrologic regime, then it is not necessary to study the ecological
consequences of altering the hydrology associated with conventional land development
practice and mitigation technology. If we cannot get the hydrology right, what hope is
there in achieving the fishable and swimmable goals?
How important is imperviousness?

The amount of total impervious area, or TIA, in a watershed is being used increasingly
as a benchmark for determining the health of watersheds. Since Klein published one of
the earliest papers on this topic (Klein, 1979), numerous authors have presented their
views. A modified summary of these findings is presented in Table I.5.1 (Schueler, 1994).
The Center for Watershed Protection (CWP) has greatly popularized the concept of a direct
relationship between watershed imperviousness and stream health and has reported that
stream health impacts tend to begin in watersheds with only 10 to 20% imperviousness
(the so-called 10% rule) threshold. Their theory indicates that sensitive streams can exist
relatively unaffected by urban stormwater with good levels of stream quality where
impervious cover is less than 10%, although some sensitive streams have been observed
to experience water quality impacts at as low as 5% imperviousness, as shown in the
study of watershed determinates (Horner et al., 1996). Impacted streams are reported to
be threatened and exhibit physical habitat changes (erosion and channel widening) and
decreasing water quality where impervious cover is in the range of 10 to 25%. The
threshold theory categorizes streams in watersheds where the impervious cover exceeds
25% as typically degraded, having a low level of stream quality and an inability to support
a rich aquatic community.
Some researchers and practitioners in urban water resources have questioned this
imperviousness threshold theory. A recent study of the relationship between subbasin TIA
and the Qualitative Habitat Index (Horner and May, 1999) raises a number of issues. The
studies show that, for 31 basins with less than 5% TIA, 14 ranked as excellent, 13 ranked
as good, while 4 ranked as fair. No explanation is provided for this variance. Also, for a
group of 45 basins with TIA values greater than 25%, 9 were ranked as good habitat
quality, 13 were ranked as fair, and 16 were ranked as poor. Obviously, factors in addition
to TIA are affecting the ranking of these watersheds because the impervious threshold
theory would have categorized all of the watersheds as typically degraded and having a
low level of stream quality.
From a statistical standpoint, it can be pointed out that other things being equal, with
a large data set, there should be some overall relationship between impervious cover
measures and stream quality measures. This type of relationship would also hold for a
number of parameters, including population density, but one might expect that these
should be noisy, messy relationships as indicated in the Horner/May study, and that a
significant threshold found in one data set from one setting is unlikely to hold for other
settings. If one starts to consider the complexity of controls on stream quality (measured
various ways), and the ways in which urbanization impacts these, it is clear that impervious cover on its own is but one of a large number of controls. Figure I.5.1 shows the
wide number of factors that can impact the ecological integrity of a receiving stream.
Many of these factors are not dependent on the degree of impervious cover.
The pros of using impervious cover as an indicator are that it is simple to measure,
planners can work with it, and it makes intuitive sense to a wide range of people. The
con is that it is a rather blunt instrument but one that does not do justice to the complex
physical, chemical, and biological processes we must be concerned with if we are to restore
the ecological integrity of watersheds.
Chapter I.5:
Table I.5.1 Summary of Key Findings of Urban Stream Studies Examining the Relationship
of Urbanization on Stream Quality
Ref. (Year)
1. Klein (1979)
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Location
Maryland
Biological
Parameter
Key Finding
Aquatic
insects/fish
Aquatic
insects
Salmon
Macroinvertebrate and fish diversity declines

rapidly after 10% TIA
Benke et al. (1981) Atlanta
Negative relationship between number of
insect species and urbanization in 21 streams
Steward (1983)
Seattle
Marked reduction in coho salmon populations
noted at 1015% TIA at 9 sites
Garie and
New Jersey Aquatic
Drop in insect taxa from 13 to 4 noted in urban
McIntosh (1986)
insects
streams
Pedersen et al.
Seattle
Aquatic
Macroinvertebrate community shifted to
(1986)
insects
chironomid, oligochaetes, and amphipod
species tolerant of unstable conditions
Jones and Clark
Northern
Aquatic
Urban streams had sharply lower diversity of
(1987)
Virginia
insects
aquatic insects when human population
density exceeded 4 persons/acre (estimated
1525% TIA)
Steedman (1988) Ontario
Aquatic
Strong negative relationship between biotic
insects
integrity and increasing urban land
use/riparian condition at 209 stream sites.
Degradation begins at about 10% TIA.
Limburg and
New York Fish spawning Resident and anadromous fish eggs and larvae
Schmidt (1990)
declined sharply in 16 tributary streams
greater than 10% impervious
Booth (1991)
Seattle
Fish habitat/ Channel stability and fish habitat quality
channel
declined rapidly after 10% TIA
stability
Yoder (1991)
Ohio
Aquatic
100% of 40 urban sites sampled had fair to very
insects/fish
poor index of biotic integrity scores
Schueler and Galli Maryland Fish/aquatic
Fish diversity declined sharply with
(1992)
insects
increasing TIA. Loss in diversity began at
1012% TIA; insect diversity metrics in 24
subwatersheds shifted from good to poor
over 15% TIA
Luchetti and
Seattle
Fish
Marked shift from less tolerant Coho salmon
Fuerstenberg
to more tolerant cutthroat trout populations
(1993)
noted at 1015% TIA
Taylor (1993)
Seattle
Wetland
Mean annual water fluctuation was inversely
plants,
correlated to plant and amphibian density in
amphibians
urban wetlands; sharp declines noted over
10% TIA
Galli (1994)
Maryland Brown trout
Abundance and recruitment of brown trout
declines sharply at 1015% TIA
Insect diversity at 19 stream sites dropped
Shaver et al.
Delaware Aquatic
sharply at 815% TIA; strong relationship
(1994)
insects/
between insect diversity and habitat quality;
habitat
majority of 53 urban streams had poor habitat
quality
Black and Veatch Maryland Fish/insects
Fish, insect, and habitat scores were all ranked
(1994)
as poor in 5 subwatersheds that were greater
than 30% TIA
Source: Modified from Schueler, T.R., The importance of imperviousness, Center for Watershed Protection, Watershed Prot. Tech., 1, 100, 1994.
106
How well do we maintain the ecological integrity

(functions) of aquatic systems (small streams)?
Scale / Spatial / Temporal / Species
Nutrients
Temperature
D.O.
pH
Turbidity
Organics
Toxics
Chemical
Flow
Variables
Regime
Ecosystem
Integrity
Biotic
Habitat
Structure
Factors
Disease
Reproduction
Feeding
Predation
Competition
Energy
Sources
Sunlight
Nutrients
Seasonal Cycles
Organic Matter
1&2 Production
Velocity
Frequency
Run off
Evaporation
Ground Water
Flow Duration
Rain Intensity
Canopy
Siltation
Gradient
Substrate
Current
Instream Cover
Sinuosity
Width/Depth
Channel
Morphology Soils
Stability
Riparian Vegetation
Figure I.5.1 Factors that impact the ecological integrity of a receiving stream.
Furthermore, such a simplistic view of cause and effect of urban impacts does little
to help our understanding of the relationship between ecological processes and technology
necessary to maintain these processes. Promoters of the impervious threshold theory
assume that mitigation of impacts will maintain an acceptable hydrologic regime and
maintain the ecological integrity of the receiving waters. Many are not comfortable with
this assumption because there is little data to prove that an altered hydrologic regime will
maintain a natural homeostasis state to ensure the long-term sustainable vitality of the
ecosystem. With LID the goal is to maintain the natural hydrologic regime and its homeostatic state. Intuitively, if the natural hydrologic regime can be maintained, the long-term
sustainability of the ecosystem can be assured.
The TIA threshold theory has some appropriate uses as a very rough watershed
planning tool, but it would be hard to defend in detail in either a scientific or legal setting
(e.g., if one were to write an ordinance setting a limit on imperviousness based solely on
TIA, a good set of experts could easily get it overturned).
Alternative approaches to the TIA threshold theory can be obtained by identifying
the individual parameters, or metrics, that define stream health and better examine the
relationship between TIA and these metrics. Some of these watershed impairment metrics
include the following:
Runoff volume
Bankfull and subbankfull discharge
Change in bankfull hydraulic geometry
Channel modifications
Upstream channel erosion and sediment loads
Decrease in base flow
Decrease in wetted perimeter (width/depth ratio)
Instream habitat structure
Chapter I.5:
Large woody debris (LWD)

Stream crossings and fish barriers
Riparian forest fragmentation, width, and diversity
Chemical water quality
Stream temperature
Aquatic diversity
Many practitioners have viewed these parameters or metrics on an individual basis

and have thus developed a series of fragmented techniques, which for the most part have
had limited success or failed. For examples and more details of the various metrics that
others have examined to relate urbanization to stream degradation, refer to references
(e.g., Alley and Veenhuis, 1983; Barton et al., 1985; Beyerlein, 1996; Booth, 1990, 1991, 1993,
1996; Galli, 1991; Graf, 1977; Gregory et al., 1991; Horner et al., 1996; Leopold, 1968; Macrae
and Marsalek, 1992; Wigmosta et al., 1994). The impacts associated with urban land-use
change, however, can and should be viewed from a hydrologic cycle perspective (i.e.,
hydrologic functions or regime). A cause-and-effect relationship can be established for the
full range of impact metrics, and logical, process-related solutions can be developed.
With LID, reducing the amount of impervious surfaces is only one of many impact
reduction techniques. LIDs goal is the reestablishment of the natural hydrologic regime
and ecosystem functional relationships, so the amount of impervious surface is not as
important as how the runoff is managed and functionally restored to the site. LID provides
numerous techniques and strategies to manage and reestablish a functional equivalent of
the natural hydrology regime independent of the amount of impervious surfaces.
Another troubling implication of the impervious cover threshold theory involves the
large expanses of urban areas that have already been densely developed. The impervious
threshold theory suggests that it is impossible to restore stream quality in watersheds with
high impervious cover; however, a growing number of stream restoration projects
throughout Maryland use innovative concepts and techniques based on the application
of fluvial geomorphology. These projects suggest that restoration is possible, particularly
if adequate stream buffers or riparian zones exist. Several additional projects in Maryland
offer promise that, even under adverse conditions, applying sound fundamental analysis
of hydrologic functions and processes can restore urban watersheds. These include urban
retrofit demonstration projects such as the Bladensburg Port Towns project in Prince
Georges County and targeted watershed projects such as Sawmill Creek in Anne Arundel
County (Arnold et al., 1982).
The principles and practices of LID also allow for the retrofit of existing developed
areas by integrating microscale management into the urban landscape. Reducing TIA in
existing developed areas alone is neither practical nor effective in retrofitting existing
urban areas to reestablish ecological functions or to achieve the fishable and swimmable
objectives.
Impacts of urbanization
In general, the degradation process to receiving waters caused by urbanization has the
following pattern. Land-use and cover changes, compacted soils, and creation of efficient
drainage/conveyance systems using connected impervious surfaces lead to very significant changes in the watersheds hydrologic cycle or regime. These changes consist primarily of a reduction or loss in the initial rainfall abstractions, Ia. These processes include
water intercepted by vegetation, water retained in surface depressions, evaporation, and
infiltration. The reduction or loss of these initial abstractions, together with accompanying
108
decreases in the time of concentration, Tc (both sheet flow and overland flow), leads to
significant increases in both the runoff volume and peak discharge values. These basic
hydrologic principles are explained in the Department of Agriculture, Soil Conservation
Services technical release series (SCS, 1986).
The receiving streams respond to these hydrologic regime changes by increasing their
cross-sectional areas, usually through a combination of channel down-cutting and bank
erosion that produces staggering volumes of sediment and results in the destruction and
loss of stream habitat and the related reduction and loss of biologic species. In addition,
the surface runoff flowing over the impervious surfaces typically displays elevated pollutant levels and temperatures that can be very damaging to most fish species. The
reduction or loss of the initial abstractions can reduce groundwater recharge, lower water
tables, and reduce or cease baseflow to small streams, particularly during dry periods. A
watershed hydrologic cycle is changed by the way we design and construct sites and
choose to design our drainage systems, that creates an efficient drainage system that
completely alters the natural hydrologic regime.
The key to addressing and controlling these impacts must focus on controlling or at least
minimizing the changes to the hydrologic cycle or regime. Simply relying on TIA reduction
and conventional detention is not feasible, practical, or sustainable. The challenge of LID
technology is to develop a comprehensive toolbox of techniques that allows the recreation
of the initial abstraction volumes and frequency of discharges to mimic natural hydrologic
processes and thus preclude the traditional impairment associated with urbanization.
The LID approach

Currently, we design and construct every site with one basic overriding goal to achieve
good drainage. In other words, it is important to get runoff off the site as quickly as
possible to the conveyance system and centralized BMP treatment device. As a site is
developed, its hydrologic functions are first altered on a microscale to create a highly
efficient drainage system. The cumulative impacts of these microscale changes results in
drastically altered hydrologic regimes that we typically try to mitigate using end-of-pipe
management practices.
If we can design sites to achieve good drainage, why not design sites with the opposite
objective to maintain predevelopment hydrologic functions? Can we intelligently engineer sites to replace the microscale hydrologic functions, and would the cumulative
beneficial effects result in the preservation of natural watershed hydrologic functions? Can
a site be designed in a way to remain a functional part of an ecosystems hydrological
regime or at least more closely mimic natural hydrologic functions? To create a hydrologically functional site, there must be a radical change in our thinking. We need to restore
hydrologic functions, not just mitigate development impacts.
Five basic steps for LID designs

The five basic steps of the LID approach are discussed in the following paragraphs. These
steps follow a systematic approach to site planning and design. The LID techniques are
not necessarily new, but the principle of combining all these techniques in a manner that
produces a comprehensive approach of distributed management is new. What is also new
is that the LID guidance manuals provide an analytical methodology based on TR-55 (SCS,
1986), which allows one to quantify the hydrologic impacts of the combined affects of all
the LID practices.
The objective of LID site design is to manage, recharge, detain, and retain runoff
volumes uniformly throughout the site to mimic predevelopment hydrologic functions.
Chapter I.5:
Uniform distribution of small on-lot retention and detention to control both runoff discharge volume and rate is the key to better replication of predevelopment hydrology. The
relative change in the frequency and duration of runoff is also much closer to predevelopment conditions than can be achieved by typical application of conventional centralized
BMPs such as ponds. Management of both runoff volume and peak runoff rate are both
included in the design of controls. This is in contrast to conventional end-of-pipe treatment
that completely alters the watershed hydrology to create a new modified hydrologic regime.
The LID site analytical analysis and design approach focuses on four major hydrologic
elements. These fundamental factors affect site hydrology and are introduced in the
following list. For a more detailed explanation, refer to the LID manuals and the unpublished paper explaining LID hydrology (Cheng et al., 2001).
Curve number (CN) A factor that accounts for the effects of soils and land cover
on the amount of runoff generated. Minimizing the magnitude of change from the
predevelopment to the post-development CN by reducing impervious areas and
preserving more natural vegetation will reduce runoff storage requirements and
help to maintain predevelopment runoff volumes.
Time of concentration (Tc) The time it takes runoff to travel through the watershed.
Maintaining the predevelopment Tc reduces peak runoff rates and can be achieved
by lengthening flow paths, reducing the use of pipes and paved channels, and
conserving natural drainage and depression storage.
Permanent storage areas (retention) Retention storage is needed for volume and
peak control, as well as water quality control and to maintain the same CN as the
predevelopment condition.
Temporary storage areas (detention) Detention storage may be needed to maintain
the peak runoff rate or to prevent flooding.
Step 1. Conservation. The first step of LID is to minimize or prevent runoff to reduce
the change in the CN. This step is similar to traditional techniques of maximizing natural
resource conservation, limiting disturbance, and restoring impacted natural resources.
This includes considering conservation requirements and watershed plan components
such as parks, open space, streams, step slopes, and permeable soils. These are the typical
conservation techniques that help to define the buildable area of the site. Conservation
techniques also include maintaining natural drainage patterns, topography, and depressions; preserving as much existing vegetation as possible in pervious soils, specifically
hydrologic soil groups A and B; and revegetating cleared and graded areas. These measures help to minimize the change in the pre- and post-development CN; see Figure I.5.2
for basic conservation strategies for site planning and design. Chapter I.3 also discusses
conservation techniques and benefits.
Step 2. Minimization. Calculation of the LID CN is based on a detailed evaluation
of the existing and proposed land cover so that an accurate representation of the potential
for runoff can be obtained. This calculation requires the engineer/planner to investigate
the following key parameters associated with LID: (1) land cover type; (2) percentage of
and connectivity of impervious cover; (3) hydrologic soils group (HSG); (4) hydrologic
conditions (average moisture or runoff conditions; and (5) maintaining existing drainage
patterns and natural retention features. Reducing the change in CN alone will reduce both
the post-development peak discharge rate and volume.
The following are examples of LID site planing practices that can be utilized to achieve
a substantial reduction in the change of the calculated CN: narrower driveways and roads
110
Figure I.5.2 Conservation strategies for site planning.
Figure I.5.3 Example of limited lot disturbance (site finger printing).
(minimizing impervious areas), site finger printing (minimal disturbance), open drainage
swales, preservation of soils with high infiltration rates, location of BMPs on high-infiltration soils, disconnecting impervious surfaces to direct and disburse runoff to soil groups
A and B, flattening slopes within cleared areas to facilitate on-lot storage and infiltration,
and construction of impervious features on soils with low-infiltration rates. See Figure I.5.3
for an example of limited lot disturbance or site finger printing. Chapter I.3 also discusses
minimization techniques.
Step 3. Maintaining the predevelopment time of concentration (Tc). The LID hydrologic
evaluation requires that the post-development Tc be close to the predevelopment Tc. This
is important because LID is based on a homogenous land cover and distributed retention
Chapter I.5:
Figure I.5.4 Example of a vegetative swale.
and detention of on-site BMPs. The following site planning techniques can be used to
maintain the existing Tc:
1. Maintain predevelopment flow path length by dispersing and redirecting flows
using open swales and natural or vegetated drainage patterns.
2. Increase surface roughness (e.g., preserving woodlands, vegetated swales).
3. Detain flows (e.g., open swales, bioretention).
4. Minimize disturbances (minimizing compaction and changes to existing vegetation).
5. Flatten grades in impacted areas.
6. Disconnect impervious areas (e.g., eliminating curb/gutter and redirecting down
spouts).
7. Connect pervious areas to vegetated areas.
See Figure I.5.4 for an example of a vegetated swale to slow down runoff to allow
more infiltration and evaporation.
Combined use of these techniques, and those to reduce the change in the CN, can
modify runoff characteristics to effectively shift the post-development peak runoff time
toward that of the predevelopment condition.
Step 4. Maintaining the redevelopment curve number and runoff volume. Once the postdevelopment Tc is maintained at the predevelopment conditions and the change of predevelopment to post-development CN is minimized, any additional reductions in runoff volume
must be accomplished through distributed on-site stormwater integrated management practices, or IMPs. The goal is to select the appropriate combination of IMPs that simulate the
hydrologic functions of the predevelopment condition to maintain existing CN and corresponding runoff volume. LID design strives to maximize the runoff use and retention practices distributed throughout the site to provide the required volume controls at the source.
Retention storage allows for a reduction in the post-development volume and the
peak runoff rate. The increased storage and infiltration capacity of retention BMPs allow
the predevelopment volume to be maintained. The most appropriate on-lot retention BMPs
include (1) bioretention cells (rain gardens), (2) infiltration trenches, and (3) rain barrels.
Other possible retention BMPs include retention ponds, rooftop storage, cisterns, and
irrigation ponds. It may be more difficult to distribute these types of controls throughout
a development site, but often they are part of the site drainage patterns and have only to
be left untouched. Chapters I.6 and I.7 discuss other techniques in greater detail.
112
Storage Detention & Filtration

LIDs IMPs
Uniform Distribution at the Source

Open drainage swales
Rain gardens /bioretention
Smaller pipes and culverts
Small inlets
Depression storage
Infiltration
Rooftop storage
Pipe storage
Street storage
Rain water use
Soil management
Figure I.5.5 Examples of integrated lot-level management practices.
As retention storage volume is increased, a corresponding decrease occurs in the peak

runoff rate in addition to runoff volume reduction. If a sufficient amount of runoff is
stored, the peak runoff rate may be reduced to a level at or below the predevelopment
runoff rate. This storage may be all that is necessary to control the peak runoff rate when
there is a small change in CN. When there is a large change in CN, however, it may be
less practical to achieve flow control using volume control only. Figure I.5.5 lists examples
of integrated practices that can be used to treat, filter, retain, and detain runoff at the lot
level.
Step 5. Pollution prevention. Pollution prevention and maintenance of on-lot BMPs
are two key elements in the overall LID comprehensive approach. Effective pollution
prevention measures can reduce the introduction of pollutants to LID BMPs, thereby
enhancing their ability to reduce pollutant levels and extend the life of the facilities. Public
education is essential to successful pollution prevention and BMP maintenance. Not only
will effective public education complement and enhance BMP effectiveness, it can also be
used as a marketing tool to attract environmentally conscious buyers, promote citizen
stewardship, awareness, and participation in environmental protection programs, as well
as help to build a greater sense of community based on common environmental objectives
and the unique environmental character of LID designs.
Education is the key to effective public participation. With LID techniques all stakeholders (public officials, engineers, builders, realtors, and buyers) must be educated about
the positive environmental impacts of LID and its maintenance savings and burdens. With
LID, most controls are integrated into the lot landscape features; therefore, property
owners will need to know how to maintain these features. This can be achieved through
easements, covenants, brochures, and environmental committees. Although additional
landscape means more landscape maintenance, there is no major stormwater infrastructure
(ponds and pipes and structures) to maintain, and the scale of the maintenance is reduced
to what an individual property owner can afford for routine landscape maintenance costs.
Property owners have responded to LIDs landscape level of control in two very
important ways. First, they feel good about their propertys helping to protect the environment. Second, they believe that the additional landscape material adds greater value
to their property. Thus, LID controls provide a strong economic incentive to maintain LID
practices because property values are also maintained.
Chapter I.5:
Figure I.5.6 Pollution prevention strategies.
From 1992 through 1997 PGCDER conducted a public education demonstration project
in an 1,150-unit residential area. Monitoring and modeling results showed that it is possible
to achieve a 30 to 40% reduction in the use of fertilizer through a very aggressive outreach
program. The study further indicated that even more gains could be made if manufacturers
provide instructions on their products, suggesting how to limit the use of the fertilizers.
Figure I.5.6 summarizes some of the findings of the Kettering study (Coffman et al., 1998).
Other important LID considerations

Potential requirement for additional detention storage
In cases where very large changes in CN cannot be avoided, retention storage practices
alone may either be insufficient to maintain the predevelopment runoff volume or peak
discharge rates or require too much space to represent a viable solution. In these cases,
additional detention storage will be needed to maintain the predevelopment peak runoff
rates. A number of traditional detention storage techniques are available that can be
integrated into the site planning and design process for a LID site. These techniques
include: (1) swales with check dams, restricted drainage pipes, and inlet/entrance controls;
(2) wider lower gradient swales; (3) rain barrels; (4) rooftop storage; (5) shallow parking
lot storage; and (6) constructed wetlands and ponds. These detention practices can easily
be integrated into the site design features.
Where downstream flooding is a problem, additional flow control may be necessary
to protect property and ensure safe conveyance. Also, when there is a need to retrofit
existing development, additional detention may be needed to control off-site flows using
regional ponds.
Determination of design storm event

The hydrologic approach of LID is to retain the same amount of rainfall within the
development site as that retained prior to any development (e.g., woods or meadow in
114
LID Site
Conservation
Porous
Pavement
Lot Level
Source Controls
Narrower
Streets
Amended
Soils
Open
Drainage
Rain
Barrel
Rain Gardens
Create a Hydrologically
Functional Lot
Figure I.5.7 Example of an LID residential lot design.
good condition) and then release excess runoff as the woods or meadow would have. By
doing so, it is possible to mimic, to the greatest extent practical, the predevelopment
hydrologic regime to maximize protection to aquatic ecosystems and groundwater
recharge. This approach allows the determination of a design storm volume that is tailored
to the unique soils, vegetation, and topographic characteristics of the developing watershed. This approach is particularly important in watersheds that are critical for groundwater recharge to protect stream/wetland base flow and ground- or surface water supplies. For each watershed, there is a unique amount of runoff that must be retained to
mimic the natural conditions. With LID, the volume of runoff to be controlled changes
with each site in order to replicate the natural ecological conditions. Figure I.5.7 shows
residential LID site design with a number of integrated management practices. Figure I.5.8
shows complete LID residential site design using a variety of techniques.
Integrated management practices (IMPs)

Site design techniques and IMPs can be organized into three major categories, as follows:
(1) runoff prevention measures designed to minimize impacts and changes in predevelopment CN and Tc; (2) retention facilities that store runoff for infiltration, exfiltration, or
evaporation; and (3) detention facilities that temporarily store runoff and release through
a measured outlet. Table I.5.2 lists examples of only some of a wide array of LID IMPs
and their primary functions. Placing these IMPs in series (treatment train) and uniformly
dispersing them throughout the site provides the maximum benefits for hydrologic controls.
Urban Retrofit
Many highly urbanized areas are almost entirely covered by impervious surfaces such as
roadways, parking, sidewalks, and buildings. Control of runoff in urban areas using
conventional stormwater management practices is difficult and severely limited due to
the lack of open space, cost of land, high construction costs, and high operational and
maintenance costs. Typical urban stormwater controls involve construction of expensive
end-of-pipe detention facilities, infiltration systems, underground storage, systems to
Figure I.5.8 Example of an LID residential site design.
Conservation
Minimization
Soil Management
Open Drainage
Rain Gardens
Rain Barrels
Pollution
Prevention
LID
Development
Chapter I.5:
Disconnected
Decentralized
Distributed
Multi-functional
Multiple Systems
116
Table I.5.2 Examples of LID IMPs and Primary Functions
BMP
Bioretention
Infiltration trench
Dry wells
Roof top storage
Vegetative filter strips
Rain barrels
Vegetated swales and small
culverts
Swales
Infiltration swale
Reduce imperviousness
Strategic clearing/grading
Engineered landscape
Eliminate curb and gutter
Vegetative buffers
Stream
Prevent
Water
Channel
Runoff Detention Retention Conveyance Quality Protection
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
optimize in pipe storage, and the use of water quality BMPs (filters and hydraulic separators). Conventional stormwater management retrofit approaches (ultra-urban BMPs and
end-of-pipe storage) also have limitations due to costs and physical requirements
To adequately address urban runoff problems, cities must have effective low-cost and
politically acceptable tools. If dramatic improvements in urban runoff management are to
be achieved, it will be necessary to fundamentally rethink current approaches and radically
redesign and reengineer urban stormwater management technology.
LID management strategies and practices must be modified to address the unique
characteristics of each individual watershed. This makes the protection strategy even more
effective as it requires the designer, regulator, residents, and stakeholders to have a better
understanding of the complexities and water resource protection needs of each watershed.
To apply LID to any land use is simply a matter of developing numerous ways to creatively
manage, prevent, retain, detain, use, and treat runoff within multifunctional landscape
features unique to that land use. Factors that influence LID strategies and practices include
specific land use, soils, climate, rainfall distribution, water/natural resource protection
objectives, and the regulatory framework.
Initially LID practices and principles were developed to accommodate typical suburban land uses. During the initial development of the technology it was realized that LID
principles such as creating a hydrologically functional landscape, uniform distribution of
controls, strategic timing of flows, micromanagement and use of plant soil filter technology
such as bioretention have universal applicable regardless of the land use. LID can be
applied to address stormwater management goals and water resources objectives for
urban, suburban, and rural development.
In order to apply LID to highly urbanized areas, a new set of specific LID urban
management principles and management practices must be developed to address the
unique landscape features and water resources protection objectives of cities. Each city
may have different urban runoff management objectives that are based on compliance
with required local, state, and federal regulations and the economic, environmental, and
human health needs to protect receiving waters, sensitive environmental features, and
living resources. The ability of cities to meet urban runoff objectives will depend greatly
on the cost of control measures and the available economic resources. The cost of current
Chapter I.5:
state-of-the-art conventional CSO and BMP technologies can be staggering. The potential
for cost savings using LID for volume and water quality control to reduce or eliminate
these systems in urban areas is tremendous. This is because many of the techniques used
in suburban applications are extremely cost effective and can be modified for use in the
existing urban infrastructure. Also see Chapters I.1 and I.4 for more information on urban
retrofit strategies and techniques.
International microscale experiences and case studies

Conventional stormwater institutions in this country have not encouraged the use of
incorporating multiple objectives at small scales. In order to determine how effective
sustainable microscale techniques and small-scale multiple objective programs can be, we
can look toward the experiences of other countries. The following section will highlight
and explore some of the experiences that other countries have had using microscale
techniques that are similar to LID approaches for wet weather control. The information
obtained in this section is from personal communications with Neil Weinstein, Executive
Director of the Low Impact Development Center, and the findings from an unpublished
EPA report on the use of LID technology for combined sewer overflow control (Weinstein,
2000). Although many communities in the United States are rapidly accepting and planning for the use of LID and other integrated microscale techniques, a limited number of
sites have been constructed using these approaches. Examples of these developments in
the United States include Village Homes in California; Somerset and Greenbelt Plaza in
Prince Georges County, MD; Wonderland Creek in Colorado; and the NEMO project in
Jordan Cove, CT. The collection of long-term monitoring and modeling data at these sites
is also limited. Because of our limited experience with LID and other microscale approaches,
we must look to other regions in order to evaluate the effectiveness of these programs to
meet receiving water objectives. We can build on this existing experience to develop a
knowledge base of planning, design, and program management tools to help communities
evaluate and develop successful LID programs.
European and Asian countries have a significant amount of experience with microscale
approaches. Some of the critical environmental and economic factors that have been
responsible for the development and acceptance of microscale techniques in other countries include the following:
In Europe, much of the infrastructure is based on combined sewer systems. For
example, England has over 70% combined sewers. The original design capacity
cannot account for the amount of infill and increased imperviousness and the
systems are surcharging. It is difficult to construct new systems because of the
expensive conflicts with other infrastructures and disruptions to the city due to
construction. New ways to reduce runoff volumes to these systems must be developed.
The cost of land in many urbanized and core city areas in Europe and Asia makes
the sacrifice of land for developable area for conventional BMPs economically and
politically prohibitive. Techniques that provide environmental protection must be
incorporated into the infrastructure, and site designs that do have minimal impact
on the development are essential.
Land subsidence in many Asian countries and saltwater intrusion in European
countries such as The Netherlands are due to reduction of groundwater levels that
result from decreased infiltration capacity of soils and increased impervious cover
that has resulted from urbanization. Techniques for groundwater recharge in order
to maintain a sustainable water balance must be developed.
118
International examples (Weinstein, 2000)

National, local government, and watershed compacts have adopted comprehensive strategies
that incorporate microscale planning programs, demonstration projects, financing districts,
and subsidies in order to improve water quantity and water quality control using microscale
and sustainable techniques. Some key studies that have been published in the literature that
demonstrate the planning, design, and environmental effectiveness of these programs as well
as some of the institutional issues associated with microscale approaches are described next.
German watershed planning using microscale approaches

The Emscher catchment in Germany is approximately 330 square miles with 2.5 million
inhabitants residing in over 17 cities. Although traditionally a coal mining and a farming
area, over the last 100 years it has evolved into a more industrialized region. The beginning
of the transition caused a decision between economic growth and good environmental
conditions. Water resources were exploited and watercourses and soils were contaminated.
The long-term records have indicated a frequency of flooding corresponding with
increased urbanization during the recent years. In addition, over 98% of the infrastructure
utilized combined sewers and there were many open systems.
In the early 1990s, a strategy was developed in order to reduce runoff volume and
improve environmental quality. The first step was to find strategies to maintain a river
ecology that met the aesthetic and recreational needs of the citizens, although the planners
recognized the challenge of restoring the system to its predevelopment state due to the
intense urbanization. Certain precepts such as not increasing peak flows and volumes and
reconstructing straightened channels to a more natural morphology were employed. This
approach helped improve the opportunity for the reestablishment of ecological systems
and had a minor, 3 to 5%, reduction in flood frequency. The planning for the drainage
area was restructured so that subbasins were analyzed in more discrete units that enabled
the development of targeted strategies for volume and pollution reduction. The next step
was the introduction of source control. Because of the longstanding tradition of efficient
conveyance systems, this new concept was at first difficult for many communities within
the drainage district to embrace. A competition was held and grants were awarded for
communities to incorporate source controls as pilot projects. The emphasis of the project
areas was disconnection of imperviousness and incorporation of small-scale simple source
controls. In one project area of approximately 400 acres, of which 40% was impervious,
70% of the facilities were constructed by private land owners, 20% were constructed with
neighborhood development grants, and 10% were constructed by contractors.
Polls showed that environmental protection was the main reason, and reducing fees
and receiving financial incentives for construction secondary. The level of disconnection
in this area was 5%. In the seven areas studied, which included this area, the range of
disconnectivity achieved varied widely, from 5% to almost 100%. It is estimated that the result
of this program will achieve a 10% reduction in the peak of the 2-year interval storm event.
Japanese experimental sewer system

In Japan, the Experimental Sewer System (ESS) has been a highly effective method to control
runoff volume. This approach is based on the development of an integrated, highly efficient
infiltration approach in order to reduce runoff volumes and increase groundwater recharge
in urban areas. From 1983 to 1995, the Tokyo Metropolitan Government built within a
5.5-square mile area 33,300 infiltration pits, 122 acres of permeable pavement, and over 175
miles of infiltration trenches. This approach has also been adopted in many other Japanese
cities including Sapporo, Shiogama, Chiba, Yokohama, Naboya, and Amagaski. Representative areas within the ESS were analyzed in order to determine the cost-effectiveness and
Chapter I.5:
efficiency of the program. It was determined that the cost of using microscale infiltration
techniques such as permeable pavement and infiltration inlets cost approximately 33%
less than conventional open pond detention systems and 10% of the cost of storage vaults.
This alternative was so much more cost-effective due to the high cost of land and complexity of the existing infrastructure. Perhaps equally impressive was the reduction in
monitored storm drain flow volumes of up to 50%. Most important was the reduction in
CSO events from 36 to 7. Some of the keys to the success of the acceptance of these
programs have been:
Evaluation of positive effects

Development of maintenance programs
Obtaining cooperation of the public
Evaluating disadvantages
Evaluating effect on groundwater
Subsidizing private construction and maintenance
Providing administrative guidance
Improvement of Administrative Model District
Promotion of technology
Inclusion of infiltration in planning
Development of a political base
Demonstration projects have been one of the most effective ways to educate people
and gain acceptance about this approach and to gain acceptance of this. In Yokohama,
model areas were constructed to demonstrate the technology. Within a 15-acre section of
the city, 1.8 miles of infiltration pipes (typically smaller than 8 in. with a gravel base), 2.5
acres of permeable pavement, and 10 acres of conventional pavement were constructed
as part of the infrastructure. The results, including volume reduction, maintenance, and
public acceptance, are continuously being monitored.
Lyons residential infiltration strategies

The area of Lyons, France, has a history of over 50 years of developing infiltration strategies
for stormwater runoff control. Lyons is approximately 230 square miles in area with 1.2
million inhabitants. The storm sewer and sewage system infrastructure includes 1,600
miles of sewers, 56 pumping stations, and 9 treatment plants in a combined system.
Because of the cost and disruption required to build large-scale relief sewers to reduce
CSO events, alternative strategies had to be developed. Although one large-scale interceptor project was planned and built in the 1970s, it was recognized that these systems
could not continuously be built and that a reduction in flows to the interceptor was also
required. In the 1980s, a strategy of large-scale centralized infiltration ponds and pits with
pretreatment devices was incorporated as part of the control strategy. Concerns over largescale loadings on groundwater, maintenance of the systems, construction of long networks
of storm drain pipes to the end-of-pipe facilities showed the limitations of this approach.
In the early 1990s, a new decentralized approach to infiltration was proposed as part of
the water and sewer master plan. Specific approaches were recommended:
In housing areas, stormwater from private surfaces must be infiltrated.

In commercial and industrial areas, stormwater from roofs must be infiltrated.
Infiltration areas must take place as close as possible to the source.
Stormwater management must preserve the natural water cycle and not affect the
water quality.
Stormwater facilities must fit into their urban surroundings as well as possible.
120
Figure I.5.9 Roof garden.
As an example of this approach for residential development, houses are required to

have private infiltration pits, cisterns, or rain barrels to divert rainwater from roofs away
from the public drainage system. Residents are responsible for constructing and maintaining these systems. Overflow from the private systems, as well as water from public streets
and driveways, is collected in a similar system along the street before entering the collector
system. Education and maintenance programs were included in this approach. Over 2,000
private devices were inspected. The maintenance program also included the development
of a standard system for the retrofit and rehabilitation of older systems.
Aesthetic, educational, and social lessons

One of the most powerful lessons of microscale applications is the adaptability, sustainability, and multiplicity of other functions that microscale approaches have. For example,
green or eco-roofs, which have been in place for many years in Europe, and are now being
used and accepted as stormwater controls for residential areas such as Portland, OR, and
have added energy conservation benefits. The roofs have been modified to provide the
most energy efficient and water retention properties feasible. Figure I.5.9 shows an example of a roof garden.
Microscale stormwater facilities have been designed as small areas of relief and berms
in open space and courtyards, pedestrian streets with rainwater channels, fountains, and
pavement relief. The experience of other countries has shown that the opportunities, size,
function, and appearance of microscale controls are limitless. These techniques not only
constitute stormwater facilities but also have become an expansion of the aesthetic components of the urban fabric. Figure I.5.10 shows examples of rain gardens used in parking
lots, which are good examples of multifunctional, multibeneficial, and cost-effective use
of space with LID practices.
Once educated about the experience, importance, and potential of these microscale
techniques in protecting property values, meeting environmental objectives, and improving the quality of life, communities, stakeholders, engineers, and property owners will
initiate programs to develop and implement these strategies as part of their receiving
waters protection programs.
Chapter I.5:
Figure I.5.10 Examples of parking lot bioretention (rain gardens).
Costs
LID case studies and pilot programs show at least a 25 to 30% reduction in site development,
stormwater, and maintenance costs for residential development. These data were obtained
for the Prince Georges County LID case study on Patuxent Riding subdivision (PGCDER,
1998). This is achieved by reducing clearing, grading, pipes, ponds, inlets, curbs, and paving.
This, in turn, lowers construction costs, allowing builders to add greater value (features) to
the property or to be more flexible and competitive in pricing their products.
One of interesting results of LIDs on-lot microscale approach is that the stormwater
management controls become a part of each property owners landscape (natural areas,
rain gardens, open space, open swales, etc.). This reduces the public burden to maintain
large centralized management facilities and reduces the cost and scale of maintenance to
a level the homeowner can easily afford the cost of routine landscape/yard care and
pollution prevention.
Prince Georges County does not rely on enforcement to ensure maintenance of LID
landscape practices. Instead, they believe the economic incentive of maintaining property
values will ensure that most property owners will adequately maintain their LID landscape.
Roadblocks to LID
A number of roadblocks must be overcome for the successful implementation of LID.
Regulatory agencies, the development community, and the public may all have concerns
about the use of new technology. In the development of the PGCDER LID design manual,
a multiagency task force spent over two years to address all the concerns and issues. Some
of the major concerns include:
1. Develop a hydrologic analytical methodology to demonstrate the equivalence of
LID to conventional approaches.
2. Develop new road standards that allow for narrow roads, open drainage, and use
of bioretention.
122
3. Streamline the review process for innovative new LID designs, which allows for
easy modification of site, subdivision, road, and stormwater requirements.
4. Develop a public education process that informs property owners about how to
prevent pollution and maintain on-lot LID BMPs.
5. Develop legal and educational mechanisms to ensure BMP maintenance.
6. Demonstrate the marketability of green development.
7. Demonstrate the cost benefits of the LID approach.
8. Provide training for regulators, consultants, public, and political leaders.
9. Conduct research to demonstrate the effectiveness of bioretention BMPs.
10. Limited field monitoring data is available to demonstrate the effectiveness of LID
in controlling runoff quantity and quality.
Summary
LID is a viable, cost-effective alternative approach to stormwater management and the
protection of natural resources. LID is designed to provide tangible economic incentives
to a developer to save more natural areas and reduce stormwater and roadway infrastructure costs. LID can achieve greater natural conservation by using conservation as a stormwater BMP. As more natural areas are saved, less runoff is generated and stormwater
management costs are reduced. This allows multiple uses of landscape features to achieve
environmental, economic, aesthetic, and natural resources benefits.
Additionally, developers have economic incentives to provide better environmental
protection by reducing short- and long-term infrastructure costs by reducing impervious
areas and eliminating curbs/gutters and stormwater ponds to achieve LID stormwater
controls. Reduction of the infrastructure also reduces infrastructure maintenance burdens,
making LID development more economically sustainable. LID allows for the same, or in
some cases, higher lot yields compared to conventional approaches. Because stormwater
management is controlled on each lot using multifunctional landscapes, that portion of
the building area that would have been used for stormwater ponds can, in some cases,
be used for additional flood control or recovered and used for building, parking lots, open
space, or habitat enhancements.
LID promotes public awareness, education, and participation in environmental protection. As every property owners landscape functions as part of the watersheds ecosystem,
they must be educated on the benefits and the need for maintenance of the landscape and
pollution prevention measures. LID developments can be designed in a very environmentally sensitive manner to protect streams, wetlands, forests, and habitat and to save energy.
The unique environmental protection objectives of an LID development can create a
greater sense of community pride based on environmental stewardship.
In the development of the LID hydrologic analysis, NRCS/SCS TR-55 (SCS, 1986) was
used because this model is most widely used by site engineers in Prince Georges County
and throughout most of the country. During the development of LID, it was learned that
current analytical models such as TR-55 are not well suited for use with very small
watersheds. A significant amount of work is needed to upgrade current models to better
quantify the effects of microscale site design control techniques.
One extremely fascinating aspect of LID is that when controlling runoff on a microscale, there exists a whole new world of possible control practices and strategies. So get
out of the box of conventional pipe and pond technology, take up the LID challenge, and
try thinking small.
Chapter I.5:
Literature cited
Alley, W.A. and Veenhuis, J.E., Effective impervious area in urban runoff modeling, J. Hydrol. Eng.,
ASCE 109, 313, 1983.
Arnold, C.L., Boison, P.J., and Patton, P.C., Sawmill Brook: An example of rapid geomorphic change
related to urbanization, J. Geol., 90, 155, 1982.
Arnold, C.L. and Gibbons, C.J., Impervious surface coverage: The emergence of a key environmental
indicator, J. Amer. Plann. Assoc., 62, 243, 1996.
Barton, D.R., Taylor, W.D., and Biette, R.M., Dimensions of riparian buffer strips required to maintain
trout habitat in southern Ontario streams, North Amer. J. Fish. Manage., 5, 364, 1985.
Beyerlein, D., Effective impervious area: The real enemy, Proc. Impervious Surface Reduction Conf.,
City of Olympia, WA, 1996.
Booth, D.B., Stream-channel incision following drainage-basin urbanization, Water Resour. Bull., 26,
407, 1990.
Booth D.B., Urbanization and the natural drainage system Impacts, solutions, and prognosis,
Northwest Environ. J., 7, 93, 1991.
Booth, D.B. and Reinelt, L., Consequences of urbanization on aquatic systems measured effects,
degradation thresholds, and corrective strategies, Proc. Watershed 93 Conf., 1993.
Booth D.B., Stream channel geometry used to assess land-use impacts in the PNW, Watershed Prot.
Tech., 2, 345, 1996.
Cheng, M.S, Coffman, L.S., and Clar, M.L., Low Impact Development Hydrologic Analysis, Prince
Georges County, MD, 2001.
Coffman et al., Kettering Demonstration Project Final Report Public Education and Public Participation, Prince Georges County, MD, 1998.
Graf, W.L., Network characteristics in suburbanizing streams, Water Resour. Res., 13, 459, 1977.
Galli, J. 1991. Thermal impacts associated with urbanization and stormwater management best
management practices, Metropolitan Washington Council of Governments, MD Dept. Environ., Washington, DC, 1991.
Gregory, S.V., Swanson, F.J., McKee, W.A., and Cummins, K.W., An ecosystem perspective of riparian
zones: Focus on links between land and water, Biosci., 41, 540, 1991.
Horner, R.R., Booth, D.B., Azous, A.A., and May, C.W., Watershed determinants of ecosystem
functioning, Proc. ASCE Conf., Snowbird, UT, 1996.
Horner, R.R. and May, C.W., Regional Study Supports Natural Land Cover Protection as Leading Best
Management Practice for Maintaining Stream Ecological Integrity, Univ. Washington, Seattle,
WA, 1999a.
Horner, R.R. and May, C.W., Watershed Urbanization and the Decline of Salmon in Puget Sound Streams,
Univ. Washington, Seattle, WA, 1999b.
Klein, R.D., Urbanization and stream quality impairment, Water Resour. Bull., 15, 948, 1979.
Leopold, L.B., The Hydrologic Effects of Urban Land Use: Hydrology for Urban Land Planning A
Guidebook of the Hydrologic Effects of Urban Land Use, USGS Circular 554, 1968.
Macrae, C. and Marsalek, J., The role of stormwater in sustainable urban development, Proc. Canad.
Hydrol. Symp., Winnipeg, Canada, 1992.
National Park Service, Ecosystem Management in the National Park Service, U.S. Dept. Interior, Sept. 1994.
Prince Georges County, MD, Dept. Environ. Resour., Low-Impact Development Design Manual, PGC, 1997.
Prince Georges County, MD, Patuxent Riding Low Impact Development Case Study, 1998.
Prince Georges County, MD, Dept. Environ. Resour., Low-Impact Development Design Strategies: An
Integrated Approach, Jan. 1999.
Schueler, T.R., The importance of imperviousness, Center for Watershed Protection, Watershed Prot.
Tech., 1, 100, 1994.
SCS, Urban Hydrology for Small Watersheds, Technical Release 55, U.S. Dept. Agric., Soil Conservation
Service, Eng. Div., Washington, DC, 1986.
Weinstein, N.K.P., Low Impact Development for Combined Sewer Overflow Control, U.S. EPA, 2000.
Wigmosta, M.S., Burgess, S.J., and Meena, J.M., 1994. Modeling and monitoring to predict spatial
and temporal hydrologic characteristics in small catchments, USGS Water Resour. Tech. Rep.
no. 137, 1994.
124
Response
Thinking big, acting small:
Multi-tasking and the benefits of dispersed micromanagement
Prince Georges County, MD, is the originator and leading implementer of low-impact
development (LID). As Larry Coffman outlines in this chapter, LID is based on the simple
strategy of nature knows best. In other words, the best we can do is to try to imitate
nature as closely as possible when undertaking water sensitive designs. At its core, LID
is a system of land development principles and practices based on creating a more hydrologically functional urban landscape to protect water bodies. In terms of scale, LID represents an important shift away from end-of-pipe, centralized, and costly solutions to
stormwater management toward decentralized micromanagement at the scale of individual house lots or parking lots.
Coffman identifies a whole suite of conservation strategies, impervious minimization
measures, phased timing techniques, and integrated small-scale BMPs that will accomplish
these objectives. In terms of design, the goal of LID is to create as many small-scale
retention, detention, pollution prevention, and treatment technologies as possible for any
site. Also discussed in this chapter is the systematic stepped approach from planning to
final design in terms of implementing LID techniques. Another benefit from undertaking
such water sensitive management is a reduction in costs compared to conventional, highly
engineered technologies. Similar findings are predicted by Jennifer Zielinski in Chapter I.3
and shown by Michael Clar in Chapter I.15. Because LID operates at such small spatial
scales, it is very amenable for application to water sensitive retrofits made to aging
infrastructure in dense urban settings, as proposed by Richard Pinkham and Timothy
Collins in Chapter I.4.
Coffman stresses that, though site-specific in design, LID approaches must be placed in
the broader spectrum of comprehensive watershed planning. A very important point that
Coffman raises is the controversial nature of LID, representing as it does, a way of turning
traditional water management on its head. At least for the near future, until a more general
acceptance of the procedures becomes established, water sensitive designers proposing such
techniques must recognize that they are challenging the existing paradigm (for example, see
the case study described by Robert France and Philip Craul in Chapter I.7). On a broader
scale, it is possible that building codes may have to be reexamined before LID can be
applied. This underscores an important point raised in this chapter the need for client
education.
In the end, as Coffman informs, LID offers microscale approaches to water sensitive
design that are adaptable and sustainable and have multiple functions in addition to
stormwater management. He even goes as far as to suggest that with the widespread
adoption of LID methods, the very term BMPs, or best management practices, might
be better replaced by IMPs, or integrated management practices.
chapter I.6
Water gardens as stormwater

infrastructure (Portland, Oregon)
Thomas Liptan and Robert K. Murase
Abstract
Water gardens are aspects of the built environment that emulate natures processes. Water
is not a feature unto itself, but an integral element of the site and architecture. Water
gardens are a synergetic result of landscape, biology, architecture, and engineering. The
principle of integrating water in urban design introduces water as a friendly companion,
but always with attention to its potential power and negative aspects. This symbiotic
design principle includes soil and vegetation within the urban hardscape. Trees, plants,
and soil are employed to function with water in urban spaces previously not used for
stormwater management. We might call this a new paradigm, urban-nature, not at the
expense of our human habitat . . . but as enhancement of this habitat . . . earth, water,
plants . . . all have an artful place with people in the urban context. Documentation of
tests, monitoring data, costs, and observations show the viability of the techniques discussed. Application of this new paradigm will result in lower development costs, more
cost-effective retrofits, improved livability, and a more natural urban environment.
Introduction
Urban development provides the essentials of human community life. Usually within this
human community exists some degree of nature. These human essentials are, to a large
extent, forms of impervious surfaces and pipes. Stormwater runoff, the physical phase of
precipitation after it falls in the urban community, is almost foreign to the natural environment. The cause of this runoff comes from the impervious surfaces needed within the
urban community. To reverse, mitigate, or eliminate the negative effects of stormwater
runoff, new ways of designing or retrodesigning the urban community are being explored
and tested. Many of these new ideas are actually modern applications of age-old
approaches, which for some reason had faded away. Maybe the resurgent interest in
bringing aspects of nature back into the community is caused by some of the federal laws
governing water, air, and threatened species. Certainly within the Pacific Northwest, water
and nature are synonymous with salmon and forests. It is also becoming apparent that
the building blocks of nature are also the building blocks of a healthy urban community.
These new ideas begin to take shape in the form of urban design techniques, which are
1-56670-562-2/02/$0.00+$1.50
125
126
methods of integrating water with land and vegetation. Water gardens, or perhaps better
stated as eco-gardens, are an aesthetic and cost-effective approach to design with water.
In Portland ideas that regreen or mitigate the effects of impervious surfaces are being
developed. These approaches include development of a healthy urban forest, revegetation
and preservation of riparian corridors and habitat, identification and removal of unnecessary impervious surfaces, improved street designs to reduce environmental impacts,
improved zoning codes to reduce hard surfaces, and identification and implementation
of green/sustainable building and site design practices. The physical characteristics of
impervious surfaces are essentially rooftops and pavement. These surfaces are at best
environmental dead zones but are certainly not benign, because they have other than
direct water impacts, such as contributing to urban heat island conditions, smog, loss of
wildlife and habitat, increased carbon dioxide, and reduced oxygen and photosynthesis.
So what paradigm best describes these new ideas or approaches? Simply stated, it is
the careful integration of water with site and architectural design. The applications of
design elements allow the urban hydrology to better mimic nature. The design elements
are those usually within the landscape architects purview, including soil, plants/trees,
rock, and wood with stormwater added.
This chapter introduces scientific information obtained either from direct testing or
by recently published technical literature. Next, design techniques are presented with
reference to their functions and benefits. Projects that have employed one or more of these
techniques are presented. Each demonstration project offers built examples. It is expected
that these techniques will improve with each new application. Projects are discussed with
a description of the site, techniques employed, and commentary on what works and what
does not.
Studies: Pacific Northwest precipitation and hydrology

Portland rainfall analysis
Most precipitation in the Pacific Northwest occurs in the form of small storms. Using U.S.
Weather Service data, an analysis of 24-hr precipitation events was conducted. Storms
were divided into sizes from 0 to 0.2 in., 0.2 to 0.4 in., 0.4 to 0.6 in., and so on, up to and
including the largest storms during the past 16 years. To simplify the analysis at the risk
of accuracy, a 24-h event was based on calendar-day rain totals. Figure I.6.1 indicates that,
on average, 145 days per year had storm events, or calendar days of rain, of 1.0 in. or less.
These storm events accounted for 81% of the average annual rainfall. This characteristic
of precipitation is an important part of understanding how to better manage urban runoff
and pollutants associated with these numerous events. Limitations of this study include
the lack of distinction between calendar days, i.e., 0.2 in. could fall at midnight and another
0.8 in. could fall in the hours immediately following on the next day. Albeit simple, it is
not encumbered by the debate of what antecedent dry period should be used to define
the beginning and end of a storm event. Even so, the information is relatively consistent
with other studies, such as Chapter I.1.
Comparison of predevelopment forests versus post-development peak flows

and volumes
An analysis of rainfall distribution was conducted to determine the pre- and post-runoff
from storm events using 0.1 in. increments, starting with 0.1 in./24 h events and continuing every additional 0.1 in. up to 2.4 in. (the Portland 2-year event). The Santa Barbara
Urban Hydrograph (SBUH) method was employed for these calculations. Here again, this
Chapter I.6:
Water gardens as stormwater infrastructure (Portland, Oregon)
127
120
113
105
Average # Days
90
75
60
45
30
20
10
15
0.25 0.5 0.5 0.75 0.15
0
0.2 0.4 0.6 0.8
1.2 1.4 1.6 1.8
2.2 2.4
Rainfall - inches
Figure I.6.1 Annual rainfall 19831998. Average event frequencies. (Source: BES Hydradata.)
Peak flow (runoff cfs)
2.5
2
1.5
Post-Dev
3 ac. Impervious
2 ac. forest
1
0.5
Pre-Dev
5 ac. forest
0
0
0.5
1
1.5
Rainfall - inches, 24 hr event
2.5
Figure I.6.2 Pre-/post-development comparison. (Source: SBUH method, perv. Cn 70/imp. Cn 98.)
may not be the most accurate method to use, but the city has currently approved it for
sizing storm facilities. A hypothetical site of 5 acres was used, with predevelopment
assumed to be a forest condition with a curve number CN70. Post-development was 3
acres of impervious surface with a curve number CN98, and the other two acres assumed
to be left undisturbed but hydrologically connected to the new storm system. Results for
the predevelopment condition indicate that no runoff occurs for storms of 0.0 to 0.8-in.
24-h events. Figure I.6.2 shows that predevelopment peak flows do not begin until at least
a 0.9-in. event occurs and then have peak flow rates of up to 0.14 cfs (cubic feet per second)
for the 2-year, 24-h event. For post-development conditions, runoff occurs with the first
0.1-in. event. For 0.3-in. events, the post-peak flow is 0.16 cfs and then increases to more
than 10 times the predevelopment peak for the 2-year event. Figure I.6.3 shows a comparison of runoff volumes, with predevelopment at zero cf. (cubic feet) for all storms up to
0.8 in. and 7,396 cf. for the 2-year event. Post-development discharge volumes were 383
cf. for the 0.1-in. event and 27,810 cf. for the 2-year event. In the predevelopment condition,
little, if any, runoff occurs for storms of less than 1.0 in. In the post-development condition,
runoff occurs with each event of 0.1 in. or more.
Bolton and Watts (1998) state, Very little precipitation ends up as overland flow in a
mature, undisturbed forest. At the same conference, Beyerlein and Brascher (1998) state
that in the Puget Sound area with annual rainfall at 40.7 in., 18.8 in. evapotranspirates and
128

30000
Volume runoff (cf)
25000
20000
Post-Dev
3 ac. Impervious
2 ac. forest
15000
10000
Pre-Dev
5 ac. forest
5000
0
0
0.4
0.8
1.2
1.6
2.4
Rainfall - inches, 24 hr event
Figure I.6.3 Pre-/post-development comparison. (Source: SBUH method, perv. Cn 70/imp. Cn 98.)
only 0.1 in. precipitation becomes surface runoff. Based on the above SBUH analysis, prior
to development, no surface runoff occurs in a conifer forest for over 100 storm events per
year, on average. This would also indicate that the energy of water within streams is only
affected by surface runoff occasionally. In addition, the pollutants carried into streams
from surface flows would also occur occasionally. Surface runoff for post-development
conditions occurs more than 100 times a year and carries not only naturally occurring
pollutants, but also human-generated pollutants. Energy from the storms is also conveyed
to the stream with peak flows equal to the 2-year event occurring with >0.3 in. events.
Basically, theres a whole-lotta runoff and pollutants that did not previously occur, and
it is happening almost every time it rains. Note: conventional detention methods (vaults
and ponds) would reduce only the energy of the higher peaks, but not the 100 times per
year they occur. Another factor of detention would be the extended duration of time
needed to get the higher volumes out of the facility.
In a post-development condition, significant flow, volume, and energy are entering
urban streams. Because predevelopment pollutant loadings are zero for these predominant
events, then any post-development discharge, regardless of the constituent concentration,
would be greater than the predevelopment reference point. In the post-development
condition, urbanized sites discharge pollutant loads of many orders of magnitude above
predevelopment conditions and occur more often. If there is no runoff, then where does
the rainfall go? According to Bolton and Watts, and others, much of it is intercepted in
the tree foliage, bark, and branches, which then evaporates. Some rain falls to the forest
floor, where it is absorbed and then makes its way into the ground and gradually seeps
into streams.
Portland Bureau of Environmental Services (BES) test swales

These swales were constructed to test various design characteristics that, at the time, had
not been documented by anyone else (at least not to the knowledge of BES). Issues of
concern include pollutant removal or capture efficiencies of different plant species, soils,
and flow attenuation and velocities. Additional issues may be tested in the future. Both
swales are identical in geometric shape and soil type. Tests to date have been based on a
difference of vegetation in each swale. One swale is planted with native grasses and forbes,
and the other swale is planted with turf grasses. The turf grass was mowed regularly, and
the native vegetation was left to grow naturally. A portion of stormwater runoff from a
Storm Event
East Swale West Swale East Swale West Swale East Swale West Swale East Swale West Swale East Swale West Swale East Swale West Swale
02/16/1999 02/16/1999 04/19/1999 04/19/1999 10/25/1999 10/25/1999 02/22/2000 02/22/2000 03/27/2000 03/27/2000 06/10/2000 06/10/2000
#1
#2
#3
#4
#5
#6
#1
#2
#3
#4
#5
#6
Figure I.6.4 Comparison of test swale inflow and outflow volumes.
100
200
300
400
500
600
700
800
Outflow
Inflow
Chapter I.6:
Runoff Volume (ft3)
900

129
130

East Swale
West Swale
Grab Parameters
pH (field)
Dissolved oxygen (field)
Temperature
Conductivity (field)
Total oil and grease
Nonpolar oil and grease
3
20
4
13
32
50
4
25
3
3
31
34
Composite Parameters
Total suspended solids
Total dissolved solids
Total solids
COD
Total Kjeldahl nitrogen (TKN)
Total phosphorus
O-phosphate-phosphorus, DISS
Nitrate-nitrogen
Hardness
Cadmium, total
Copper, total
Lead, total
Zinc, total
Cadmium, dissolved
Copper, dissolved
Lead, dissolved
Zinc, dissolved
81
28
60
65
54
50
75
16
46
73
65
72
76
47
52
53
64
69
18
51
52
40
38
45
8
33
61
53
62
63
50
38
36
48
Figure I.6.5 Test swale percent pollutant load change, average

six storm events.
50-acre urban area is pumped into each swale with almost identical volumes. Flow meters
measure the flow at the end of each swale. Three pollutant samples are taken at each swale
every 30 min during a storm event and then combined for analysis. A total of 6 events
over the last 2 years have been sampled. Results to date indicate the following.
Runoff attenuation occurs in both swales. The swale with native vegetation retains
up to 41% of the flow and the turf swale retains about 27%. No identifiable conditions
exist to explain the difference; it is assumed that the native vegetation and lack of mowing
allow the swale to facilitate infiltration. This may be due to the robustness of the root
systems and presence of more organic material in the native vegetation swale versus the
turf grass swale. Figure I.6.4 shows the flow comparison for each storm event.
Pollutant removal efficiency is good during all seasons but is better for warm-season
load reductions. Generally, the swale planted with native grasses captures more pollutants
than the turf swale, except for O-phosphate-phosphorus. This may be due to accumulation
of organic matter in the swale, whereas grass clippings were removed from the turf swale.
Recent storms monitored dont show a measurable difference in swale performance even
though the turf has not been mowed for over a year and the inflow has been increased
from 0.04 cfs to 0.08 cfs. Figure I.6.5 shows that both swales perform relatively well.
Summary
Pollutant removal relative to concentrations is good; for example, average total suspended
solids (TSS) removal is 59% for turf and 68% for native vegetation. When loads are
Chapter I.6:
131
calculated based on runoff volume captured in each swale and concentration removal, the
TSS removal percentages are 69% for turf and 81% for native vegetation. Vegetation
maintenance is not necessarily required. A messy or somewhat natural-looking planting
does not indicate the stormwater management functions have been impaired.
Design techniques or green solutions

Techniques are being developed nationwide by many who are helping to shape this new
paradigm (Chapters I.1 and I.5). Water, soil, and vegetation are purposely introduced into
site and building elements previously isolated from each other. Four basic functions result:
water soaking into soil/vegetation; water flowing over soil/vegetation and inanimate
objects; water transpired by vegetation; and precipitation intercepted by vegetation and
evaporated (water even evaporates during dry periods of storm events; see BES tree study,
2001). The combination of these functions is often advantageous to achieve better water
distribution. Pollutant capture is achieved by water having to filter through the soil and/or
vegetation. Atmospheric pollutants are captured in plant foliage and then trapped in the
soil, where many of them have an opportunity to break down. All these approaches allow
runoff to be diffused, which also allows pollutants to be distributed in the landscape
instead of concentrated. Increases in urban air and water temperature are minimized or
eliminated by shade on impervious surfaces and, in the case of eco-roofs, nearly all
precipitation is retained during warm-season months (May to October). The physical forms
of these techniques are almost infinite, but are described here as:
Eco-roofs Figure I.6.18
Stormwater planters Figures I.6.13 and I.6.14
Infiltration garden Figure I.6.6
Landscape swales Figure I.6.11
Vegetative filters Figure I.6.28
Portland demonstration projects

Buckman Heights
A redevelopment project in the combined sewer area was designed in 1996 and opened in
1998. The sites previous use was for a new-car dealership parking lot. The site is 2 acres,
with a 150-unit, 4-story apartment building and some surface parking and underbuilding
parking. The project has many environmental attributes including car sharing for the tenants.
The owner, Prendergast Associates, wanted to do its part to remove the sites runoff from the
combined sewer system. The buildings are organized around a main courtyard; the traditional
layout is articulated with low seating walls off the sidewalk and two large planting beds
designed as landscape infiltration areas to filter and absorb the stormwater from the buildings downspouts. The parking areas are designed with care and detail to reduce the presence
of the automobiles and absorb the water runoff from the paved surface.
Landscape infiltration
Courtyard. Figure I.6.7 shows two 18 ft 45 ft infiltration gardens integrated with
the site. The gardens were designed to accept runoff from the rooftops and the surrounding
courtyard paved areas to flow into the vegetation. The planter area tapers from 6 in. along
the perimeter of the surrounding walkways to 18 in. at the center. Moisture-tolerant plants
of spirea, iris, Oregon grape, and astibe were planted within a Japanese holly border.
132
Optional
overflow
Max. 2:1 side
slopes
Flow
12 max.
reservoir depth
Optional
piped inflow
18 topsoil
4 ft. minimum
Figure I.6.6 Landscape infiltration.

Section Not to Scale
Description: Landscape infiltration areas can be integrated into the site landscaping. The design
can be formal or informal in character. They may be used in courtyards, parking lots, or where
other planting areas are available. Although the area is saturated during storm events, infiltration
occurs quickly.
Stormwater management function and sizing: The system works by holding runoff and allowing
pollutants to settle as the runoff infiltrates. Flow and volume are also managed with these facilities.
Allows evapotranspiration and groundwater recharge and retains warm weather runoff. Depending
on soil type and infiltration rate, this facility may provide 10- to 25-year event disposal. Using above
proportions size at 0.045 impervious area for the 0.9 in., 24-h storm event.
General specifications: Acceptable soil types A & B. Minimum soil infiltration rate of 2 in./h. Facility
storage depth may vary from 2 to 12 in. Filters can be planted with a variety of trees, shrubs, and
ground covers, including grasses appropriate for periodic inundation. Freeboard not required.
Figure I.6.8 displays an overflow pipe set 9 in. above the bottom of the basin. Runoff enters
the landscape infiltration area and soaks into the soil, except for large storms that flow
out the overflow. This keeps the area at a maximum depth of water during storms of 9 in.
with percolation rates estimated at 2 in./h; the area is expected to drain within 5 h.
What have we learned? Aesthetically it is a very pleasing design. Based on visual
observations, infiltration rates appear to be increasing. Infiltration tests will be conducted
this spring to confirm these observations. If true, then the overflow inlet can be raised
several inches to allow more runoff to be infiltrated. Vegetation appears healthy and is
growing well. Pruning is done annually in the fall to maintain the overall desired aesthetic.
From a functional perspective, the pruning could be reduced. No fertilizers or pesticides
will be needed, although the entire site has an irrigation system. Piping from the downspouts to the planter has a tendency to clog with sediments from the roofing materials
and tree leaves. This requires at least annual cleaning to maintain good drainage; however,
during a very intense 1.4 in./3-h storm on October 1, 2000, the pipes were full and runoff
Chapter I.6:
133
Figure I.6.7 Buckman Heights courtyard infiltration garden. (Color version available at www.gsd.
Figure I.6.8 Buckman Heights courtyard infiltration garden, with outlet riser. (Color version available at www.gsd. harvard.edu/watercolors. Password: lentic-lotic.)
simply overflowed across the lawn and walkways into the planter. Two alternatives, and
probably better designs, would have been shallow surface channels integrated with the
walkway design or larger pipes.
Parking lot. Figure I.6.9 shows the uniquely designed landscape infiltration perimeter. Perimeter landscaping is required by city code; for a project like this, it requires a
5-ft wide strip the entire length of the parking lot, 200 ft. The lot was one of the first to
use substandard dimensions with a 20-ft-wide aisle and 17-ft-long 8.5-ft-wide stalls.
134
Figure I.6.9 Buckman Heights parking lot landscape infiltration. (Color version available at
www.gsd. harvard.edu/watercolors. Password: lentic-lotic.)
Parking is at 90. Unique about this integrated landscape is the lack of a freeboard
(precautionary measure to contain flows). When considering the configuration and grading of the site, a freeboard was not needed to protect property or people. This allowed a
smooth transition from the pavement edge into the planting area. Runoff flows into the
landscape via curb cuts and then infiltrates into the relatively porous soil. A 2-in. storage
capacity is available for runoff as it soaks in or for large storm flows to move by displacement toward two inlets at each end of the strip that drain into underground dry wells.
Plantings include only Red Sunset maple trees (Acer rubrum, red sunset) and Oregon
grape as a hedge. Approximately 8,000 ft2 of concrete surface drains into the landscape,
which is about 1,000 ft2 in area. It is estimated that this area will infiltrate the 10-year
storm event (3.2 in./24h).
What have we learned? The plantings look good and the runoff is captured in the
landscape area. Visual observations to date have indicated the area will hold at least a
2-year storm event (2.4 in./24 h), as occurred in November 1998. In one area the pavement
was not sloped at the specified 2% gradient and causes some ponding in the corner of
one car stall. Grinding the existing curb cut about in. will allow the water to drain off.
The lack of a freeboard has not posed any concerns. Figure I.6.10 shows one of the curb
cuts that must be kept clear to allow flow passage. These curb cuts have not been cleaned
since installation and are still relatively unobstructed.
Buckman Terrace
Buckman Terrace is another redevelopment project by Prendergast Associates and is across
the street from Buckman Heights. The project was designed in 1997 and opened in 1999.
This is a 0.8-acre site with 150 apartment units, all with under-building parking, and a
1,500 ft2 commercial section in a 4-story structure. The building also has car sharing and
numerous other environmental attributes. An eco-roof has been installed on the commercial portion and another at the main entrance of the building. Landscape planters are
Chapter I.6:
135
Figure I.6.10 Buckman Heights parking lot curb cut. (Color version available at www.gsd.harvard.
being used on the east side, and a landscape swale has been installed on the west side of
the building. The landscape techniques integrate lush, moisture-tolerant planting with the
function of stormwater quality and environmental enhancement.
Landscape swales
Westside swale. Figure I.6.12 shows the westside swale adjacent to the building
where all roof downspouts discharge. Approximately 13,000 ft2 of rooftop drains into a
430-ft swale. The swale is 6 ft wide and 3 in. deep; it has rock check dams every 15 ft. The
swale is sized to convey the 25-year storm event flows but also to provide detention of
all storms up to the 10-year event. These storms are detained behind the check dams. All
flows that exceed the infiltration capacity of the soil discharge to a catch basin at the
downstream end of the swale. A pea gravel mulch was used to slow the flow and provide
opportunities for sediment deposition and some infiltration. The swale gradient is approximately 2% until it reaches the last 100 ft, where it increases to 4%. Purposefully excluded
from the design was the standard 12-in. freeboard. This was done for two reasons: first,
the added safety of a freeboard was not needed since there is no possibility of damage to
the building or adjacent property; second was aesthetics (i.e., because of the narrow area,
a 12 in. deeper swale would have been unsightly and dangerous). Plantings include
sedges, miscanthus, spirea, Oregon grape, and Japanese iris.
What have we learned? The plantings are attractive and add considerable interest to
this side of the building. Visual observations immediately after a large storm event
(1.4 in./3 h) were conducted on October 1, 2000. Design and construction quality is important for best water management. The check dams were incorrectly constructed parallel to
the flow (they were supposed to be perpendicular), which caused peak flows to bypass
and erode part of the pea gravel on the 4% slope section. The 4% slope section only had
grasses planted in a single row and should have had three rows with triangular spacing.
The pea gravel has worked successfully as mulch to protect the swale and help filter flows.
136
Check dams @ 12 intervals,

with 6 maximum water depth
6min.
For parking
lots, tire
stop or curb
flow
For parking lots 12

x 12 clear flow area
at cut-outs
18 topsoil
4 ft. minimum
Figure I.6.11 Landscape swale.

Description: Landscape swales are planting areas with a slight depression of up to 6 in. that allow
runoff to enter, infiltrate, and flow through. They are usually long and narrow in width, which makes
them well suited for parking lots and other narrow landscape spaces. Swales are constructed with
a variety of trees, shrubs, grasses, and ground cover, depending on soil.
Stormwater management function and sizing: Swales capture pollutants as runoff is detained
and absorbed in the soil, vegetation, and organic matter. Using above proportions, size at 0.05
impervious area. Detention is provided for storms up to the 10-year event. Swales help mitigate
runoff temperatures by retaining most of the runoff in warm seasons. Groundwater recharge occurs
as check dams facilitate infiltration.
General specifications: Acceptable for all soil types. Minimum swale length is 20 ft. Maximum
slope is 6%. Clay soils shall be amended with 50% sandy loam in the top 18 in. of the swale.
Check dams to be of durable, nontoxic materials i.e., rock, brick, and old concrete. Check dams
shall be width of swale 35 in. height. Swales using these design criteria need not bypass larger
storms. Liners are not needed unless required for groundwater protection or to protect building
foundations. Freeboard not required.
Stormwater planters
Eastside planters. Two designs were used here; Figure I.6.15 shows the north section
planter is at grade and somewhat inconspicuous next to the building. Figure I.6.16 shows
the raised planters in the south section at 18 to 36 in. above grade. Figure I.6.17 shows
that runoff is directed into the planters via scuppers from the roof. Gravel soakage trenches
accept water as it filters through the upper soil/vegetation portion of the planter. Plantings
include Japanese spurge, iris, vine maple, and Oregon grape. The planters have more
surface area than required for Portland conditions, and thus the reservoir space is only
2 in. deep.
Chapter I.6:
137
Figure I.6.12 Buckman Terrace west side landscape swale. (Color version available at www.gsd.
What have we learned? Aesthetically, the design is equal to the original concepts
considered without the use of water. Water is very visible to the tenants as they walk by.
Because the soakage trench was within 8 ft of the building foundation, the owner and his
consultants expended quite an effort to get building bureau approval. If the facility had
been only 2 ft farther away, no special approval would have been required. In tight urban
settings like this, the building bureau prefers the Portland CD planter. The CD unit is
not designed to allow infiltration, other than some incidental amounts. Although very beautiful, the non-native plants might not perform as well as some native moisture-loving species.
Eco-roofs
The entire building has a roof area of approximately 25,000 ft2, and the building is constructed with sufficient weight capacity to hold an eco-roof. As a test, eco-roofs were
placed on two sections. Figure I.6.19 shows a 200-ft2 eco-roof above the front entrance. A
small, 25-ft2 rooftop above drains into the eco-roof. Figure I.6.20 shows the main eco-roof
over 1,500 ft2 of commercial space, which has full solar exposure. An additional 750 ft2 of
impervious roof drains into the main eco-roof. Figure I.6.21 shows a closeup of Oregon
sedum on the entrance eco-roof, which is also planted with sword fern, licorice fern, and
white stonecrop. It is on the east side and in the shade of a north-facing wall. Both were
planted in March 2000. The commercial eco-roof was planted with two species of Oregon
sedum, various wildflowers, native grasses, and a few licorice ferns. Grasses and wildflowers were planted from seed, and mulch was hand broadcast to protect against wind
erosion. Figure I.6.22 shows a globe flower (Gilia capitata) blooming on the main eco-roof.
An irrigation system has not been installed for either eco-roof. The soil profile is 20 lb/ft2
when saturated and 4 in. deep. An American Hydrotech waterproof membrane and reservoir drain system was used. BES staff specified the soil mix and vegetation.
What have we learned? Grasses and wildflowers achieve a graceful, flowing appearance. It is reminiscent of eastern Oregon or the Midwestern American prairie. During the
138
Downspout or
Other conveyance
Plantings
flow
Overflow option
Reservoir: 4-12
Gravel as splash
block and mulch
to protect soil
18Sandy loam topsoil
Gravel, or
recycled
glass
min. 12
wide
Topsoil or subsoil
Note: Gravel
trench may
not be needed
if existing soil
has sufficient
infiltration
rate
Existing soil
Existing soil
Figure I.6.13 Stormwater planter AB.

Description: Planter AB is designed with a pervious bottom. The planter is used where infiltration is
desirable. Planters are excellent for dense urban development.
Stormwater management functions and sizing: Planter AB is designed to allow runoff to filter
through the planter soils and vegetation (thus capturing pollutants) and then infiltrate into the native
soils (flow control). The planter is sized to accept runoff and temporarily store the water in a reservoir
on top of the soil. Reservoir sizing, above the 18-in. topsoil, is for a 0.9-in., 24-h storm event. To
calculate, use impervious area square feet (sf) .045 = reservoir cubic feet (cf) of storage required.
The infiltration gravel area can be designed to accommodate any storm event.
General specifications: Acceptable soil types A & B. There are numerous design variations. The
planters shall be designed to allow captured runoff to drain out in 2 to 6 hours after a storm event.
Plantings shall be appropriate for moist and seasonally dry conditions and can include rushes,
reeds, sedges, iris, dogwood, currants, and numerous other shrubs, trees, and herbs/grasses.
Topsoil shall have a minimum infiltration rate of 2 in./h. Sand/gravel area may not be required if
existing soil has at least 5 in./h infiltration rate. The sand/gravel trench width, depth, and length
are to be determined by a qualified professional. Minimum planter width is 30 in.; there is no
minimum length or required shape. The structural elements of the planters shall be stone, concrete,
brick, wood, or other durable material. If treated wood is used, it shall not leach out any toxic chemicals.
Planters within 10 ft of structure will probably require special approval from the local building code
agency.
warm season, storm event runoff was visually observed to be very low or nonexistent.
The eco-roof had the capacity to hold much of the additional flow from the other roofs.
During winter storms, runoff occurs often but is detained. Many of the plants survived
Chapter I.6:
139
Reverse bend trap

Plantings
Downspout
Reservoir 4-12 max.

Stone or other
inert mulch to
protect soil
18 Sandy loam topsoil
Perforated pipe
@ 10 O.C.
12 Gravel, sand
or other material
Sub-grade or
existing soil
Waterproof
building as
needed
Foundation drains
as required
Pipe to main
storm system
Figure I.6.14 Stormwater planter CD.

Description: Planter CD is designed with an impervious bottom or is placed on an impervious
surface. This planter is used where infiltration is not possible or desirable, such as unstable slopes
or brownfields. Planters are excellent for dense urban development.
Stormwater management function and sizing: Pollutant reduction is achieved as the water filters
through the soil; flow control is obtained by storing the water in a reservoir above the soil. Nominal
infiltration can be allowed if soils and other geotechnical issues are addressed. The planter is sized
to accept runoff and temporarily store the water in a reservoir on top of the soil. Reservoir sizing,
above the 18-in. topsoil, is for a 0.9-in., 24-h storm event. To calculate, use impervious area square
feet (sf) .05 = reservoir cubic feet (cf) of storage required.
General specifications: Acceptable for all soil types. There are numerous planter design variations.
The planters shall be designed to hold water for no more than 4 to 6 hours after a storm event.
Plantings shall be appropriate for moist and seasonally dry conditions and can include rushes,
reeds, sedges, iris, dogwood, currants, and numerous other shrubs, trees, and herbs/grasses.
Minimum planter width is 18 in.; there is no minimum length or required shape. Topsoil shall have
a minimum infiltration rate of 2 in./h. Sand/gravel shall have a minimum infiltration rate of 5 in./h.
The structural elements of the planters shall be stone, concrete, brick, wood, or other durable
material. If treated wood is used, it shall not leach out any toxic chemicals. Irrigation is optional,
although plant viability shall be maintained.
or reseeded with only one hand watering. Although no maintenance was conducted this
year, it appears the grasses will need to be mowed at least once a year. Should the owner
prefer a different plant association, sedum might be added at some future time. It is very
important to assure good vegetation coverage, especially over the lightweight soils to
prevent wind erosion.
On another eco-roof project, the Hamilton Apartments in downtown Portland, almost
an inch of soil was lost to wind erosion. Depending on the initial planting scheme, cover
crops, such as common clover, may provide excellent soil coverage, which happened on
a section of the eco-roof. Water from air conditioning condensate is a possible source of
140
Figure I.6.15 Buckman Terrace stormwater planter at grade. (Color version available at www.gsd.
Figure I.6.16 Buckman Terrace raised stormwater planters. (Color version available at www.gsd.
free, nonpotable water for irrigation. Condensate flows were significant during the hottest
part of the summer, with flows measured at 12 oz/min in the afternoon and 6 oz/min in
the late evening. This might prove to be a free source of irrigation water, if considered
during the design phase. On this project, BES is testing to determine characteristics of
planting methods, measurement of runoff flows and precipitation, and viability of soil
and vegetation. Other issues may be addressed in the future. Figure I.6.23 shows the
Chapter I.6:
141
Figure I.6.17 Buckman Terrace raised planter and scupper. (Color version available at www.gsd.
southeast quadrant of the eco-roof. Figure I.6.24 shows one of several stonecrop species
in bloom last August. Figure I.6.25 shows some of the moss that colonized certain areas
of exposed soil and helped reduce wind and soil erosion. Lightweight soils must be fully
covered to prevent erosion. A Garland Co. waterproof membrane and planting design
was used on this project.
Bureau of Environmental Services (BES) Water Pollution Control Laboratory

The project site was previously a 6-acre industrial site along the east bank of the Willamette
River and adjacent to a city park on its north side. It was used for industrial activities for
almost 100 years. The upland neighborhoods are a mixture of old residential and industrial
land uses, with new residential conversion and infill occurring rapidly. The area was
served by a combined sewer system that overflowed into the Willamette River about 75
times annually. BES installed a $5.4 million separated stormwater pipe system for 50 acres
that drains to the water garden. At the time, treatment of separated stormwater was not
required, but it was decided that a water quality pond would be constructed to reduce
pollutants entering the river. Neighborhood citizens were very concerned that BES was
going to create a problem pond; however, after considerable effort by the citizens and
BES, a mutually agreeable plan was prepared, and a water garden, as it was eventually
named, was constructed.
Simultaneously, BES was preparing to build a new Water Pollution Control Laboratory
on another portion of the site. The new building would have laboratory, office, and field
operations space, as well as a public meeting room for the neighborhood and others. The
building footprint is approximately 12,000 ft2, with parking for 60 cars, and 1.5 acres was
allocated for the water garden. About one acre still remains vacant for potential expansion.
The parking lots were designed with landscape swales (see subsection on BES test swales).
The project was designed in 1995 and opened for use in 1997. Combining the best of the
artistic and the utilitarian, the design transformed this once-industrial site into a meaningfully sculpted landscape, integrated with ecological processes. Each component of the
site speaks to the inherent poetry of water and its role in our environment.
142

Plants:
Herbs, grasses,
succulents (such as
sedum)
Mulch or material to
prevent surface wind
and rain erosion
Soil: 2- 6
Drainage layer
Waterproof membrane
and root protection

Roof structure
Figure I.6.18 Eco-roof (green roof).

Description: An eco-roof is a lightweight roof system of waterproofing material with a thin soil layer
and protective cover of vegetation. The eco-roof can be used in place of a traditional roof.
Stormwater management function and sizing: The eco-roof captures and then evapotranspirates
10 to 100% of precipitation, depending on the season. Roof runoff rates are significantly retarded
because the rain must first soak through the soil before running off the roof. An eco-roof provides
peak flow detention for storms up to the 10-year event. Eco-roofs mitigate runoff temperatures by
retaining most of the runoff in warm seasons. Groundwater recharge can occur where roof drains
flow to landscape areas. Sizing is equal to the square footage of eco-roof.
General specifications: Quality waterproof material appropriate for eco-roof application. Soil of
adequate fertility and drainage capacity at depths of 2 to 6 in. Building structure adequate to hold
an additional 10 to 25 psf weight. Self-sustaining vegetation, without the need for fertilizers or
pesticides. Soil coverage to prevent erosion must be established immediately upon installation, by
using mulch or protective blanket or vegetation mats (sod). Ninety percent plant coverage to be
achieved within 2 years. Temporary irrigation to establish plants is recommended. Permanent
irrigation systems using potable water may be used, but the water application shall not exceed
0.25 in. every 10 days for JuneSeptember season. Irrigation is not needed from October to May.
Projects are encouraged to have alternative means of irrigation, such as cooling tower condensate
or other nonpotable sources. For roof slopes greater than 10%, measures (such as geotextile
webbing, and sleepers) shall be used to prevent soil slippage.
The water garden (pond)

From the very beginning, this facility had to be special. Attention to aesthetics and integration with the neighborhood was essential. The upland catchment contains approximately 40 acres of impervious surfaces and 10 acres of mixed pervious surfaces. Some of
these include gravel yards for storage of heavy equipment. The pond was designed to
accommodate the peak flow from a 0.83-in./24-h storm event, with a diversion structure
to bypass large storm flows directly to the river (bypass flows are assumed to carry lower
pollutant concentrations). Figures I.6.26 and Figure I.6.27 show the focal point of the
design, a 1-acre pond formed from two converging circles. Elements include a circular
stone wall to house the pond outlet structure, a 100-ft-long rock-filled concrete shute that
conveys flow, yet provides an artful sculpture during dry weather, and a lushly planted
pondscape that is integrated with the building landscape design and the adjacent park.
The curvilinear flume in the upper cell is reminiscent of a glacial moraine. It slows the
Chapter I.6:
143
Figure I.6.19 Buckman Terrace entrance eco-roof. (Color version available at www.gsd.harvard.
Figure I.6.20 Buckman Terrace main eco-roof. (Color version available at www.gsd.harvard.
stormwater while directing it into the detention cells. The cells are planted with a variety
of aquatic and emergent plant material that naturally facilitates sedimentation and biofiltration of pollutants. Circular weep holes on both sides of the flume uniquely display the
flow of water. The plantings include a mix of native and nonnative species, and include
Oregon ash, red alder, red maple, several grasses, redtwig dogwood, Douglas spirea,
Oregon grape, and numerous wetland species. An observation platform was designed as
144
Figure I.6.21 Buckman Terrace entrance eco-roof Oregon sedum (Sedum oreganum). (Color version
Figure I.6.22 Buckman Terrace main eco-roof globe flower (Gilia capitata). (Color version available
an extension of the main spine of the building and extends over the water. Three monitoring stations were set up to measure flows and pollutants to determine pond efficiency.
What have we learned? Pollutants and flow have been monitored over 2 years for a
total of 11 storm events. Generally, pollutant removal is good all year for most concentrations except cadmium, copper, and zinc. Some of these pollutants are seeping in from
subsurface flows and surface flows during high-intensity storm events. Phosphorous and
Chapter I.6:
145
Figure I.6.23 Hamilton Apartments eco-roof southeast quadrant. (Color version available at www.
gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Figure I.6.24 Hamilton Apartments eco-roof stonecrop. (Color version available at www.gsd.
nitrogen are also high on occasion. The pond remains wet all year, but summer discharges
are small to nonexistent. Evaporation and soakage into surrounding soils are significant.
Moderate amounts of excess irrigation water and unknown nonstormwater flows from
the pipe system flow into the pond. Because the pond does not have a constant summer
flow, however, the pond water occasionally becomes anaerobic and objectionable odors
have been recorded. This problem remains to be addressed; one solution may be to drain
146
Figure I.6.25 Hamilton Apartments eco-roof moss. (Color version available at www.gsd.harvard.
Figure I.6.26 Water garden stone wall outlet structure. (Color version available at www.gsd.harvard.
the pond every summer. BES tried to accurately measure inflow, outflow, and bypass, but
the subsurface and surface flows cannot be measured by the flow meter in the intake pipe.
These uncontrolled sources also compromise the pollutant removal information. Education
is needed in the neighborhood to reduce pollutants at their sources.
After construction an inspection of the diversion structure showed that the concrete
dam had been installed in the wrong place and low flows were not entering the pond.
Chapter I.6:
147
Figure I.6.27 Water garden curvilinear inlet flume. (Color version available at www.gsd. harvard.edu/watercolors. Password: lentic-lotic.)
The dam was immediately removed. It is very important to assure good design and
construction quality control. Check the engineering and construction details. Concerns
about warm-season water temperatures impacting the river are still being investigated,
however, the pond does not discharge very much in the warm season. Sediments are
accumulating in the rock flume and vegetation is starting to grow. It has been decided to
let this continue and try to determine if an unmaintained pond loses its efficiency, although
aesthetic pruning will continue. Many undesirable and potentially hazardous materials
have been removed from the pond, including hypodermic needles, rat poison, trash, and
debris. Use of a forebay to trap these things would have been desirable.
Another issue the pond vegetation and presence of water has been attractive to
wildlife. Ducks are always around, and at least two families have nested there. The ducks
sometimes feed in the bottom sediments where the pollutants are being trapped. It is yet
to be understood whether this is a problem to the wildlife. Domestic dogs also like to play
in the pond. Raccoons and other wildlife have been observed. Another note concerning
fish and flooding: during the 1996 flood, the pond became a backwater for the river flows.
Carp and potentially other fish found the pond and, as the flood receded, they were
isolated from the river. When the hot summer of 1998 occurred, all the fish died. It is
unknown whether any currently threatened species might have used the pond for refuge
during the flood, but they surely could not survive the summer conditions. Designers
should keep this is mind when proposing facilities near natural water bodies.
Vegetative filters
Figure I.6.29 demonstrates the use of scuppers as roof runoff cascades to the vegetative
filter located on the south side and adjacent to the building. Most of the building has a
metal roof of about 12,000 ft2 that directs flow into a gutter with several steel scuppers.
These scuppers allow the runoff to freefall into the garden area below. Large stone and
rock are at the impact point of each freefall to diffuse energy and spread the water into
the plantings. Planters have a mixture of lush ornamental, native, and wetland plants and
are lined with crushed stone to provide a visually unique image year-round. Runoff is
148
Flow spreader
check dam at 5 ft
intervals
Impervious area
sheet flow or pipe to
spreader
flow
Optional inlet, adjust height

to allow infiltration,
depending on soil
18 topsoil
10 ft. minimum
Optional
pipe to
discharge
Figure I.6.28 Vegetative filter.

Description: Vegetative filters are gently sloped areas. Stormwater enters the filter as sheet flow
from an impervious surface or is piped and then converted to sheet flow using a flow spreader.
Stormwater management function and sizing: Flow control is achieved using the relatively large
surface area and a generous proportion of check dams. Pollutants are removed through filtration
and sedimentation. Using above proportions, size at 0.6 impervious area.
General specifications: Acceptable for all soil types. Filters shall be a minimum of 20 ft. 10 ft.
Maximum slope is 10%. Check dams shall be of durable, nontoxic materials i.e., rock, brick,
and old concrete. Check dams shall be the width of the filter 35 in. height. Filters designed
using these criteria will not need to bypass larger storms. Runoff shall enter the buffer as predominately sheet flow. Check dams and flow spreaders are required. Filters can be planted with a
variety of trees, shrubs, and ground covers, including grasses. Freeboard not required.
detained in the planters where some infiltration occurs. Large storm events have runoff
that overflows to a catch basin.
What have we learned? It works great and looks good, too. Unfortunately, the planting areas are unnecessarily deep to allow for a nonessential freeboard. If a failure occurred
at the catch basin inlet, the excess runoff would flow into the parking lot drains an
important reason to discontinue use of the freeboard, where they are not needed, is to
allow more surface area for runoff to spread within the planting areas. The addition of a
few check dams and raising the inlet grate a few inches will help achieve this goal.
The Oregon Museum of Science and Industry (OMSI)

This is an 18-acre redevelopment site, with a 100,000 ft2 museum and exhibit space and
parking lots for 700 cars. The project was Portlands first major demonstration of on-site
parking lot stormwater management using 10 landscape swales to convey, infiltrate, and
Chapter I.6:
149
Figure I.6.29 Roof scuppers and vegetative filter. (Color version available at www.gsd.harvard.
filter runoff. These swales were used in place of the originally proposed, conventionally
raised landscape medians. Based on visual observations, as the soil and vegetation
mature, significant infiltration now occurs and runoff is only associated with large storm
events. (See subsection on BES test swales.)
This project first came to the attention of the Portland Bureau of Environmental
Services (BES) in 1990 when plans were submitted for review. At that time, neither the
city of Portland nor the state of Oregon had specified site design requirements for stormwater quality discharges to the Willamette River; however, BES was gearing up to address
forthcoming NPDES stormwater regulations and a combined sewer overflow problem of
significant proportions, so clean rivers were, and still are, of concern. Following a review
of the preliminary OMSI site plans, BES staff approached OMSI with the unprecedented
request that it voluntarily redesign the parking lots and landscape to capture pollutants
in stormwater runoff. This proposal did not affect the layout of the buildings and parking.
The BES suggestion was to adjust site grading and change the already proposed landscape
medians to accept rather than shed runoff.
OMSI was very interested in the environmental approach, but the nonprofit organization was under a tight budget and timeline. It agreed to change the design on the
condition that the existing schedule be met with no overall increase in costs. The OMSI
consultant team determined that even with the redesign fees, the related construction cost
savings would result in a net reduction of project costs. Taking the environmental approach
would actually be less expensive to construct.
The parking lots were initially redesigned to include grass/turf swales to filter out
pollutants as the runoff traveled through them. OMSI took this a step further and required
the landscape architects to improve the design to detain water longer and to incorporate
native and wetland vegetation. OMSI considered this mini linear wetland concept a
more attractive and educational approach for its new facility. Interpretative signs were
installed to educate the public about the benefits of the swales and wetlands in improving
water quality. The city, for its part, had established a special team from various bureaus
to assist in moving the project smoothly through the city approval process. This took an
150
Figure I.6.30 OMSI Willamette riverbank. (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
enormous effort, because no existing policies or codes existed which allowed developers
to use water quality site design techniques.
Figure I.6.30 shows some of the restoration and pedestrian improvements along the
Willamette River bank. The sites western boundary is along the riverfront, which was
stabilized with rock riprap and included intensive plantings of native riparian vegetation.
Landscape swales
Figure I.6.31 shows one of the ten swales. The swales are 6 ft wide and vary in length
from 100 to 250 ft, for a total length of 2,330 ft. They were originally designed as biofilters
but have continued to exhibit good infiltration characteristics. Check dams were installed
every 50 ft to slow the flows and encourage infiltration. The parking dimensions were
modified to allow more space for plantings and minimize impervious surfaces. (Special
approval was required; minimum stalls were 9 ft by 18.5 ft.) Stalls are 8 ft wide by 16 ft
long with bumpers designed to overhang the curbs. Parking is at 90 and the aisle is 24 ft
wide. Recently, the Portland city council approved a new parking lot code that allows
20-ft aisles and 9-ft by 16-ft stalls. Landscape space in front of the stalls is required at 6- to
8-ft widths to allow for stormwater management. No parking spaces are lost, because the
code just provides more room for landscape and less for cars.
What have we learned? In 1996, the BES prepared a water quality audit and estimated that the bioswale system captures 50% of the average annual total suspended solids
(TSS) loadings from the site. It was also estimated that with minor design improvements,
such as additional check dams and more curb cuts, TSS capture would increase to 90% of
the average annual site loading. Curb cuts were installed at 30 ft on center. Performance
could be improved by installing curb cuts at 10 ft on center. These improvements have
not been implemented because a major portion of the site was modified when a public
street was constructed. Many of the improvements may still be made in the future. Visual
observations continue to show that the swales allow much of the runoff to infiltrate.
Chapter I.6:
151
Figure I.6.31 OMSI landscape swale. (Color version available at www.gsd.harvard.edu/watercolors.

Swale maintenance is incorporated into the normal operations of the site. The landscape medians were always intended to be a landscape feature of the site, which would
require maintenance. The bioswales only require a little more attention. Curb cuts must
be checked and cleared at least once a year. City staff have observed that some curb cuts
were poorly constructed and/or located and need modification to reduce build-up of
sediments, which blocks runoff from entering the swales. In hindsight now all swales
were unnecessarily designed with 12-in. freeboards. Because the bioswales are oversized,
it will take several decades before any accumulation of pollutants will need removal, if
ever. The parking lot has significant use and lots of trash. Wind blows the trash into the
swale vegetation, where it is trapped. This has been considered a desirable aspect of the
swales, because the trash is somewhat camouflaged until OMSI staff removes it.
The success of OMSI demonstrates that water quality and stormwater runoff measures
can fit into constrained spaces, save construction costs, and have an attractive appearance.
The owner documented a savings of about $78,000 in construction costs. These savings
were achieved through the reduction of pipe, manholes, and catch basins.
Custer Park
Custer Park is a small neighborhood park with a playground, open lawn, and softball and
soccer fields. This was a 4-acre park renovation project, with high demand for a full-sized
soccer field. Before the park was developed in the 1950s, a small seasonal creek ran through
the site. Several seasonal springs also contributed to the small creek flows. The original
1950s park design called for the piping of the creek to drain the springs and stormwater
runoff from an adjacent residential neighborhood upland of the park. The Parks Bureaus
desire was to have the water out of site and out of the way, but Parks has been struggling
with wet turf ever since. In 1996 and 1997, BES, Parks, and Murase Associates worked
with very little budget and redesigned the site to daylight a 400-ft section of piping.
Appearing as a daylighted swale, it runs adjacent to a new pathway and is planted with
152
Figure I.6.32 Custer Park landscape swale. (Color version available at www.gsd.harvard.edu/
watercolors. Password: lentic-lotic.)
native hydrophytes and riparian shrubs and trees. A series of stone weirs slows the velocity
of the flow, allowing sediment to drop out, while the plantings provide filtration. The
swale terminates in a detention pond that removes much of the remaining pollutants from
the flow. Although the swale provides improved water quality and makes visible those
parts of the hydrologic cycle that are usually hidden, it also introduces habitat that is
beneficial for wildlife and adds an additional layer of interest to park users. The completed
project responds to the neighborhood desire to return nature to the park and provides a
larger playing field as well.
Landscape swale
The site terrain has a grade change of 20 ft over a distance of 500 lineal ft. It was determined
that only the city water quality storm event would be used to determine the bioswale
design. These relatively small storm events would not pose as much risk to the park, and
the larger storms are allowed to continue to flow within the original pipe system. A
diversion structure was installed to direct small storm flows into a swale cut into the slope
along one side of the park. At the same time the soccer field was extended up to the edge
of a pedestrian path that runs adjacent to the bioswale on the other side. Figure I.6.32
shows the swale from the downstream end looking up toward the park.
What have we learned? It works great and looks good, too. Plantings were not sufficient to allow adequate coverage and establishment in the first two years but now are
doing fine. The swale included an unnecessary 12-in. freeboard, which diminishes the
aesthetics. Because the project is within an open area and a high flow diversion occurred
upstream, the freeboard was overdesigned. A positive aspect of the excessive depth of the
swale is its capacity to accumulate sediment for many decades without the need for
maintenance.
Chapter I.6:
153
Wrap-up
Water is the main theme of this chapter, but as many philosophies espouse, all things are
connected. The essence of this chapter is to present techniques that help the urban environment function in a more natural way. The reason for doing such is to reduce negative
impacts caused by human development. Although the case studies are within the city of
Portland, the principles are universal and can be applied to any region. The success of
these projects is proof that ecological design not only benefits the environment of humans
and wildlife but also often costs less to implement and sustain. Perhaps, we can have
our cake and eat it, too. It is no easy task, however; many institutional barriers and
professional mindsets must be overcome. Design, research, demonstration projects, and
education are all key elements in helping to bring these and other new approaches to the
professional community.
Acknowledgments
The authors thank Ryan Retzlaff, Emily Hauth, Nicci Lambert, Emily Brown, Elizabeth
Liptan, Scott Murase, and Sue Brantley for their assistance in the preparation of this
chapter.
Literature cited
Beyerlein, D. and Brascher, J., Traditional alternatives: Will more detention work? in Salmon in the
City, Tom Holz, Amer. Public Works Assoc., Washington Chapter, 45, 1998.
Bolton, S. and Watts, A., Results from forest hydrology studies: Is there a lesson for urban planners?
in Salmon and the City, Tom Holz, Amer. Public Works Assoc., Washington Chapter, 49, 1998.
Portland, Fifth Stormwater Monitoring Report (July 1, 1999June 30, 2000), National Pollutant
Discharge Elimination System (NPDES) Municipal Separate Storm Sewer System Discharge
Permit Number 101314, Annual Compliance Report No. 5, prepared for Oregon Dept. Environ. Quality, submitted by City of Portland, Multnomah County, Port of Portland, Nov.
30, 2000.
Portland, Bureau of Environmental Services, Tree Monitoring Report, Portland, OR, Mar. 12, 2001.
Disclaimer
The discussion in this chapter is not intended to substitute for professional advice applicable to specific project circumstances. Design approaches are offered to facilitate understanding of the concepts and must be considered in terms of the project, local building
codes, and regional climate. Readers are urged to seek professional assistance before
applying any of the techniques of this chapter. The techniques and other information
presented may not represent the latest, approved approaches of the City of Portland, OR.
154
Response
Letting it soak in
It, here, refers to both the rainwater and the message about how to effectively manage
it. This chapter by Thomas Liptan and Robert Murase identifies significant elements in
the paradigm shift now taking place in stormwater management of moving away from
centralized and expensive infrastructure toward localized and dispersed micromanagement. Whereas larger-scale reconfigurations of impervious surfaces are important, as, for
example, discussed by Jennifer Zielinski in Chapter I.3, one very significant element of
low-impact development championed by Larry Coffman and Michael Clar in Chapters I.5
and I.15, respectively, is attention directed to the small, but numerous, impervious surfaces
of roofs and parking lots.
Liptan and Murase make the point that rain gardens are both an aesthetic and costeffective approach to managing stormwater in urban settings. In particular, they invoke
a clarion call for water sensitive designs to integrate earth, water, and plants with the site
and the surrounding architecture. In this concept of urban nature, stormwater becomes
a friendly companion to be celebrated as a design feature rather than tucked away as
a nuisance. The results of engaging in such a process, as Liptan and Murase state, will be
lower development costs, more cost-effective retrofits, improved livability, and a better
natural urban environment for all to benefit from.
This chapter goes on to offer a set of cogent messages for water sensitive designers:
1. Water gardens need not be restricted to only high-visibility locations (such as
Portlands Water Pollution Control Laboratory), because they can also offer ecological and aesthetic benefits when placed in association with much more mundane
and utilitarian sites such as apartment blocks and service parking lots (in
Chapter I.7, Robert France and Philip Craul show designs for one such a green
parking lot).
2. As Bruce Ferguson stresses in Chapter I.1, designing in the absence of good, sitespecific hydrological knowledge (as, for example, the fact that most precipitation
falls in small storms) is unwise at best, and possibly foolhardy at worst.
3. Landscape architects need to do a much better job in undertaking experiments to
determine what attributes of which water sensitive designs work, not merely look,
best.
4. A great need exists for hubris to be replaced by humility in terms of honestly
sharing information and experiences about the lessons learned from various design
projects both successes and mistakes.
chapter I.7
Retaining water: Technical support

for capturing parking lot runoff
(Ithaca, New York)
Robert France and Philip Craul
Abstract
Water sensitive design, as part of a larger framework of integrated watershed management, is, due to its inherent complexity, a daunting and onerous task for anyone to engage
in independently. Just as water is an integrator across physical landscapes, it can also be
an integrator across professional landscapes. Because few individuals have the breadth of
ability to both fully comprehend and effectively design innovative stormwater management systems, complicated problems necessitate the formation of diverse, interdisciplinary
teams. This chapter briefly outlines some of the attributes suggested by a team of technical
experts (from the disciplines of hydrology and soil science) that should be considered in
the landscape architecture of a combined bioretention-wetland system designed to treat
parking lot runoff at the same time as providing a scenic amenity at a high-visibility
location.
Introduction
Remedying environmental development pressures
Sprawling patterns of community growth and transportation systems reduce vegetative
cover and produce large areas of impervious surfaces (roads, parking lots, driveways)
that distort watershed hydrology, and produce rapid, high-volume stormwater runoff
patterns (Horner et al., 1994; Chapter I.3). Impervious surfaces typical of urban and suburban development areas prevent the infiltration of rainwater into soils (Chapter I.1). As
a result, groundwater supplies cannot be recharged, streams degrade, and flooding is
increased, with associated habitat reduction (Schueler, 1995). As the runoff moves over
impervious surfaces, it collects contaminants, increasing pollution in streams and wetlands. In addition, impervious surfaces retain solar energy, raising air and water temperatures, which in turn negatively impact aquatic communities.
An increasing body of scientific research conducted in many geographic areas and
using varied techniques supports the theory that impervious land coverage can be a
reliable indicator of stream degradation (Richman et al., 1997; Chapter I.3). In particular,
1-56670-562-2/02/$0.00+$1.50
155
156
Bannerman and Dodds (1992) document the contribution of parking lots to runoff pollution problems, demonstrating that for commercial and industrial land uses, parking lots
are a critical source of stormwater pollution . . . account[ing] for approximately one-fourth
to two-thirds of the suspended solids, total phosphorous, total copper, and total zinc loads
in the commercial and industrial areas studied.
The first portion of a storm cycle has the greatest impact on water quality (Chapter I.6).
Rapid, small storms disperse the most highly concentrated contaminants (oils, metals, and
other toxic substance), thereby producing the most negative impacts. Therefore, treating
the first flush of runoff is key to controlling nonpoint source pollution. This implies that
the primary focus in stormwater management practices should be at the source of the
runoff. Supporting this is the fact that the more distant runoff treatment efforts are placed
from the source, the more effort in terms of cost and maintenance is required to operate
them (Richman et al., 1997).
Design strategies important to stormwater management include the following:
Minimization of directly connected impervious areas

Maximization of permeability
Employment of access streets
Plans for alternative modes of transportation
Integration of the drainage system(s) with natural landforms and topography
Directly connected impervious areas are those paved or roofed surfaces that drain
into a catch basin or other conveyance structure. If runoff is collected and concentrated
in a series of drainage structures as it is transported, no filtration by soil or organic matter
occurs. Additionally, the speed and volume of the water flow are increased, thereby
increasing erosion and flood potential, contributing a cumulative impact on stream systems. Subsurface flow and filtration of runoff are accomplished by the use of pervious
areas, depressions, and swales in conjunction with drains. Foregoing single-use drainage
ways in favor of multifunctional waterways is also recommended.
Maximization of permeable areas not only improves water quality, it also can significantly reduce development costs by eliminating or reducing the need for underground
conveyance stormwater systems. This can be accomplished with the use of permeable
pavement surfaces, reduction in the footprint size of new buildings, cluster development,
and shortened or shared driveways (Center for Watershed Protection [CWP], 1998;
Chapters I.1, I.3, I.5, I.15).
A number of so-called best management practices (BMPs) can be employed by landscape architects, engineers, and designers to minimize human impacts on the coastal
environment (Schueler et al., 1992; Chapter I.4). Frequently in the past, corrections for the
adverse effects of stormwater runoff on aquatic ecosystems were not accounted for in site
planning. It is now recognized that an often underused environmental tactic is the integration of drainage and filtration into the initial stages of the site design process
(Chapters I.5, I.15). Conventional storm drain systems divert water beneath the ground
surface and fail to integrate with surface topography. If instead, drainage and filtration
considerations are addressed in the initial planning process, then proper environmental
considerations can be integrated into the landscape in an aesthetically pleasing and economically sound fashion through minimizing expensive earthwork.
Techniques for managing stormwater drainage include
Extended detention ponds: These ponds detain stormwater runoff for a short
period of time, allowing pollutants to settle out.
Stormwater wetlands: These constructed, shallow pools support growth of wetland
plants and maximize pollutant removal through plant uptake.
Chapter I.7:
157
Infiltration swales and basins: Reservoirs are created out of shallow trenches that
are lined with porous material that enables the filtration of cleansed stormwater
into the water table.
Multiple-pond systems: These pond systems combine a number of pond designs
such as extended detention, permanent pool, shallow wetland, and infiltration,
providing reinforcement of pollution removal abilities.
Sand filters: This technique diverts the first flush of runoff into a sand bed, at
which time the filtered water is then collected in an underground drainage system
and conveyed back to a stream or water body. This method is particularly well
suited to treat parking lot runoff.
Grassed swales: Swales are frequently used for their pollution removal potential;
runoff is collected and filtered in concave earthen depressions.
Filter strips: These level vegetated strips of land area (or enhanced natural buffers)
intercept overland sheet flow running off from development areas. Although the
vegetative cover enhances pollutant removal, filter strips are not especially effective in treating high-velocity runoff and should be used either in conjunction with
other management techniques or in low-density development areas.
The best solutions for stormwater management in a given design project may combine
several of these categories, depending on the individual objectives and factors (EPA, 1990;
Schueler et al., 1992). For example, existing water quality, size of site, type of development,
economics, amount of time necessary for implementation, and limitations such as space
requirements, soil type, etc. often vary. While the long-term benefits of these strategies
are not known in detail due to their relatively recent invention, these practices are widely
regarded by the scientific community to be effective in reducing the environmental damage
caused by development (e.g., Horner et al., 1994; Schueler et al., 1992). One particular type
of stormwater management BMP beginning to receive attention, particularly in relation
to parking lot designs, is the use of bioretention swales.
Bioretention swales and green parking lot design

Our urban centers are increasingly planned and built for the benefit of cars, rather than,
and sometimes even at the expense of, people. For example, accommodating transportation-related activities can contribute up to three-quarters of the total impervious surface
coverage within urban watersheds (Richman et al., 1997; Prince Georges County [PGC],
1999), the remainder being associated with buildings. Perhaps no more substantial contributor to impervious surfaces exists than parking lots. Consequently, a literature is
rapidly developing that investigates procedures to reduce the amounts and mitigate or
alleviate the effects of stormwater runoff from parking lots (Schueler, 1995; Claytor and
Schueler, 1996; PGC, 1999; Chapters I.5, I.15).
Bioretention is a method to manage and treat stormwater runoff by using plant
materials and a conditioned planting soil bed to filter runoff temporarily stored in shallow
swales (PGC, 1993, 1999; Claytor and Schueler, 1996; Chapter I.5). Water purification occurs
through both physical filtering and biological (plant and microbe) uptake processes (PGC,
1993, 1999; Claytor and Schueler, 1996). Additional benefits include wildlife attraction.
Several projects in Maryland, Oregon, and Colorado have successfully employed bioretention swales in commercial parking lots (Thompson, 1996; Chapters I.4, I.5, I.15).
Although bioretention swales and retention basins are based on functions of natural
riparian and forest plant communities, they still require regular inspections and maintenance. They should be inspected on a semi-annual basis for the first year following all
major storm events, and subsequently by annual inspections (PGC, 1993; Claytor and
158
Schueler, 1996). In such cases, soil should be tested, erosion problems corrected, regular
mulching implemented, woody vegetation pruned, inflow areas checked for clogging, and
built-up sediments removed.
In addition to the goal of creating an effective stormwater management plan for
integrated parking lot design, other goals that are guiding principles in bioretention
include:
Decrease the amount of impervious surface on the site by increasing the amount
of greenspace.
Enhance the pedestrian character of the location.
Provide opportunities for visitors and residents to experience and learn of the
functioning of the natural environment.
Minimize the environmental impacts of the parking lot while providing the same
amount of parking.
Project site description

A site design proposed by Childs Associates for Cornell Universitys new Ornithology
Laboratory at Sapsucker Woods in Ithaca, NY, is based on an ecological approach in the
creation of stormwater wetlands and bioretention swales. Prior to establishment of the
Sapsucker Woods Sanctuary in the 1950s, the 12-acre site, including its created lake, barn,
and fire pond, had been a sheep farm. Today, the relatively flat site, with a perched water
table, has naturally evolved to a wooded wetland predominately made up of invasive,
nonnative species. A new building and accompanying parking lot for 190 cars is planned
(Ulrich, 2000), which will make a substantial intrusion on the site, both physically and
visually (Figure I.7.1). These built elements have the potential to significantly impact prime
open space along the lakeshore, increase site runoff dramatically, cut off views and vistas
of the lake, and reduce both the visual quality of the site and its use as a wildlife habitat
for birds.
The design concept for the site has been drawn from the natural topography and
spatial features (Childs Assoc., 2000). The new building and the parking areas are envisioned as islands floating within the wet woodland and meadow context of the site,
similar to the existing island in the middle of the abutting lake. The parking areas will
occupy discrete spaces or islands that are enfolded and absorbed into the larger natural
context of the site. Changes in vegetation and topography will be used to define the
parking islands. The perimeter of each island will have a dense vertical edge of
wetland trees and shrubs that will provide wildlife habitat and contrast to the vegetation
at the interior of the island (Childs Assoc., 2000).
The site plan includes the design of the vehicular and pedestrian circulation systems,
the creative management of stormwater runoff, and the design of soils and plantings for
wetland mitigation and bioretention areas. All the stormwater runoff from the building
and the parking areas will be filtered through bioretention swales that will run between
the designed islands (Figure I.7.2). Vegetation and soil construction in the bioretention
swales will remove contaminants from the parking lot runoff. The cleansed water will
then flow into a newly created wetland approximately 1.5 acres in size and will eventually
drain into the existing lake.
The concept of parking islands surrounded by wetlands represents an ecological
approach to design that utilizes strategies to direct and filter runoff into newly created
treatment areas that will double as wildlife habitats. The overall intention is to improve
both the visual and ecological values of the site. This design concept is further intended
to integrate the new, large laboratory building into its larger context and to express the
Chapter I.7:
159
Figure I.7.1 Schematic design plan showing intended site development. (Adapted from preliminary
design plans by Susan Childs Associates.)
Ornithology Laboratorys commitment to the conservation of natural ecosystems and

desire to add to the the complement of trails that traverse the 220-acre wooded setting
(Ulrich, 2000).
Wetlands
Wetland enhancement of the northern pond
Motivated by need for adherence to wetland mitigation laws to compensate for wetlands
unavoidably lost during site construction, the plan calls for enhancement of the farm pond
situated several hundred meters to the north of the development site in order to increase
its wildlife potential. For construction, the following principles from France (2002) are
suggested to achieve effective wetland design:
160
Figure I.7.2 Detailed schematic site design plan showing new ornithology laboratory building and
parking lot islands surrounded by bioretention swales and treatment wetland. (Adapted from
preliminary design plans by Susan Childs Associates.) Dark gray = shallow marsh; medium gray =
wet meadow; light gray = shrub/scrub wetland; and stippled area = wetland edges and island
uplands (see Table I.7.1 for plant list).
Grading
Shape and blend in with natural contours of the landscape; i.e., ridges should be kept
as promontories or islands and depressions removed to build embayments.
Soil should be saved for later shoreline sculpting such as island construction.
Once the new wetland edge has been shaped, the compacted soil should be broken
up before breaching the final edge to the existing pond.
Chapter I.7:
161
Soil
At least 4 to 6 in. of soil is required for shallow rooting aquatic plants.
Islands
They should be at least 4 m2 or 0.02 ha in size, located 15 m apart and separated
from the shoreline by at least 15 m of permanently flooded water.
Low, cross-shaped islands are ideal as waterfowl habitats because their irregular
shape and increased edge enable establishment of distinct territories, thereby
maximizing space utilization.
Orientation
In cold climates like Ithaca, NY, maximizing southern exposure through an
eastwest alignment can be beneficial for overwintering waterfowl using emergent vegetation as shelter.
Shoreline
Creating irregular shorelines can increase their length by up to 10 to 20% per unit
wetland area, which will provide benefits to waterfowl by increasing locations
for nesting and resting.
Slope
Shorelines should be shallowly graded (3:1 to 5:1) to provide easy access for
wildlife and to limit erosion caused by wave contact with the toe of the bank.
Planting
For areas greater than 1 ha (as in this case), it will be more economical to use a
hydroseeded mixture.
The seasonal timing for seeding is not particularly critical, though spring is to be
preferred.
Wetland creation at the southern construction site

The goals for creating this wetland are twofold. First, the area created will go toward
compensating for wetlands unavoidably lost during site construction. Second, this wetland
will provide water quality treatment by receiving stormwater from the bioretention swales
draining the parking lots and building roof, and in so doing, perform final polishing
before the stormwater is discharged into the lake. For construction, the following principles from France (2002) are suggested to achieve effective wetland design:
Marking, access, and excavation
Careful surveying is required to delineate location of the selected site in relation
to access and erosion control measures during construction.
Excavation should proceed from the center toward the intended land edge, with
soil saved for later shoreline and riparian landscaping.
Sidewalls
They should be sloped no steeper than 2 horizontal:1 vertical.
Site preparation
Once the wetland has been shaped and graded, the compacted soil should be
broken up before shallow flooding to enable the material to settle and a level
bed to become established.
Soil
At least 4 to 6 in. of soil is required for shallow-rooting aquatic plants.
Increasing the organic matter content of bottom sediments will help to bind contaminants.
Planting
Given the estimated size of this wetland (about 2 acres), it will be more economical
to hydroseed a large portion of it, using direct plantings as the core (in such a
case, it will be necessary to have a dry establishment period).
162

Residence time
This is in the order of about 1 week to ensure effective removal of most contaminants.
Length
About 100 ft is effective for removal of most contaminants.
Depth
For stormwater wetlands, effective allocation of total area should be in the order
of 50% shallow (0- to 1-ft depth) marsh, 30% deep (1- to 3-ft depth) marsh, and
20% deep water (3- to 6-ft depth) including open-water pools.
Deep zones should be arranged perpendicular to flow direction to limit the chance
that water can short-circuit the intended pathway.
Because a mixture of flooding regimes expands wetland functions, interspersion
of shallow and deep water areas will increase the efficiency of contaminant
removal and attractiveness for wildlife.
Slope
The slope should not exceed 0.5 to 1% in order to maintain depth differences of
less than 3 in. from inlet to outlet.
Steep vertical banks should be avoided for safety purposes.
Length:width ratio
Between 3:1 and 4:1 is most efficient for contaminant removal.
Water distribution
Use of multiple input points along the length (termed step-feeding) is a strategy
to distribute the contaminant load evenly.
The following design particulars are suggested for the southern stormwater wetland:
Sidewalls graded down to a V-notched central gully, which is 2 ft lower than the
parking lot elevation.
The bulk of the wetland (most significantly, the area upstream of the central
bioretention swale) will impound water to a depth of no more than 1 ft during
flood events and may therefore not need to be graded so deeply.
Three regions of greater depth (at least 1 ft below the summer watertable) that will
be year-round openwater areas; these three regions (two situated at the mouths of
the inflowing bioretention swales, the other sculpting the lakeshore inland) will
be shaped to receive, concentrate and then redistribute stormwater.
For marsh areas use a 4-in.-thick layer of organic bottom soil overlaying impervious
clay.
A 1% grade along complete length will ensure that the approximate 700-ft travel
distance is more than adequate for a water residence period of about 5 days
maximum (based on calculations of estimated flow velocity derived from the
hydraulic radius and Mannings surface roughness coefficient).
Hand plantings (emergents) will fringe all three openwater pools, whereas the rest
of the wetland will be seeded with a regional mixture; riparian landwater interfaces will have a variety of water-tolerant hardwoods.
Bioretention swales at the parking lot site

The goals of the bioretention swales are to provide the first line of defense for mitigating
contamination that will arise from parking lot runoff. Other attributes of such swales are
to convey filtered stormwater to the wetland where it may be detained and further treated,
to provide an aesthetic amenity, and to provide wildlife attraction. For construction, the
Chapter I.7:
163
following criteria from Prince Georges County (1993) are suggested to ensure effective
treatment of the parking lot runoff:
Grass filter strip
Provides vegetative pretreatment of sediment laden runoff sheetflow
A 1% minimum grade to 4:1 maximum grade
Between 20 and 100 ft wide
Bioretention swale size
Minimum area of 5 to 7% total drainage site
From 10 to 25 ft wide and at least 40 ft long
Ponding area
A maximum of 6 in., which drains within 4 days
Mulch layer
Functions as a contaminant filter, protects soil from desiccation, and serves as a
habitat for microorganisms
Organic material such as shredded wood chips or commercial compost
Depth of 3 in.
Planting soil
Major purpose is to sustain plant growth
Secondary purpose is to trap contaminants
Should be of a depth of 2 to 4 ft
Loamy sand mixture, etc., with no more than 25% clay content
Underlying 1-ft sand bed increases the infiltration capacity and provide aeration
for the plant roots
Plant material
Entraps contaminants through growth and ET
Designed to mimic a terrestrial forest communitys functions
Provides a host of other site amenities, including wildlife attraction
Hydrology
Installation of an overflow storm drain inlet (2 in. above max ponding depth) with
a simple piping system to ensure that excess water will be carried offsite and
thereby prevent any flooding of parking areas.
Bedding underdrains in pea gravel will assist site draining.
Planning for incorporation of bioretention swales for stormwater treatment necessitates estimation of the size needed to hold the quantity of runoff expected. We used the
following hydrological assumptions:
1. Common storm event
Total site area = 91,000 ft2
Existing RCN = 70
Proposed post-development RCN = 84
Design storm = 3 in. (2-yr, 24-h event)
Design water depth = 3 in.
Calculated runoff based on land characteristics is about 2 in. (assuming that no
pervious pavement is used)
About 24,000 ft2 of bioretention area would be required for water quantity storage.
As our planned site is to have 77,000 ft2 of combined bioretention and wetland area, runoff
from the design storm can spread out to a depth of about 1 in. This suggests that flooding
164
is unlikely to pose a problem. With this knowledge, it becomes possible to plan the
elevation of the new parking lot accordingly.
Calculations for water quality improvement found that a smaller bioretention area
would be required for runoff treatment. Therefore, the 0.2 in. of first-flush runoff (which
will contain the majority of the contaminants) can easily be treated in the combined
bioretention-wetland system area. For drainage estimations, the water residence time
(based on estimated flow velocity derived from hydraulic radius and Mannings surface
roughness coefficient, ET, and infiltration rates) was calculated as being less than 1 day
for the 2-yr, 24-h storm event.
2. Extreme storm event
Design storm 7 in. (100-yr, 24-h event)
Other details as described previously
About 64,000 ft2 of bioretention area would be required for water quantity storage.
Working backward using the available treatment area for the proposed site design found
that an absolute minimum of 2.5 in. of freeboard would be required to contain this runoff
(this may have implications for constructed soil depths).
The following design particulars are suggested for the bioretention swales:
The 50- to 55-ft sidewalls are graded down to a V-notched central gully, which is
3 ft lower than the parking lot elevation.
The central gully is widened by about 4 ft at the depth of 3 ft below parking lot grade.
An underdrain is installed ending in the wetland, which is surrounded by pea
gravel.
Above this is 6 in. of sand with a high infiltration capacity.
Above this, and extending out to the margins of the overall swale, is 1 to 1 ft of
planting soil.
A vertical pipe and overflow drain are installed at an elevation of 2 in. above max
ponding depth.
A 3-in.-deep mulch layer of woodchips is placed over the trough, blending into a
2-in. layer of compost spreading out the sloping flanks.
Plants
Given the high visibility of this site, the landscape architects paid careful attention (Childs
Assoc., 2000) to plant selection. Two criteria were integral in the selection of candidate
vegetation: First, for those species to be associated with the wetland and bioretention
swales, their ability to withstand inundation as well as provide phytoremediation of
contaminants were paramount. Lists for potential bioretention species from Maryland
(PGC, 1993) were compared with lists of native wetland plants for New England to select
those mesic species known to thrive in this north-temperate bioregion. Second, given that
the site is a center for professional ornithological study and recreational bird watching,
priority was given to species with known abilities to attract avifauna. A screening matrix
was developed to identify those species able to fulfill both selection criteria; a selection
of these species is listed in Table I.7.1. A comprehensive landscaping quality assurance
document was produced, covering such items as sources of trees and shrubs, their acceptable sizes, details about fertilizers, mulches, and erosion control materials, site preparation
and planting, and maintenance (Childs Assoc., 2000).
Chapter I.7:
165
Table I.7.1 Proposed Plant List for Wetlands and Bioretention Swales around the Parking Lot
Islands at the Cornell Ornithology Laboratory
1. Shallow marsh
Sweet flag
Water plantain
Bearded sedge
Tussock sedge
Turtlehead
Spike rush
Yellow water lily
Royal fern
Arrow arym
Pickerel weed
Northern arrowhead
Common three-square
Burreed
Skunk cabbage
3. Shrub/scrub wetland
Trees
Red maple
Green ash
American larch
Swamp white oak
Black willow
Small trees
Box elder maple
River birch
Shrubs
American cranberry
American cranberry bush
2. Wet meadow
Shrubs
Speckled alder
Red chokeberry
Buttonbush
Silky dogwood
Red oster dogwood
Winterberry
Spikebush
Swamp rose
Pussy willow
Common elderberry
Ground covers/perennials
Marsh marigold
Fox sedge
Joe pye weed
Blue flag
Cardinal flower
Sensitive fern
Cinnamon fern
Fowl bluegrass
Narrow-leaved cattail
Common cattail
Blue vervain
4. Wetland edges/upland (island edges)

Trees
Red maple
Black gum
White pine
Eastern hemlock
Small trees
Shadblow
Gray birch
Cockspur hawthorn
Black cherry
Shrubs
Sweet pepperbush
Gray dogwood
Witch hazel
Bayberry
Highbush blueberry
Note: Additional plants were proposed for the upland parking lot islands (contact Childs Associates).
Source: Adapted from Susan Childs Associates, Boston, MA.
Soils
The general situation
Unless careful attention is paid to soil creation, landscape projects will be only marginally
successful in terms of sustainability (Craul, 1992, 1999). The major portion of the construction
166
for the new Cornell Ornithology Laboratory and its associated structures occurs in a
recognized wetland environment. It is therefore necessary to construct a series of bioremediation areas among the structures and parking areas. Their purpose is to remove the
various pollutants contained in the runoff from rooftops and parking lots before it enters
the natural drainage system. To ensure success in the construction of bioremediation
wetlands, greater attention must be given to the details of soil and plant specifications
and installation than in routine landscape projects. That process is described next.
The soils of the bioremediation site

The soils of the site are identified as the Erie-Ellery channery silt loam complex and the
Ellery-Chippewa-Alden channery silt loam complex in the Tompkins County Soil Survey
Report (Neeley, 1961). The Ellery series has since been merged into the Chippewa series
and is not discussed further.
The Erie series consists of very deep, somewhat poorly drained soils formed in
loamy till. They have a fragipan layer starting at depths of 10 to 21 in. below the
soil surface. These soils are of uniform slope and are on footslopes and broad
divides in glaciated uplands. Permeability is moderate above the fragipan and
slow in the fragipan and substratum.
The Alden series consists of very deep, very poorly drained soils on upland till
plains in depressions and low areas in the landscape. They are formed in a thin,
silty colluvial mantle overlying glacial till. They are on upland till plains in depressions and low areas in the landscape. Slopes range from 0 to 3%.
The Chippewa series consists of very deep, poorly drained and very poorly drained
soils formed in compact till deposits. These soils are in upland depressions. A
dense fragipan is present in the subsoil. Permeability is moderate above the fragipan and slow or very slow in the fragipan and substratum. Slope ranges from 0
to 8%.
Two soil pits were excavated on January 15, 2000, to confirm the soil conditions.
Supplemental piezometers were installed to refine the determination of the fluctuating
perched water table present in these soils during the dormant season (it may be present
in the growing season during wetter-than-normal years). The brief soil profile descriptions
are given in the list that follows.
Soil pit 1:
Ap-0-7 in.; very dark brown mucky silt loam; weak, medium granular structure;
friable; common fine roots
E-7-16 in.; brown silty clay loam; weak, fine subangular blocky structure; few
medium, faint-yellowish brown mottles; very friable; many fine roots
Bg21-16-25 in.; yellowish-brown silty clay loam; weak, coarse subangular
blocky structure; with common medium, distinct gray mottles; friable; water
drainage at roots and their channels
Bg22-25-28 in.; yellowish-brown silty clay loam; weak, medium platy structure;
common medium, prominent gray mottles; firm; no roots
Bg3-28-37 in.; yellowish brown silty clay loam; moderate to strong, medium
platy structure; very common medium, prominent gray mottles; very firm; no
roots
Cg-37-52+ in.; gray loam; massive and firm; no roots
Chapter I.7:
167
Soil pit 2:
Ap-0-8 in.; dark grayish-brown silt loam; weak, fine granular structure; very
friable; common fine roots
Bg21-8-16 in.; yellowish-brown silt loam; weak, medium subangular blocky
structure; many coarse prominent gray mottles; firm; few fine to no roots
Bg22-16-23 in.; yellowish-brown silty clay loam; weak, medium platy structure;
many coarse, prominent gray mottles; firm; no roots
Bg3-23-30 in.; yellowish-brown silty clay loam; moderate to strong platy structure; many coarse, prominent gray mottles; firm; no roots
Cg1-30-42 in.; yellowish-brown loam; very firm breaking into strong medium
platy peds; many medium distinct mottles; drier than above horizons; no roots
Cg2-42-47+ in.; yellowish-brown loam; weak to moderate medium platy structure; common medium distinct mottles; firm to friable; no roots
The implications of these soil conditions for a bioremediation site are:
The native soils are already classified as wetland soils being poorly to very poorly
drained.
The subsoil horizons of these soils contain either a fragipan (a dense layer impermeable to water which perches water at a shallow depth) or firm horizons that
have slow permeability. This is also a disadvantage, however, in that it may cause
ponded water at high elevation during wet seasons, requiring structural works to
be elevated or filled to higher than desired final grade.
The vegetation present on the undisturbed portions of the project site is of wetland
species composition and does not require significant modification to meet wetland
criteria.
The designed soil profiles

Grass and vegetation filter strip soil profile
This profile (Figure I.7.3) occurs over the sculptured native subsoil, but generally
above the maximum high perched water level or the shallow ponding area
level. It extends from the gravel parking lot to the bioretention areas.
The profile consists of a topsoil horizon and a subsoil, both of which are derived
from other soil materials. The specifications are given in the next section.
The native subsoil (that below the topsoil, the former plow layer, of 7 to 8 in.) is
sculptured so that a 1% grade is created extending from the longitudinal centerline of the parking area to the edge of the bioretention area or the approximate elevation of the shallow ponding area within the bioretention area.
The subsoil depth is given as residual because the total depth of the soil profile
will decrease approximately 2% from the appropriate subsoil depth at the
edge of the parking area to 0 at the edge of the bioretention area.
The bioretention area and adjacent filter strip soil profile
This profile (Figure I.7.4) is constructed as the rooting soil for the bioretention
plants in the loamy sand planting bed and the adjacent filter strip on either
side. The bioretention area will be saturated through most of the dormant
season and will have shallow ponding during much of that season.
It is expected to dry out in the growing season, especially during the latter portion
of July and August, and early September in dry years. Normally, it will become
saturated by November.
168
Figure I.7.3 Schematic of the grass and vegetation filter strip soil profile.
Figure I.7.4 Schematic of the bioretention area and adjacent filter strip soil profile.
The profile consists of a coarse mulch layer over a specified loamy sand planting
bed.
The profile vertical dimension is provided by the excavation of the natural subsoil
to the specified depth. The soil and mulch layer are installed in the excavation.
The bottom of the planting bed trench is graded longitudinally at 1% to facilitate
saturated flow through the bioretention trench and eventually to the outlet.
The trench bottom is also tapered or V-notched to the centerline in cross-section,
at a right angle to the longitudinal grade, to facilitate saturated water flow to
the center of the trench. The depth of the trench is 24 in. throughout.
The soil horizon specifications

The grass and vegetation filter strip topsoil
The soil of the filter strip should aid in reducing the runoff velocity and filter particulates from the parking area runoff. The topsoil must have the characteristics of
Chapter I.7:
169
Table I.7.2 Range in Percent Passing

Sieve Sizes for the Bioretention
Planting Bed Loamy Sand Soil
Sieve Size
10
18
35
60
140
300
Silt
Clay
Percent Passing Range

100
88100
7080
2535
1323
1020
818
212
high infiltration rate, high permeability, relatively high organic matter, and sufficient depth, coupled with the subsoil, to support vigorous growth of plants.
The existing topsoil (former plow layer or Ap horizon) on the construction site is
to be stripped and stored for use as a soil mix material.
The existing topsoil has a texture of silt loam, which renders it unsatisfactory for
the grass and vegetation filter strip topsoil. It must be modified to at least loam
(40% sand) or coarser texture by the addition of a coarse sand. A sand content
of 60% or more may be necessary to offset the silt plus clay content of the
topsoil. The existing silt loam soil must be tested for particle size distribution
(ASTM D-422-63) to determine the appropriate sandsoil mix ratio. A probable
ratio may be 2:1. Percolation and texture tests will be required for the final mix
and must be approved by the landscape architect.
The sand specified for the bioretention sand bed may be used as the coarse sand
amendment to the existing topsoil.
The final soil mix should be free of any stones, gravel, glass, plastic, masonry,
drywall, asphalt, rubber, metal, or any other type of foreign debris.
The grass and vegetation filter strip subsoil
The soil material employed as the filled subsoil in the grass filter strip must have
moderate permeability and fertility with adequate organic matter content and
sufficient strength to support both plants and the topsoil.
No acceptable material exists on site for use as the subsoil for the grass and
vegetation filter strip areas. Therefore, a soil material must be imported.
The unamended loamy sand specified for the bioretention area planting bed may
be used as the grass and vegetation filter strip subsoil.
The final soil should be free of any stones, gravel, glass, plastic, masonry, drywall,
asphalt, rubber, metal, or any other type of foreign debris.
The bioretention area mulch layer
Chipped or shredded and screened clean hardwood mulch, with 100% passing a
-in. screen. It should not contain any debris such as stones, gravel, masonry,
drywall, glass, plastic, asphalt, rubber, metal, or any other type of foreign
debris.
The bioretention area planting bed soil
The planting bed soil must be a sandy soil that exhibits a high infiltration rate (at
least 2 in./h).
Texture: the specified particle size distribution is given in Table I.7.2.
Organic matter content: the planting bed soil should have an organic matter content
of about 2% by weight and may range from 1.5 to 3.0% (approximately 5% by
170

Table I.7.3 The Particle Size Distribution of the
Bioretention Area Sand Beda
Sieve size (no.)
Percent passing
Size content, %
100
90
70
21
13
10
4
10
20
49
8
3
6
4
10
18
35
60
140
300
Silt
Clay
a
Actual values may be 2.5%.
Table I.7.4 Particle Size Distribution of

the Gravel Drainage Layer (AASHTO #8)
Sieve Size (in. or no.)
in.
in.
#4
#8
#16
Percent passing
100
85100
1030
010
05
volume). Many natural sand deposits will have an organic matter content
approximating this percentage. Fertilization may be necessary to meet the
nutrient requirements given below.
Others: the pH range for the planting soil should be 5.56.5; magnesium content:
35 lb/acre; phosphorus content: 75100 lb/acre; potassium content: 85 lb/acre;
soluble salts: not to exceed 500 ppm.
The soil bed should be free of any stones, gravel, glass, plastic, masonry, drywall,
asphalt, rubber, metal, or any other type of foreign debris.
The bioretention area sand bed
This horizon is primarily a sand and acts as an additional facilitator to water
drainage and treatment.
The texture of the sand bed should be a sand with the particle size distribution
given in Table 1.7.3.
It should be free of any stones, gravel, glass, plastic, masonry, drywall, asphalt,
rubber, metal, or any other type of foreign debris.
The bioretention area gravel drainage layer
The drainage layer is primarily to aid drainage of excess overflow. The 4-in.
perforated PVC pipe with an overflow standpipe prevents high-water levels
from reaching the parking lots and associated areas during severe, intense
storms. The material is commercial pea gravel, which may be #7 or #8 by
AASHTO designation for coarse aggregates. The particle size distribution for
#8 is given in Table I.7.4.
Installation and sequence of the soil

Installation and the sequencing of the soil work involves some details that require attention
to ensure that the soil is placed properly:
Chapter I.7:
171
Soil must be placed in lifts not exceeding 12 in. in thickness.

Work should begin with excavation of the bioretention trench, laying of pipe, and
backfilling. Work should proceed outward from the trench to the parking surface.
Minimum compaction is achieved by routine travel over the soil by lightweight soilhandling equipment. Excessive compaction is to be avoided by repeated traffic
over the same area or by frequent turning. Multiple, offset linear travel patterns
provide minimum desirable, but not excessive, compaction.
Soil must not be handled, placed, tilled, or traveled over when it is wet or the moisture
content exceeds field capacity (the water in the soil voids glistens in light on close
examination and/or the soil exhibits thixotropy when shaken).
The bioretention mulch is best placed after the planting of trees and shrubs has been
accomplished. Any ground cover specified as plugs may be installed once the area
has been mulched. Ground cover established by seeding or consisting of grass
should not be covered with mulch.
It is necessary to sequence the installation of sediment and erosion controls to minimize the contamination of the planting soil with silts and other fines (clay and
organic matter). Silts and fines will have a tendency to clog the planting soil and
impair the functions of the bioretention area.
Sediment controls are installed around the area to be disturbed before grading and
around the bioretention area before the excavation of the trench for the planting
soil.
Post-installation maintenance
Post-installation maintenance of the bioremediation areas is necessary for successful continued functioning:
Biannual mulching is recommended. The previous mulch may be removed and discarded to an appropriate disposal area or retained if it is decomposed. The mulch
is replaced to a depth not to exceed 3 in.
Dead or poor, low-vigor plants and trees within the bioremediation area should be
promptly replaced as part of a comprehensive landscape maintenance program.
The operation is necessary to maintain the function of the bioremediation area.
Lack of growing plants can greatly reduce the effectiveness of the bioremediation
area for which it is designed. Continuous inspection for insect and disease outbreaks or physical damage or injury to the vegetation should be a routine part of
the maintenance program.
Any obstructions to drainage of the bioremediation area should be promptly eliminated.
Damage or disturbance to the bioremediation area soils by burrowing animals should
be immediately rectified and the animals eliminated or removed from the area.
Education
Wetlands, once referred to as wastelands, are now regarded as important centers of
biodiversity, as purifiers of contaminated water, and as regulators of watershed hydrology.
Recently, wetlands have received increased recognition because, given all these accredited
roles in addition to their admired beauty and attraction for recreational purposes, they
can also serve as important vehicles for environmental education (France, 2002). The recent
book published by the Association of State Wetland Managers and the EPA, titled Guidebook
for Creating Wetland Interpretation Sites Including Ecotourism, is witness to the fact that
172
wetland interpretation sites are invaluable to secondary and university science and
ecology educational efforts (Kusler et al., 1998).
In addition to the obvious opportunities for passive education in the form of interpretive signage, etc., at least two ideas are worthy of serious consideration for formal
research projects concerning wetland function. Both of these research projects could be
easily initiated and implemented by wetland scientists at Cornell University and run by
volunteers and summer students. The purpose of this section is to introduce these two
research topics as potential elements to be incorporated into the education program at the
Ornithology Laboratory, not to develop the actual scientific sampling programs necessary
for their successful implementation.
Wetland enhancement of northern pond effects on avian biodiversity

As part of adherence to wetland mitigation laws to compensate for wetlands unavoidably
lost during site (mostly building, not parking lot) construction, the plan is to enhance the
existing farm pond in order to increase its wildlife potential. An oft-acknowledged need
exists for more information to be made available about how the physical conversion of
farm ponds to wetlands will increase the attraction of these improved habitats for wildlife.
With this in mind, comparative bird counts of species abundance, visitations, and nesting
sites could be conducted before and after enhancement.
Parking lot development and bioretention facilities effects on vernal pool

hydrology and biointegrity
Whereas most wetlands are important from a landscape-biodiversity perspective, vernal
(temporary) pools have been recognized as being of extreme significance in this regard
due to the absence of predaceous fish and consequent inhabitation by amphibians.
Although the design plan calls for avoidance of one such critical vernal pool habitat in
the careful placement of the parking lots and bioretention swales, no scientific information
is available as to how wide the protected vegetated buffer area should be in such circumstances. With this in mind, comparative amphibian counts could be made before and after
site development for several years. In addition, because it is the hydrology of vernal pools
that is so critical to their continued existence, several piezometers could be installed in
order to monitor water level fluctuations before, during, and after construction of the
nearby parking lots.
Acknowledgments
This landscape architecture project is currently being designed by Antonia Bellata Osborne
and Susan Child of Childs Associates, whom we thank for allowing us to participate in
the conceptual design phase as well as providing support documents for this technical
paper. Lisa Cloutier aided in writing a portion of the introduction, and Matthew Tucker
prepared the figures.
Literature cited
Bannerman, R. and Dodds, R., Sources of Pollutants in Wisconsin Stormwater, Wiscon. Dept. Natur.
Resour., Madison, 1992.
Center for Watershed Protection, Better Site Design: A Handbook for Changing Development Rules in
Your Community, Center Watershed Protect., MD, 1998.
Childs Associates, Site work quality document for the Cornell Laboratory of Ornithology, 2000.
Chapter I.7:
173
Claytor, R.A. and Schueler, T.R., Design of Stormwater Filtering Systems, Center Watershed Protect.,
MD, 1996.
Craul, P., Urban Soil in Landscape Design, John Wiley & Sons, New York, 1992.
Craul, P., Urban Soils: Applications and Practices, John Wiley & Sons, New York, 1999.
EPA, Urban Targeting and BMP Selection. An Information and Guidance Manual for State Nonpoint Source
Program Staff Engineers and Managers, U.S. EPA, Chicago, 1990.
France, R., Designing Wetlands. Principles and Practices for Landscape Architects and Land-Use Planners,
W.W. Norton, in press, 2002.
Horner, R.R. et al., Fundamentals of Urban Runoff Management: Technical and Institutional Issues, Terrene
Inst., Washington, DC, 1994.
Klute, A., Ed., Methods of Soil Analysis, Parts 1 and 2, Monograph no. 9, 2nd ed., Amer. Soc. Agronomy,
Madison, WI, 1986.
Kusler, J.A. et al., Guidebook for Creating Wetland Interpretation Sites Including Wetlands and Ecotourism,
Assoc. State Wetland Manag., New York, 1998.
Neeley, J.A., Soil survey of Tompkins County, New York. USDA-Soil Conservation Service and
Cornell Agricultural Experiment Station, Sup. Docum., U.S. GPO, Washington, DC, 1961.
Prince Georges County Maryland, Dept. Environ. Resour., Design Manual for Use of Bioretention in
Stormwater Management, Prince George County Publ., MD, 1993.
Prince Georges County Maryland, Dept. Environ. Resour., Low-Impact Development Design Strategies.
An Integrated Approach, Prince George County Publ., MD, 1999.
Richman, T. et al., Start at the Source. Residential Site Planning & Design Guidance Manual for Stormwater
Quality Protection, Tom Richman Assoc. Publ., CA, 1997.
Schueler, T., Site Planning for Urban Stream Protection, Center Watershed Protect., MD, 1995.
Schueler, T., Kumble, P., and Heraty, M., A current assessment of urban best management practices.
Techniques for reducing non-point source pollution in the coastal zone, U.S. EPA, 1992.
Thompson, J.W., Let that soak in, Land. Arch., 11, 60, 1996.
Ulrich, C., Building on the past. A lab for the 21st century, Living Bird, 19, 24, 2000.
174
Response
To build an oxymoron:
A green parking lot
For a landscape architecture firm engaged in a stormwater management project, recognizing
the point when outside consultation by scientists is required, is an important realization that
the best solutions to environmental problems result from the marriage of art and science
together. As quoted by C. Caudwell in Garrett Ekbos seminal Landscape for Living, Art is
the science of feeling, science is the art of knowing. We must know to be able to do, but we
must feel to know what to do. This chapter by Robert France and Philip Craul, similar to
those by Richard Pinkham and Timothy Collins (Chapter I.4), Thomas Liptan and Robert
Murase (Chapter I.6), Robert France and Kaki Martin (Chapter I.12), and Nicholas Pouder
and Robert France (Chapter I.16), demonstrates the strength shown by successful pairing
between disciplines. This echoes the call for greater interdisciplinary design in Chapter I.17
by Wendi Goldsmith.
Three important messages are put forward in this chapter. The first is the detailed
presentation of the mostly descriptive, but also rudimentally quantitative, methodology
needed to assess the potential for any site of applying bioretention techniques, an important tool of the emerging paradigm of low-impact development as discussed elsewhere
in this book (e.g., Chapter I.1 by Bruce Ferguson, Chapter I.5 by Larry Coffman, Chapter I.6
by Thomas Liptan and Robert Murase, and Chapter I.15 by Michael Clar). The complexity
of these issues is what may, depending on the availability of in-house expertise, necessitate
scientific consultation, as was the situation for this project led by S. Childs Associates.
The second take-away message from France and Crauls work is recognition that a
green parking lot is not really as oxymoronic as it sounds. We have to abandon the image
of parking lots as mere functional appendages to development projects and instead begin
to look upon them as worthy extensions of the aesthetically designed landscape. This,
more than perhaps anywhere else, is the nexus where science and art can mesh for effective
stormwater management. Given the high visibility of some building sites, such as the
ornithology laboratory described in this chapter, it is unrealistic to expect them to be
surrounded by ugly and environmentally dysfunctional fields of concrete when so much
more can be accomplished.
The last, and perhaps most important, message developed by this chapter is the
attention France and Craul give to assessing existing soils along with developing recommendations for constructing new soils. More than any other chapter in this book, with the
possible exception of that by Wendi Goldsmith (Chapter I.17), it educates us about the
need to remember that, in order for water sensitive design to work effectively, careful
consideration must be given to the interactions with water upon the physical landscape
of the soil.
chapter I.8
A productive stormwater park

(Farmington, Minnesota)
Diana Balmori
Abstract
The Farmington, MN, waterway provides drainage for a suburban development of nearly
500 homes approximately 25 miles from downtown Minneapolis. It is an open water
drainage system, replacing the usual underground pipe line that frequently floods streams
with its sudden discharges. In addition, the waterway provides 91 acres of planted green
space, as well as a system of connected streams that widen into larger ponds for recreational use. A park, by becoming a drainage system, acquires a function that justifies its
creation.The idea of establishing such a productive park could be implemented in many
communities. This chapter describes the design decisions and goals involved in winning
city planners approval for the project. An appendix suggests vegetation compatible with
climate and local conditions in addition to soil and water regimes.
Introduction
Suburban development produces problematic amounts of water runoff from the large,
impermeable surfaces of its new roads, roofs, and driveways (Chapters I.1 and I.3). This
water, dumped all at once into existing rivers and streams, produces sudden floods that
erode riverbanks and can potentially harm plant and aquatic life.
The Farmington, MN, waterway was designed to run alongside a suburban housing
development of 486 midpriced homes built by the Sienna Development Corporation of
Minneapolis. The project grew out of our desire to try a different approach to standard
drainage systems. The Farmington waterway was created as a drainage system alternative
to the usual underground straight-line water pipe. It functions successfully as a drainage
project, as shown in the recent peak flooding years of 2001 and 1998, when the Mississippi
River crested and remained above its floodplain for at least three months in each of those
years. In addition, and perhaps more important, the project creates public space a
park as part of the drainage solution. Because the waterway project makes the green
space useful, thus paying for itself, we call it a productive park.
1-56670-562-2/02/$0.00+$1.50
175
176
Figure I.8.1 The Farmington, MN, site in winter. (Photo by Tom Hammerberg. Copyright Balmori
Associates.)
The site: Midwest prairie

The area near Farmington is completely flat (Figure I.8.1). In winter months the temperature often dips to well below zero. A slight covering of snow seems glossy, reflecting the
low sun of winter. Around the site, a few farmhouses surrounded by acres of farmland
and fields of dry cornstalks stand in the snow. On the edge of the planned development,
a northsouth road bisects the rural terrain. On the opposite side is the town of Farmington,
which has a population of 12,500. According to local residents, other nearby developments
flood every spring.
Farmington, a small farming town, is about 25 miles northwest of Minneapolis, close
enough to the nearby urban area to create a high demand for housing. Rod Hardy, director
of the Sienna Development Corporation, proposed building a development of some 500
houses over the course of 10 to 15 years, with the initial phase totalling approximately
170 houses. In 1993, several months after the summer of the worst flooding of the Mississippi
River in 28 years, William Moorish, director of the Center of American Urban Landscape at
the University of Minnesota, proposed to the town that we would design a different kind
of drainage system for the planned Farmington development, which the town had not
approved at that point. The developers initial proposal was standard, directing the
increased runoff into a single pipe, which would then empty into the nearby Vermillion
River and continue on into the Mississippi River, about 20 miles from Farmington.
Additional water from an increased number of impermeable surfaces in the new
development would no doubt exacerbate the already dangerous risk of flooding. Hydrologists, landscape architects, and ecologists have suggested alternate drainage schemes,
such as trying to slow the runoff among vegetated edges to clean it gradually and, in large
storms, store it until the peak of the storm has passed. When the moment comes for town
boards or city planners to approve a drainage system, however, the water goes into pipes
more often than not due to the common and understandable reluctance to try something
new and unproven. In addition, such a slow-down system of managing runoff would
probably require using more land, which a development corporation is not likely to
purchase. Funding the project thus becomes a concern.
Chapter I.8:
A productive stormwater park (Farmington, Minnesota)
177
On the other hand, what if the space around a body of water or wetland system
became public space, an amenity for the whole town? William Moorish of the University
of Minnesota suggested just such a proposition to the town board in Farmington. Happily,
the small farming town had a capable town planner who supported the idea. Because
new solutions require broad training and a willingness to rethink old problems within
new constraints, Yale University hydrologist Paul Barten was added to the project team
and charged with exploring alternatives to standard drainage systems. Although little
money was available to carry out our drainage solution ideas, the Farmington prairie
waterway project was one way of trying out new ideas in a real-world setting. The town
of Farmington and the Sienna Development Corporation eventually provided funding.
The municipal approval process

The town of Farmington made the development of an innovative drainage system a
condition for approval of the development plan. The town planner supported the idea of
a prairie waterway/park throughout the lengthy approval process, and the town board,
mostly comprised of local farmers, did not object, but expressed some degree of apprehension over the need to buy land exclusively for the water system. The land was purchased with city money recovered through taxes paid on the property once the houses
had been sold (tax increment financing). The taxes were established by the town as
contributions by the developer to provide needed facilities, but the city also had to buy
additional land beyond the development to allow access for the water to flow to the
Vermillion River.
The design process and landscape aesthetic

We presented two schemes for a surface water drainage system to the Farmington town
board and the developer. The first scheme, which the town board and the Sienna Development Corporation much preferred, made the waterway a naturalistic stream
(Figure I.8.2). The second scheme, which we preferred, created a series of geometric interconnected ponds (Figure I.8.3) that cleaned and retained the water based on the system
used in cranberry bogs. The town board and developer raised objections to the gridded
bog system, claiming that it seemed unnatural. The ideal of the natural in an artificial
drainage system comes from the 18th-century Romantic landscape aesthetic. Because it is
important in any new work to find an aesthetic appropriate to it, the use of a well-known,
human-created system, such as the time-tested system of cranberry bogs, seemed appropriate to the site and the project requirements. Creating an appropriate aesthetic was not
one of the primary concerns of either the town or the developer, however, and truthfully
belongs instead to the aesthetic concerns of the designer. Still, being able to implement
our stream-like scheme was also acceptable to us, especially because convincing the town
and the developer to accept an open drainage system was a big step forward.
Our goals, as presented to the Farmington Town Board and the Sienna Development
Corporation, included the following:
Design the waterway as a civic amenity that would contribute to the long-term
prosperity of the city of Farmington, MN.
Recognize that the waterway is a created landscape, not a natural one, and that
its form can announce it as a cultural endeavor shaped by natural systems.
Create better drainage for the watershed.
Give the residents of the development, which they now call Park Place, a viable
outlet for basement water.
178
Figure I.8.2 The winning scheme made the waterway a naturalistic stream. (Drawing by Bill Coyne
and Joseph Marek. Copyright Balmori Associates.) (Color version available at www.gsd.harvard.
Help manage the housing growth of the city of Farmington in ways that would benefit
the developer, the current community, and the future residents of Farmington.
Create a stormwater pond and wetland system for the proposed Planned Unit of
Development (PUD) and keep floodwater from impacting nearby housing.
Create a design that has high ecological integrity and becomes a place of high
biological diversity.
Minimize construction and maintenance costs.
Design for the recreational needs of the community of Farmington.
Chapter I.8:
179
Figure I.8.3 The scheme that was not chosen created a series of geometric interconnected ponds based
on the system used in cranberry bogs. (Drawing by Bill Coyne and Joseph Marek. Copyright Balmori
Associates.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Design changes to the developments layout

The developments layout was modified in order to reduce the stormwater runoff
(Chapters I.5 and I.15). Street widths were narrowed, driveways were shortened, and at
180
Figure I.8.4 The park drainage system with an overview from Spruce Street, the main avenue. A
new development is on one side and agricultural fields border the other side. (Drawing by Fran
Leadon. Copyright Balmori Associates.)
the center of each block a depression was made to gather the water from spring rains
running off the roofs and lots. These latter areas at the center of each block were thickly
planted with trees (e.g., red maples) that can survive with roots in standing water for part
of the year. These trees, through their foliage (evapotranspiration), would get rid of the
water in these depressions.
Planting turned out to be the most troublesome part of the project. Basically, neither
the developer nor the city had sufficient funds to allow planting the center of the blocks
or the public green space created by the new water system. We proposed several inexpensive solutions, including gathering seeds from local vegetation and subsequent planting of those seeds by local schoolchildren as part of their science curriculum. We also
proposed that the Town Board contact the Minneapolis Department of Transportation
(MIN-DOT) to request permission to reuse plants that were being removed to widen roads
after discovering that there was just such a project not too far from the site. A cover of
winter rye was put in to control erosion. Eventually the town implemented a 10-year plan
for investing $10,000/year in small trees. As of this writing, four years of planting (largely
willow, hackberry, plum, and ash trees) have taken place.
The park
The park-like landscape for the growing city of Farmington consists of a riparian waterway, a civic lawn on axis with the downtown area, and playing fields. The wetlands
associated with this urban waterway provide a recreational component for those interested
in observing wildlife or using green public spaces. In addition, of course, they store
stormwater. The form of this landscape is revealed through the rise and fall of water levels
during and after storms (Figure I.8.4).
At the end of Spruce Street, the main thoroughfare in this development, a civic
connection is made by a semicircle of upland lawn, which is terraced down to a pond
Chapter I.8:
181
Figure I.8.5 Water movement in the drainage system. Hatched areas are locations for storing and
delaying entry of water. (Drawing by Fran Leadon. Copyright Balmori Associates.)
allowing views of the waterway, as well as access to it in both directions. Viewers can
watch the moon rise over the farm fields from this curved terrace. Tree rows line the streets
of the housing development and continue into the park, establishing a series of rooms
roughly the same size as the city blocks (i.e., about 400 ft2). Reminiscent of windbreaks,
these frames give way on the inside of the rooms to native and naturalized plants (see
list in Appendix A) that are determined in their distribution and type by the specific
conditions in which they grow. Between each pair of tree rows, separating the planted
rooms in much the same way that the roads separate the blocks of the housing area, is
a roadway of tall grasses that provides shelter to the small animals that move through
this landscape. Certain areas of the park were kept isolated to encourage a wildlife habitat,
but pathways and boardwalks allow access to many parts of the site, including one of the
wetlands. An 8-ft trail surrounds the green system, runs by, and crosses the waterway.
The waterway
Three separate water systems are used in the Farmington prairie waterway (Figure I.8.5):
1. A swale system to treat the stormwater runoff from the new housing development
that lies immediately to the west of Park Place
2. A pond and channel system designed to move and clean the water over a large
area before it enters the Vermillion River
3. Type I (periodically inundated) and Type II (sedge meadows) wetlands intended
to make up for wetlands lost to the proposed development and to provide temporary storage for elevated groundwater levels
Inputs: Sources of water

Two housing developments to the south of the Park Place development in Farmington, which
are known as Castle Rock and Henderson, will also be served by the drainage system because
their stormwater will enter the pond at the southern end of the Park Place site. Both new
developments are reported to have drainage problems; basements in the homes flood each
spring. This pond reduces the velocity and turbulence of the stormwater and settles out the
coarser particles. From here, the water continues to a larger pond that is surrounded by an
aquatic bench of emergent wetland plants 10 to 15 ft wide with a depth from 6 to 12 in.
The channel leaving the pond has the sinusoidal curved edges produced by running
water. The channel also slows the water, giving it greater opportunity to become cleansed as
it moves through the vegetation. The great conductivity of the local soils means that groundwater rises and falls quickly during and after storm events. Under these circumstances, the
great width of the channel, which varies between 80 and 110 ft, helps to reduce the velocity
182
Figure I.8.6 First stage of the housing development at Park Place, made up of 167 units. (Photo by
Charles Tooker. Copyright Balmori Associates.) (Color version available at www.gsd.harvard.
and scouring of the waterway. The vegetated, expanded streambed traps particulates and
takes up nutrients.
Finally, runoff from the Sienna Development housing itself is sent into stormwater
swales just east of the neighborhood. These wide stormwater swales are set above the
groundwater level and are temporarily inundated after a storm. The water moves very
slowly southward over sedges and grasses that help to clean the water before it subsides
or moves north through subsurface flow into the channel and on to the Vermillion River.
This doubling-back of the water flow gives increased residence time, and therefore
increased treatment, before the water enters the channel. During periods of base flow, the
channel has a small amount of water in it. During medium-sized storms, the channel
water rises to fill a shelf on either side of it. The stormwater swales become inundated
from both road runoff and the temporarily elevated groundwater. The water remains
below the level of the recreational pathways and the upland edges of the park. In addition,
the water has not spread overland into the mitigating wetlands.
During major storm events, the wetlands at the south end of the site accommodate
overflow from the ponds. The mitigating wetlands to the north temporarily store the
higher groundwater, but only in 100-year events is the berm between the wetlands and
the channel breached. The berm separating the stormwater swales from the channel is set
at an elevation that will keep them separate even during the 100-year event. On September
15, 1998, just as the construction of the drainage system was completed, a 100-year flood
event took place. The drainage system worked exactly as it was supposed to, the berm
was breached, and the mitigating wetlands stored the higher waters.
The Farmington housing development, started in 1990, was intended to be built in
three 5-year phases. But the project was so successful, the built houses sold out quickly
during the first phase of 167 houses. By 2000, seven phases totalling 486 new houses had
been completed. An interesting footnote: the third phase was partially purchased by the
builders of the project. Aerial photographs show the three stages of building (Figures I.8.6
to I.8.11). The waterway can be clearly seen in each photo because it preceded the building
of houses in the development.
Chapter I.8:
183
Figure I.8.7 The second stage of the housing development. (Photo by Bordner Aerials. Copyright
Balmori Associates.) (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
Figure I.8.8 Third stage of the housing development of homes priced at $137,000 to $153,000. (Photo
by Bordner Aerials. Copyright Balmori Associates.)
Conclusion
The Farmington, MN prairie waterway project and 486-unit housing development was finished in the fall of 2000 and was completely sold out by that time. Residents who purchased
the $150,000 homes are pleased not only with the fact that they do not have to worry about
184
Figure I.8.9 View of the third stage of the Park Place housing development with water system.
(Photo by Design Center for American Urban Landscape. Copyright Balmori Associates.) (Color
Figure I.8.10 The 486-unit, completed Park Place development in 2000. (Photo by Design Center for
American Urban Landscape. Copyright Balmori Associates.)
possible flood damage to their homes, but also that they can enjoy the beauty and proximity
of the park, pond system, and waterway at the edge of the development. Children play in
the ponds and connecting streams. Fishermen catch bass. Families spread blankets for picnics
near the waters edge. Joggers and cyclists frequently spot deer, woodpeckers, chipmunks,
and other wildlife as they use the paved paths and green spaces (Figures I.8.12 to 8.15).
In addition, neighbors from two nearby developments also use the green space and
waterways created.The drainage system in the Farmington Park Place project, which was
completed in 1998, has worked well since its timely inauguration with a 100-year flood in
September of that year. Then in April 2001, as the Mississippi River again swelled over its
banks and flooded large areas, the drainage system prevented substantial flood damage to
the immediate area. Some local residents are completely unaware that the waterway/park
Chapter I.8:
185
Figure I.8.11 The finished Prairie Waterway/Park Place project in 2000. (Courtesy of City Planning
Department.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Figure I.8.12 A view of the series of interconnected ponds and streams at Park Place with houses
in the background. (Courtesy of William Morrish.)
system they enjoy for recreation is also responsible for solving the stormwater and drainage problems that would likely have existed without it. The productive park is now 4
years into its planting and is on its way to becoming established. In this case, our goal to
combine a successful drainage system to control stormwater with usable parkland alongside a stream/pond area indicates that function need not take precedence over a public
amenity like green space. The Farmington waterway project shows that a stormwater
186
Figure I.8.13 Planted wetlands help to clean and slow runoff at Park Place. (Courtesy of City
Planning Department.
Figure I.8.14 The gridded street pattern and drainage improvements as suggested to the Sienna
Development Corporation. (Courtesy of William Morrish.)
solution, such as a productive park, can indeed work successfully and improve the quality
of life for local residents.
Acknowledgments
Design, Diana Balmori (Principal), Patricia Crow, and William Coyne, Balmori Associates, Inc., Landscape and Urban Design, New Haven, CT
Hydrologist Paul Barten, Ph.D., Yale Univ. School of Forestry and Environmental
Studies, New Haven, CT
William Moorish and Tom Hammerberg of the Design Center for American Urban
Landscape, Univ. of Minnesota, Minneapolis, MN
Chapter I.8:
187
Figure I.8.15 Residents who purchased the homes are pleased by the beauty and proximity of the
91-acre park, pond system, and waterway at the edge of their development. (Courtesy of William
Morrish.)
Developer Rodney D. Hardy, V.P., Sienna Development Corporation, Minneapolis,

MN
Richard Krier and James R. Hill, Inc., Planners/Engineers/Surveyors, Burnsville, MN
Ronald P. Peterson, Peterson Environmental Consulting, Inc., Edina, MN
John Corrigan, Soil Conservation Service, Farmington, MN
Glenn R. Cook, Mustafa Emir, and John Smyth of Bonestroo, Rosene, Anderlik and
Associates, Engineers and Architects, St. Paul, MN
John Never, MN Soil and Water Conservation, Minneapolis, MN
Dave Ballman, Corps of Engineers, St. Paul, MN
Ken Brooks, MN Dept. of Natural Resources, St. Paul, MN
Susan Galswich, Society for Wetland Scientists, Des Plains, MN
Jim Perry, and Erville Gorham, Univ. of Minnesota, Minneapolis, MN
Neil Diboll, Prairie Nursery, Westville, WI
Pat Lynch, MN Dept. of Natural Resources, St. Paul, MN
Bill Penning, MN Dept. of Natural Resources, St. Paul, MN
Organizations including The Nature Conservancy, Ducks Unlimited, and the Freshwater Institute could be useful in planting or monitoring.
Literature cited
Boon, B. and Groe, H., Nature Is Heartland: Native Plant Communities of the Great Plains, Iowa State
Univ. Press, Ames, IA, 1990.
Kusler, J.A. and Kentula, M.A., Wetland Creation & Restoration: The Status of the Science, Washington,
DC, 1990.
Mitsch, W.J. and Gosselink, J.G., Wetlands, Van Nostrand Reinhold, New York, 1993.
Minnesota Dept. Natural Resources, Minnesotas Native Vegetation: A Key to Natural Communities,
State of Minnesota, Dept. Natural Resour., St. Paul, 1993.
Schueler, T.R., Design of Stormwater Wetland Systems: Guidelines for Creating Diverse and Effective
Stormwater Wetlands in the Mid-Atlantic Region, Metropolitan Washington Council of Gov.,
188
Appendix A: Guidelines
Up to this point, I have given a general description of all the parts that went into designing
a water system to serve as both a drainage system and a public space, i.e., a productive
park. What follows is a synthesis of the main guidelines given to the Farmington Town
Board and the Sienna Development Corporation to enable them to carry out the project.
Vegetation
Planting recommendations
Because of the difficulty in predicting post-construction water levels, a final Planting Plan
for the Prairie Waterway can be done as soon as the waterway has been dug and the
conditions for planting are known. It is possible, however, to set forth some guidelines
for planting at this time. Guidelines for plant choice:
Use material compatible with soils, climate, and local conditions (see Plant Lists).
This criterion is of the utmost importance.
Use material initially that will stabilize banks and prevent erosion. Thereafter, plant
with as much local material as possible, because that material has proven to adapt
to local circumstances well and is available free of charge, minus labor, which can
be voluntary.
Plant according to hydroperiods use the right plant for the specific water regime
of a given location. (See Wetland Vegetation section that follows.)
Plant many trees. They provide shade for the waterway, which is important for
many reasons including bank stabilization, spatial interest, habitat for birds and
other animals, variation from farm fields, windbreak, evapotranspiration, and help
in creating a sense of scale.
Plant shrubs, which will provide shade for small animals, habitat and food for
birds, protection of wetlands from windblown particles, enclosure of space, and
edge definition.
Plant grasses and sedges, which provide protection from soil erosion and from
drying out, a valuable source of food for wildlife, the basis for rich prairie soil, a
medium that slows the progress of the water and catches fine particles, and shelter
that allows animals to get around safely.
The range of plant material types will provide a healthy environment for water and
a diversity of wildlife. The next step after initial planting to stabilize the waterway is
planting trees, shrubs, and permanent grasses, as well as wetland plants.
A number of things will ultimately determine the planting patterns in the Prairie
Waterway park, including soils and water regimes, waterway function (such as stability
of the form of the waterway, its function as a habitat, and treatment of stormwater),
recreational and aesthetic concerns, and maintenance issues.
Wetland vegetation
This list of wetland vegetation excludes Typha spp. and Phragmites communis, because they
are of little value to wildlife. William Mitsch holds that local soil underlain with clay is
Chapter I.8:
189
best: limestone-based gravel rather than silica-based gravel should be used for retention
of phosphorus and anaerobic conditions are better for denitrification.
Primary species
Phalaris arundinacea, Sagittaria latifolia, Scirpus fluviatilis, Scirpus americanus, Scirpus acutus,
Scirpus validus, Sparganium eurycarpum, Caresx lacustri, and Carex spp.
Secondary species
Acorus calamus, Cephalanthus occidentalis, Hibiscus moscheotos, Hibiscus laevis, Nuphar luteum,
Peltandra virginica, Pontederia cordata, Saururus cernus, and Zizania aquatica palustris.
Soils and water regimes

Soils
As shown in the Dakota County Soil Survey, soils in the area of the Prairie Waterway Park
Place are made up of Cylinder (129) and Marshan (252) soils.
Cylinder 129
The pH of the Cylinder soils becomes more alkaline with greater depth. The range is from
pH 5/6 to8.4. Windbreaks and environmental plantings for this soil include (815) Siberian crabapple, tatarian honeysuckle, Siberian peashrub, lilac, (1625) eastern red cedar,
hackberry, northern white cedar, bur oak, white spruce, Scotch pine, and (2635) green ash.
Habitats most compatible with Cylinder soils are open-land wildlife and woodland
wildlife, because the soil is good for grain and seed crops, grasses and legumes, wild
herbaceous plants, hardwood trees, and coniferous plants. Cylinder oils allow only fair
conditions for wetland wildlife.
Marshan 252
Marshan soils range with depth from pH 5.6 to 7.3. Windbreaks and environmental
plantings for this group include (815) redosier dogwood, northern white cedar, American
plum, purpleosier willow, tatarian honeysuckle, (1625) amur maple, white spruce,
(2635) hackberry, golden willow, silver maple, green ash, and (35+) eastern cottonwood.
Habitats most compatible with Marshan soils are open-land wildlife and wetland
wildlife because the soil is good for grain and seed crops, grasses and legumes, wild
herbaceous plants, wetland plants, and shallow water areas. Marshan soils will allow only
fair conditions for woodland wildlife.
Once the waterway has been built, it is intended that Wetland types will be:
Type I: Periodically flooded wetlands
Type II: Sedge meadows
Type V: Ponds with an aquatic bench of emergent wetland plants at the edge
Waterway function
Stability
A cover crop such as winter ryegrass or millet (in summer) should be seeded immediately
after excavation to stabilize the soil. It is recommended that the topsoil removed from the
190
excavation of ponds and wetlands be stockpiled and redistributed in the wetlands because
this soil will contain seedbanks that should flourish in the new waterway. This has been
found to be a successful way of planting constructed wetlands. Whole plants and plugs
of 8 to 10-cm diameter from existing local wetlands have the greatest chance of success.
Buying plants from a nursery is also successful, but expensive.
Habitat
Varied habitats are included in the waterway because of the different types of wetlands
planned.Type I wetlands contain water for a short time after heavy rains or snowmelt and
are vegetated with bottomland hardwoods and grasses. Nineteen species of reptiles and
amphibians, 11 species of mammals, and 83 species of birds are attracted to this habitat.
Type II wetlands are inland fresh meadow (sedge meadows > 0). These are shallow
depressions without standing water but waterlogged within a few inches of the surface
during the growing season. They are vegetated with grasses, sedges, and rushes. Seven
species of reptiles and amphibians, 10 species of mammals, and 47 species of birds frequent
this habitat.
Type V wetland is inland open fresh water with emergent vegetation restricted to an
aquatic bench at the edge. Fifteen species of reptiles and amphibians, 12 species of mammals, and 74 species of birds inhabit this area.
Stormwater treatment
Four types of areas related to stormwater treatment will require planting: upland, stormwater swales, the channel basin, and the wetlands. Upland area vegetation, although above
the high-water level, protects the wetlands from windblown particles. It is a vital buffer
zone that separates the wetlands from windblown particles. In addition, its roots will help
stabilize the banks of the channel and break the rainfall, allowing it to filter into the ground,
where it hits, rather than washing into the waterway. It also performs a vital role in evapotranspiration. The trees that would be recommended for this area should be long-lived,
because they provide the most visible structure of the waterway planting. Shrubs should
be dense to block windblown particles during all seasons.
Stormwater swales and the dry NURP will contain sedges and grasses to increase
residence time of the water and to provide plenty of surface for catching fines. The main
requirement of these sedges and grasses is that they survive varied water conditions, from
fairly dry to inundated. In addition, they need to perform well in taking up nutrients.The
channel basin planting will also need to survive frequent inundation, as the channel water
will frequently rise to fill this area. A dense root system, quickly established, to prevent
erosion, will be important.
See the sections that follow for recommendations on planting the wetlands.
Maintenance issues
The path used for bicycling will also allow small maintenance vehicles to get around the
site for general grooming of the area. If most trees and shrubs are well adapted to their
site, and there is plenty of diversity in the planting, little maintenance will need to be
done other than pruning to open up view or remove branches that enter the pathway. The
areas that will require somewhat more maintenance are the ball fields and the lawn areas
overlooking the ponds. The grass will need to be kept short for families who plan to play
and picnic here. The pathway system will also allow access to the water for periodic
monitoring.
Chapter I.8:
191
Aesthetic and recreational concerns

Whether walking, jogging, bicycling, picnicking, playing in the streams, or running in the
grass, residents benefit from a diverse, well-planted landscape. Planting can assist with the
perception of scale. A users enjoyment increases if one can locate oneself in the landscape.
Planting for all color and early spring greening will draw people to the park for longer
periods of time and increase their enjoyment.
Civic involvement and education

The following are some suggestions of projects related to the waterway and groups that
might be interested in becoming involved.
Horticultural organizations in and around Farmington may know of a regional expert
in species identification who would be able to make an appraisal of the wetlands to be
replaced. Some of the plants from these wetlands may be incorporated into new wetlands.
Local genotype planting material will succeed better than material from another
location. A local Heirloom Seed Group and local nurseries would be possible sources of
such material.
Most service organizations, such as Rotary Clubs, Veterans groups, or Girl Scout and
Boy Scout troops, plan projects for their communities. Investment in the park/waterway
project could take many forms, including donated time or the donation of special elements
such as picnic tables, benches, or bridges.
The Waterway will be a destination for school groups, 4-H clubs, the Photography
Club, or garden clubs. Some projects that these groups might find interesting include:
Making and installing swallow and bat boxes to control mosquito populations near
the waterway
Collecting seeds from local plants to use in the waterway or removing plugs from
existing wetlands to plant in the new wetland
Planting trees as a town-wide effort the trees that outline the rooms can be
planted with proper supervision as soon as the waterway seeding has taken hold.
Various organizations could choose one or two specific rooms to plant.
Monitoring this would be a possible enrichment to the curriculum of a science,
geography, or social studies teacher as a short- or long-term project for students.
The Freshwater Institute offers seminars to educate teachers about monitoring. A
short-term project could entail monitoring the runoff from the housing areas and
determining the effects of lawn fertilizer on the waterway, for example.
Other types of projects include wildlife monitoring, including birds, or testing the
quality of water where it enters the system, compared with the exit point. James Perry at
the University of Minnesota might also be interested in using the Park Place waterway
as a study area for graduate students.
192
Response
Successfully marrying form and function in stormwater management
This chapter by Diana Balmori demonstrates the successful incorporation of many elements of stormwater management described in other chapters in this book: low-impact
development approaches (Larry Coffman and Michael Clar in Chapters I.5 and I.15,
respectively), detention ponds (Desheng Wang in Chapter I.2), swales and streams (Glenn
Allen in Chapter I.13 and Robert France and Philip Craul in Chapter I.7), and wetlands
(James Bays in Chapter I.14).
Balmoris strategy in the Minnesota project outlined in this chapter most closely
matches, in both spirit and execution, that of Catherine Berris in Chapter I.9 for a site in
British Columbia. The water sensitive design that Balmori presents perfectly encapsulates
the paradigm shift occurring in wetland creation in terms of movement from functional
stormwater wetlands to multipurpose stormwater wetland parks. In this new paradigm,
wetlands are no longer seen as just means to improve environmental conditions; they are
also regarded as an excuse to create green spaces to improve sociological conditions.
Recreation, aesthetics, and other human amenities, once marginalized as afterthoughts,
are instead important elements incorporated into the overall water sensitive design from
the very start. As this chapter states, stormwater function need not take precedence over
creation of a park with corresponding benefits to the quality of life for nearby residents.
Three other features of this wetland park project are worthy of attention:
The requirement for a large, multidisciplinary team was integral to project fruition.
Innovative financing in terms of taxes to the developers helped to raise funds.
The interesting approach of fielding real, viable alternatives, any of which would
have been appropriate for selection, instead of stacking the cards with inappropriate, substandard alternatives other than the obvious one the public is forced to
choose, is noteworthy.
chapter I.9
A stormwater wetland becomes a nature

park (British Columbia, Canada)
Catherine Berris
Abstract
The study area is a 57-acre park surrounded by suburban development in Abbotsford, a
small city in the Fraser Valley of British Columbia, Canada. The original purpose of the
project as defined by the client was to provide a stormwater detention facility to minimize
downstream flooding of agricultural lands, with a path surrounding it. The landscape architects expanded the purpose to include a variety of ecological and recreational objectives.
The site consisted of overgrown farmland with a small, channelized creek inhabited
by two endangered species of fish. The design and construction resulted in a lush, diverse
nature park. Some of the design features and methods included:
Relocation of the existing creek to include meandering and a larger cross-section

A large constructed wetland for storage of 10 acre-ft of water to a 5-ft depth
Retention of the best existing vegetation
Shoreline designed for variety and edge effect with peninsulas and islands
Wetland, riparian, forest, and meadow plantings, with extensive use of bioengineering
Landforms and planting designed to shade the creek to the highest degree possible
Major and minor path system, with some islands inaccessible to humans
Six structures to provide use and interpretive opportunities, with design reflecting
regional context, and built in an environmentally sensitive manner
Phasing of construction over 3 years to satisfy environmental authorities
Introduction
Stormwater runoff has become a pervasive problem in suburban settings (Chapters I.1
and I.4). The removal of original forest vegetation, stripping of top soil, paving of roads
and parking areas, and construction of housing have changed the hydrological cycle.
During storm events, the rate of runoff is extremely high, causing erosion along streams
unsuited for the strong flows. In the winter when the ground is saturated, storm events
often cause flooding because there is no capacity to hold the excessive amount of runoff.
Conversely, in the summer, seepage and gradual runoff from forested areas are no longer
1-56670-562-2/02/$0.00+$1.50
193
194
Figure I.9.1 Channelized creek through previous farm field prior to construction. Fishtrap Creek,
Abbotsford, BC. (Photo by Catherine Berris Associates.) (Color version available at www.gsd.
available. Stream flows are very low, or nonexistent, water temperatures rise, and water
quality decreases.
It has been shown that the total watershed impervious area and riparian forest integrity are measurable indicators of the ecological health of urban watercourses. In British
Columbia, ecological health is rated based on the following criteria, among others: excellent (streams that support wild salmon, particularly coho); good (hatchery salmonids and
some species of wild salmonids); fair (trout and hatchery fish only); and poor (no fish or
coarse fish only, e.g., stickleback and sculpins) (Page et al., 1999; Kerr, Wood Leidal Assoc.,
2001).
New approaches to stormwater management are attempting to reduce the extent of
impervious surfaces and to promote new forms of development that direct water back
into the ground as close to its source as possible (Gibb et al., 1999; Chapters I.3 and I.5).
This desire to cure the disease of urban runoff (Chapter I.1), however, is not always
possible in settings that have already been fully developed.
As in several other chapters in this book (e.g., Chapters I.2 and I.8), this chapter
describes a stormwater management system that has been incorporated into a developed
area. It presents a story about palliative care. It describes how a stormwater management
wetland was introduced into a community to address the flooding and habitat problems
caused by suburban development.
The context
The project was initiated in 1990. The site was a 57-acre area of overgrown fields containing
a small channelized creek (Figure I.9.1), some scrubland, and a small area of mowed grass
with a path installed by the Parks Department (Figure I.9.2). Fishtrap Creek drains 7,500
acres of urban upland and agricultural lowland south to the Nooksack River in Lynden,
WA.
The problem was frequent flooding of rich agricultural lands. The cause was the
suburban development in the upper watershed, development undertaken in the only way
known at the time. The land was cleared, roads installed, storm sewers constructed, houses
built, and then street trees (sometimes) added.
Chapter I.9:
A stormwater wetland becomes a nature park (British Columbia, Canada)
195
Figure I.9.2 Small park area near creek prior to construction. Note mowed grass park, walkway,
and flooding of creek. (Photo by Catherine Berris Associates.) (Color version available at www.gsd.
The purpose of the project, as the client defined it, was to provide a stormwater
detention facility. The project function was to attenuate post-development flows from 1to 10-year storm events lasting 1 to 24 h. A predevelopment runoff of 92 ft3 per day (which
became the outlet control flow) and a post-development runoff of 512 ft3 per day were
predicted for a 10-year, 1-h duration storm (Hicks et al., 1993). The governing detention
of 10 acre-ft was determined for 10-year, 6- and 12-h duration storms.
Engineers were hired to design the stormwater management system. Fortunately, early
in the process, the engineer called a landscape architect and said, The City has purchased
some land for a stormwater detention facility. They would like you to design a path to
go around it. By the way, two endangered species of fish live in the creek system, so the
environmental agencies will have some input into the project. That started a planning,
design, and construction process that spanned five years.
The project had significant requirements. The need to store 10 acre-ft of stormwater
meant that most of the existing vegetation had to be cleared. The creek supports cutthroat
trout and two endangered fish species: the Salish sucker and Nooksack dace. Environmental authorities required the project to be sensitive to the habitat needs of those fish.
The main requirements were to allow instream construction only during the 2-month
period least critical to the fish life cycle, to minimize siltation, and to limit water temperature increases in the future. Because of the concerns about potential impacts on fish
habitat, a 2-year delay was required between completion of the first phase the south
basin and construction of the north basin.
The primary landscape architecture design concept was to give people that little bit
of Thoreau that we all crave. Many people moved to Abbotsford for a small-town, rural
experience. It is a town surrounded by agricultural fields, streams, and hedgerows. Within
the community, however, those bird sounds and the ability to wander through a natural
space had become scarce. The purpose of the park design therefore included:
The development of a park with a variety of natural habitats, to bring back a
semblance of the nature that was there before the land was cleared for agriculture
Promotion of use by wildlife, for the benefit of all species, including the human one
196
Figure I.9.3 Site during construction. Extensive removal of material was required for stormwater
storage. (Photo by Catherine Berris Associates.) (Color version available at www.gsd.harvard.edu/
Provision of nature-oriented recreational opportunities, like bird watching, walking through clean air, and appreciation of natures sights and sounds
Provision of ecological and cultural interpretive facilities, to enable learning about
the natural systems on the site and its past history of early pioneer homesteads
Wetland design
The basic approach to the site design included relocation of the existing creek to include
meandering and a larger cross-section. The channel was designed to provide a 6-ft depth
at low water, in order to remain cool for fish and to maintain hydraulic efficiency. The
meandering was introduced in order to provide a more natural appearance than the
previously constructed drainage ditch.
Adjacent to the creek is a large constructed wetland capable of storing the excess
stormwater that overflows the creek banks. During storms the water level can rise 5 ft,
and then it is gradually released through a control structure downstream of the wetlands.
Riparian slopes around the wetland range in slope from 3:1 to 5:1. This grade provides
stability from erosion and the opportunity to support riparian vegetation easily, and it
allows for a return to the existing grade within a reasonable distance, thereby enabling
the protection of the best existing vegetation. Even with these slopes, most of the site had
to be cleared and excavated (Figure I.9.3).
The process of designing the shoreline configuration around the creek and the wetland
involved a significant collaboration among the engineers and landscape architects. The
required storage volume for stormwater was a major limiting factor. From a design that
began as a large oval, every pushing of the shoreline toward the wetland to provide
habitat, lookout points, and shade for the water required a respective pulling of the
shoreline back in another location (Figure I.9.4). Space was extremely limited, and consideration was given to provision of appropriate buffers for residents and protection of
important existing habitat, such as a raptor nest.
The resulting configuration includes variety and a significant amount of edge with
peninsulas and islands. The landforms and planting were designed to shade the creek to
the highest degree possible; however, effective cooling was difficult due to the size of the
wetland and the extent of the area to be planted.
Chapter I.9:
197
Figure I.9.4 Plan of north basin. Note meandering creek surrounded by wetland, small patches of
retained forest, and proposed walkways and structures. Ball fields were preexisting. (Photo by
Catherine Berris Associates.) (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
Enhancement of water quality through trapping, transformation, and storage is an

important function of the wetlands. Sediment traps were built on all urban storm flow
inlets to the wetlands to capture heavier sediments.
Planting
Planting in the park included wetland, riparian, forest, and meadow plantings, with
extensive use of bioengineering. The park involved revegetation of a very large area. About
35,000 trees, shrubs, and ground cover plants and 66,000 wetland and riparian plants were
installed. The plants are almost all native, with imported species only in the more manicured portions of the park. The growth rate has been outstanding.
Wetland test plots of different species were established immediately after the excavation was complete. Because of the depth of water inundation, cattails were by far the most
successful wetland plant. Although extensive planting of cattails can limit diversity, in
this case the environmental agencies were concerned that vegetation in the wetland
become established quickly, and cattails offer high value for biofiltration. One challenge
was that a severe winter resulted in freezing of the wetland after a fall planting. The ice
pulled numerous wetland plants from the ground.
The riparian slopes were planted almost exclusively with live stakes and brush layering
of willow and dogwood. Their success was primarily dependent on the slope material. Where
there was native peaty soil or an imported growing medium, the live cuttings established
extremely well. In locations with a high proportion of gravel on the slopes, the survival rate
was lower. In some areas, the growth of willows was so significant that views of the creek
and wetland along major stretches of the paths, as well as views from benches, were blocked.
In these areas, the city was advised to clear view windows after several years.
All areas of the site above high water were hydroseeded. Trees and shrubs were
planted within the grass areas. This method resulted in some loss of shrub material,
198
Figure I.9.5 Entry pier extending over creek as seen across new wetland. (Photo by Simon Scott.)
(Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
because shrubs were outcompeted by the grass, but the eventual appearance achieved the
desired effect of naturalized scrubland. In other projects, a ring of mulch around each tree
and shrub helped to increase the success rate of plants. The alternative of mulching beds
around shrubs results in a more formal appearance, especially in the short term, and it also
requires a high level of maintenance in the first few years; however, more shrubs survive.
The tree planting program was very successful, and there was a high survival rate.
Some trees were planted in monoculture groves, with others as mixed forest species. The
attempt to establish a grove of Garry oak, which is native to a site not far from the study
area, was not successful. The site is probably too low-lying and peaty for that tree, even
though the Garry oaks were planted on the upper portion of a hillside.
Wildflowers were planted with the grass at several key locations. Native wildflower
seeds were acquired from a specialized supply company. The wildflowers were very
successful in the first year, but as is often the case, they became less profuse over time
due to competition from the grasses. Lupines have remained a highlight year after year.
Park amenities
A major paved path meanders along one side of the wetland, and a secondary gravel path
provides access on the other side. The paths form three major loops, one around the south
basin, and another around the north basin, with a pedestrian bridge dividing the latter
into two smaller loops. Some of the islands have small gravel paths onto them. Others
have wildlife habitat as the primary goal, and these islands are inaccessible to humans.
Six structures within the park support various types of viewing and access and
incorporate interpretive information. The design of the structures reflects the regional
context, with barn-like roof forms and timber construction. To address the environmental
sensitivity of the site, all timbers were precut and pressure-treated off-site. The structures
over the water were built on piles to minimize ground disturbance. The structures were
designed to blend into the setting, to appear light from a distance yet solid from nearby.
The entry pier (Figure I.9.5) is located near the parking lot. It extends over the wetland
and is open to the sky and the water in the center. An orientation map and interpretive
information are on a sign where the walkway extends onto the pier.
Chapter I.9:
199
Figure I.9.6 Picnic shelter on peninsula with pedestrian bridge crossing creek on right. (Photo by
Catherine Berris Associates.) (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
Figure I.9.7 Deck on old railway embankment ends in lookout over park. (Photo by Catherine Berris
Associates.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
A picnic shelter (Figure I.9.6) is located on a peninsula central to the north basin, near
the pedestrian bridge. The picnic shelter has space for gathering along its side and has an
open grass area for associated play. A patch of domestic blueberries, which are a remnant
from the farm use, is located near the picnic shelter.
A deck located on an old railway bed (Figure I.9.7) provides the greatest opportunity
for historic interpretation. The deck was designed to mimic the railway pattern and
provides an excellent overlook because it is on a raised embankment that previously led
to a trestle. Old piles from the trestle found during construction were placed vertically
within the wetland as wildlife habitat. Laborers from the local correctional center participated in building a stairway onto the railway grade.
200
Figure I.9.8 Reading shelter. Note interpretive signage on structure and seating steps overlooking
wetland. (Photo by Simon Scott.) (Color version available at www.gsd.harvard.edu/watercolors.
Figure I.9.9 Boardwalk to island with interpretive sign panel and bench. (Photo by Catherine Berris
Associates.) (Color version available at www.gsd.harvard. edu/watercolors. Password: lentic-lotic.)
Other structures include (1) a small reading shelter (Figure I.9.8) that provides a
stopping place in sun or rain and gives a view the length of the wetland and (2) a
boardwalk over a wetland area to an island (Figure I.9.9). Small tables and benches are
distributed in special locations throughout the park.
Interpretive signs throughout the park (Figure I.9.9) explain (1) the overall concept
and design of the park, (2) the vegetation, fish, and wildlife and their ecological importance, and (3) the rural history of the area. The signs are almost all integrated into the
structures.
Chapter I.9:
201
Figure I.9.10 Fallen tree left in park provides habitat, play environment, and display of root system.
(Photo by Catherine Berris Associates.) (Color version available at www.gsd.harvard.edu/watercolors.
Park management
An important aspect of the park development has been the ongoing management and
maintenance. Several trees on the edge of existing forested areas fell during construction
due to windthrow. The trees were left in place, where they provide wildlife habitat,
climbing opportunities, and a magnificent display of a root system (Figure I.9.10).
Close contact is maintained with the city regarding the extent of mowing in the park.
Parks maintenance crews are accustomed to mowing all grass areas. At this park, because
the trees and shrubs were planted within hydroseeded areas, it is critical that there be no
mowing throughout most of the park. This policy was implemented, and a swath about
3 ft wide immediately adjacent to the main pathway is now mowed, mainly for security
(visibility) reasons and to keep the area clear for service and emergency access. This has
allowed naturalization to occur throughout most of the park.
Guidelines on weeding are also followed within the park. A naturalized appearance
is desired, so volunteer native plants are welcome. For example, during establishment,
numerous red alders seeded in one area. Only nonnative, aggressive plants are to be
removed, e.g., Scotch broom and Canada thistle. Fortunately, no major invasions of such
plants have occurred, and purple loosestrife has not found its way into the park.
Park use
The project supports a high level of use by surrounding residents. As soon as paths were
built, they were being used. Only 2 years after completion of the landscape work, the park
had a very natural appearance. As the years pass, previously scrubby areas gradually
become more forest-like as growth takes place. Many people who visit the site report that
it appears completely natural. The paths through the park are now part of a larger
greenway network.
Use by wildlife is also high with inhabitants including turtles, small mammals, and
numerous birds such as waterfowl, herons, and owls. The local naturalist club was concerned about the project before construction. Now the naturalists are delighted with the
202
outcome and they use it as an example of a nature-friendly park design. The overwhelming
sensations in the park, especially in spring, are the natural fragrances and the constant
chorus of birds, providing a sharp contrast with the surrounding developments and road
traffic.
In the initial design stages of the project, some concerns were raised about safety,
given the 3:1 slopes into the wetland. Children are occasionally seen playing at the edge
of the wetland, but there have been no safety problems. The thick, tall growth on most of
the riparian slopes makes it difficult to reach the wetland in most locations. Where the
vegetation is not thick, the grass is deep and the slopes are relatively easy to climb (i.e.,
not slippery), so children can readily scamper up the bank.
At several points during construction, neighboring residents expressed concerns about
the park development. Site meetings were held with residents to resolve all concerns. Any
neighboring resident wanting a fence was offered one. The city now receives many positive
compliments on the park from neighbors and visitors, and there is a high level of community pride in this engineered facility that has such natural, peaceful qualities.
Summary
This project does not reduce stormwater runoff. It has served to ease the pain that the
disease of runoff has caused in this local area. The wetlands manage stormwater as
intended. The park has also improved the quality of life of visiting humans and wildlife,
providing a poetic response to the prosaic issue of stormwater retention (Weder, 1997).
The intent of the project was to develop a different type of park, focused on providing
people with the opportunity to experience and learn about nature in the city. It has
succeeded in providing a sanctuary for fish and wildlife and shows how an environmentally sensitive design can be incorporated into a community.
Acknowledgments
The author thanks the client, the City of Abbotsford Engineering Department, particularly
Ed Regts, for the enthusiastic support for all aspects of this project. The primary consultants were Dayton and Knight Ltd., engineers; Catherine Berris Associates Inc., landscape
architects; Brad Cameron, architect; and Envirowest Consultants Ltd., environmental consultant.
Literature cited
Gibb, A. et al., Best management practices guide for stormwater, prepared for Greater Vancouver
Sewerage and Drainage District, Burnaby, BC, 1999.
Hicks, R.W.B., Regts, E., and Kelly, H.G., Urban wetlands for multidisciplinary stormwater management, Eng. Hydrol. Symp., Hydraulics Division/ASCE, July 1993, San Francisco, 1090,
1993.
Kerr, Wood Leidal Associates Limited, Description of Ecological Health Rating System, North
Vancouver, BC, 2001.
Page, N. et al., Proposed watershed classification system for stormwater management in the
GVS&DD area, prepared for Greater Vancouver Sewerage and Drainage District, Burnaby,
BC, 1999.
Weder, A., Creating a natural diversion, Azure, March/April, 32, 1997.
Chapter I.9:
203
Response
Naturalized design:
Triumph of imagination and innovation
The stormwater wetland project that Catherine Berris describes in this chapter is a signature success story highlighting the triumph of the new paradigm of wetland park creation
(as described elsewhere in this book James Bays in Chapter I.14, Diana Balmori in
Chapter I.8, and Robert France and Kaki Martin in Chapter I.12) over the old paradigm
of highly engineered detention basin construction. The choice of the word creation over
construction is informative here, for although a thing constructed can be appreciated
after it is built, a thing created can be appreciated before it is even designed. Berris
perseverance in shifting the original plans for a simple, single-purpose detention basin to
the development of a multipurpose wetland park is a model lesson in the valued role that
landscape architects can play in water sensitive designs.
Elements of this project that are particularly meritorious include:
Phased development
Integration into the existing residential neighborhood
Restoration of the semblance of nature such that recent visitors unfamiliar with
the wetlands creation imagine it to be completely natural (as is the case for most
modern visitors to Olmsteds famous Back Bay Fens in Boston)
Creation of wildlife habitat such as islands for waterfowl or shaded trees to regulate
temperatures for salmon
Acknowledgment of human history
Great attention paid to recreation opportunities and to allowing people to come
into close contact with the water (easier to accomplish in less-litigious Canada than
in the safety-paranoia legal environment of the United States)
Use of architectural structures of a style endemic to the region
Need for education of city parks maintenance crews with respect for curbing their
desire to mow all grass right to the waters edge (also an issue experienced in the
project that Nicholas Pouder and Robert France describe in Chapter I.16)
chapter I.10
Wetlands-based indirect potable reuse

project (West Palm Beach, Florida)
Larry N. Schwartz, Lee P. Wiseman, and W. Erik Olson
Abstract
As the population in the city of West Palm Beach grows, and demands on water resources
for urban use increase, the development of alternative sources of water supplies becomes
increasingly important. The city has developed a program to use highly treated wastewater
from its East Central Regional Wastewater Treatment Plant (ECRWWTP) for beneficial
reuse including augmentation of its drinking water supply. To protect and preserve its
surface water supply system and to develop this reuse system to augment the water
supply, the city purchased a 1,500-acre parcel, referred to as the Wetland Reuse Site. The
Wetland Reuse Site is one component of the Wetlands-Based Indirect Potable Reuse Project.
The other component of the project is the Standby Wellfield. The Wetland Reuse Site and
the Standby Wellfield sites consist of wetlands and uplands dominated by Melaleuca. An
important goal of the project was to develop an Advanced Wastewater Treatment facility.
The facility produces reclaimed water that, when discharged, will be compatible with the
hydrology and water quality at the Wetland Reuse Site.
Introduction
A great deal of research has been performed documenting the ability of wetlands, both
natural and constructed, to provide consistent and reliable water quality improvement.
Wetlands are effective in the treatment of BOD, TSS, nitrogen, phosphorus, pathogens,
metals, sulfates, organics, and other toxic substances (Kadlec and Knight, 1995). Several
factors are important in determining the appropriate design of a wetland treatment system.
For natural wetlands these include the type of wetlands as defined by the dominant
vegetation and soils, the direction and extent of surface water flow to and from the
wetland, the location and type of downstream water bodies, the presence of protected
species, and the regulatory requirements. Although wetland treatment systems are
designed to maximize water quality improvement, the appropriate design will establish
a hydroperiod to encourage and maintain valuable wetland communities and wildlife
habitat. Therefore, the ancillary benefits that wetland treatment systems provide become
a valuable educational and public resource, as indicated in Chapter I.14.
The city of West Palm Beachs water supply system includes a surface water allocation
from Lake Okeechobee to the M Canal, via the L-8 Canal, through the Water Catchment
1-56670-562-2/02/$0.00+$1.50
205
206
Figure I.10.1 Hydrologic features in northern Palm Beach County.
Area (WCA) to the citys water treatment plant at Clear Lake (Figure I.10.1). As part of
the Everglades restoration program being implemented by the South Florida Water Management District (SFWMD), the timing, location, and quantity of water releases to canals
from Lake Okeechobee will be modified. More water will be directed toward the Everglades for restoration and less water will be sent to the canals (SFWMD, 1999). This
translates into less water available for the citys water supply.
Concurrent with increases in potable water demand, the average rate of wastewater
flow from the ECRWWTP is expected to increase to approximately 60 mgd by the year
2010. Presently, all of the wastewater effluent from the ECRWWTP is injected over 3,000 ft
into the ground (boulder zone) at the facility using six deep wells. Rather than continuing
to dispose of the wastewater effluent from this facility, the city of West Palm Beach, with
the assistance of Camp Dresser and KcKee, Inc. (CDM), developed the Wetlands-Based
Indirect Potable Reuse Project.
In order to protect and preserve its surface water supply system, and to develop an
appropriate reuse system to augment the water supply, the city of West Palm Beach, in
conjunction with Florida Communities Trust, purchased land immediately east of the
WCA and west of Floridas Turnpike (Figure I.10.1). This parcel, referred to as the Wetland
Reuse Site, is one component of the Wetlands-Based Indirect Potable Reuse Project. The
other component of the project is the citys Standby Wellfield. The Wetland Reuse Site
covers an area of 1,415 acres and consists of a combination of wetlands (wet prairies and
forested hammocks dominated by Melaleuca) and uplands. The Standby Wellfield site
covers an area of 323 acres and consists of similar wetlands and uplands dominated by
Melaleuca. An important goal of the project was to develop an Advanced Wastewater
Treatment (AWT) facility at the ECRWWTP. The facility will produce reclaimed water that,
when discharged, is compatible with the hydrology and water quality at the Wetland
Reuse Site.
Chapter I.10:
Wetlands-based indirect potable reuse project (West Palm Beach, Florida)
207
Development of the wetlands-based indirect potable reuse project

Advanced wastewater treatment constructed wetlands demonstration project
In 1992 the city of West Palm Beach implemented an AWT/constructed wetlands demonstration project. The Florida Department of Environmental Protection (FDEP) defines AWT
as producing an effluent quality of total suspended solids (TSS), 5-day carbonaceous
biochemical oxygen demand (CBOD5), total nitrogen (TN), and total phosphorus (TP)
goals of 5, 5, 3, and 1 mg/L, respectively. This 150,000-gal/day demonstration facility was
constructed in 1995 and was in continuous operation for over one year. Effluent from the
ECRWWTP was treated to AWT levels at the demonstration facility using a solids contact
unit for phosphorus reduction and deep-bed denitrification filters for nitrogen reduction.
High-level disinfection was provided before discharge to the constructed wetlands. Planting, application of reclaimed water, volunteer colonization of vegetation, and maintenance
for removal of undesirable species resulted in the establishment of a viable wetland
ecosystem in the two constructed wetland cells.
The pilot AWT facility met the treatment goals as well as all of the state and federal
drinking water standards (except for iron), Florida groundwater guidance concentrations,
and surface water quality standards (except for total residual chlorine and coliforms).
During the year of operation of the demonstration project, TP concentrations in the AWT
effluent were in excess of the TP concentrations at the Wetland Reuse Site. A coagulation
process optimization study was conducted to evaluate the feasibility of achieving appropriate reclaimed water TP concentrations for discharge to the Wetland Reuse Site.
The first phase of the optimization study focused on the iron salt coagulation process
to maximize phosphorus removal and minimize iron floc and soluble iron carryover on
a bench-scale (jar-test) basis. The bench-scale study included testing of three coagulants
(ferric sulfate, ferric chloride, and polyferric sulfate) at varying dosages, optimization of
pH conditions, and examination of several flocculation and sedimentation polymers for
floc conditioning. Alum was also originally considered as a coagulant but was disregarded
due to concerns with possible toxicity effects and problems associated with dewatering
and disposal of alum sludge. Based on results of the bench-scale study, operational criteria
for demonstration scale operation were recommended to optimize and evaluate the iron
coagulation process. Results of the demonstration-scale study indicated that, during a
significant majority of the time, the existing treatment achieved effluent parameter concentrations at or below the anticipated permit compliance limits for iron, TSS, CBOD5,
and TN, and the operational goal of 0.05 mg/L for TP. Based on the performance of the
demonstration project, the new reclaimed water production facility was designed as an
AWT process with coagulation/clarification (for additional phosphorus removal) and
deep-bed denitrification filters followed by ultraviolet (UV) light disinfection.
Baseline monitoring of wetlands

As required by the FDEP, baseline monitoring was performed at the Wetland Reuse Site
from July 1996 to August 1997. The purpose of this monitoring was to establish baseline
conditions in the wetlands prior to reclaimed water application and to determine the
appropriate quality of the reclaimed water that will be applied to the Wetland Reuse Site.
In addition to the monitoring of background hydrology, groundwater quality, and surface
water quality, the baseline-monitoring program investigated sediment quality, vegetation,
fish, and the presence of listed threatened and endangered plant and animal species.
Groundwater samples from the Wetland Reuse Site and the Standby Wellfield met the
requirements for drinking water except for iron. Iron was detected in excess of the Secondary
208
Drinking Water Standards of 0.3 mg/l at all of the wells, but not in excess of the Class III
surface water quality criteria of 1.0 mg/l. TN concentrations in the wetlands ranged from
0.67 mg/l to 3.85 mg/l, with an average value of 1.36 mg/l. The concentrations of TP were
low throughout the wetlands, ranging from <0.01 to 0.13 mg/l, with an average value of
0.027 mg/l.
Hydrologic modeling
A hydrologic model capable of simulating both groundwater flow and overland flow was
constructed and calibrated to assess the hydrology, hydrogeology, and potential hydraulic
conveyance characteristics within the project area. The groundwater flow computer code
used in this study is MODFLOW (McDonald and Harbaugh, 1984). The SFWMD Wetland
Package (Restrepo and Montoya, 1997) was used to represent overland (sheet) flow in
MODFLOW, the associated hydroperiods, and the interaction between groundwater and
surface water within selected wetlands in the modeled area.
Model simulations were performed using water levels and flows at the end of an
average wet season followed by a 7-month dry season based on a 1-in-10-year drought
condition. Additionally, the water levels and flows were evaluated with current and
projected pumping at the Standby Wellfield. The model indicates that maintenance of
viable wetlands (i.e., no extended wet or dry periods) on the Wetland Reuse Site and the
Standby Wellfield and aquifer recharge to augment the water supply can be achieved.
Reclaimed water will initially be applied to the Wetland Reuse Site at a rate of 2 in./week,
which corresponds to a reclaimed water flow of approximately 6 mgd over 770 acres of
the 1,415-acre site. The results of the modeling indicate that up to 6 mgd of reclaimed
water can be applied to the Wetland Reuse Site without producing more than an 8-in.
average rise in surface water levels in the wetlands over the 19961997 baseline hydroperiod.
A one-mile reclaimed water distribution header will be constructed along the WCA
berm (Figure I.10. 2). The reclaimed water will sheet flow across the site to a collection
ditch where it will be pumped to the Standby Wellfield for aquifer recharge. The collection
ditch and containment berm will parallel the Turnpike Canal and will sever the existing
surface water connection. Routing of flow from the Wetland Reuse Site to the Standby
Wellfield and more frequent operation of the Standby Wellfield are necessary for project
implementation. It is anticipated that the Standby Wellfield will be pumped at an average
rate of 7 mgd to increase recharge and provide raw water to the M Canal. The existing
raw water main for the Standby Wellfield extends to the M Canal.
Preliminary design
AWT design goals
Consideration of demonstrated treatment technologies and the results of activities to
optimize the removal of phosphorus from the AWT discharge indicate that the high-quality
reclaimed water facility can be designed to consistently meet permit compliance limits for
a TN and TP of less than 3.0 mg/L and 0.1 mg/L (on an annual average basis), respectively.
The high-quality reclaimed water facility should achieve operational goals for a TN and
TP of less than 2.0 mg/L and 0.05 mg/L (on an annual average basis), respectively, to
minimize change in the wetland vegetation communities. Permit compliance levels, project
AWT reclaimed water goals, and backgroundwater quality at the Wetland Reuse Site are
presented in Figure I.10.3. A commitment to construction and operation of a high-quality
reclaimed water facility has been provided to meet these stringent discharge requirements.
Chapter I.10:
Figure I.10.2 Wetlands-Based Indirect Potable Reuse Project.
209
210
Average Concentration (mg/l)
15
15
16
14
15
13.2
12
10
8
4
2
1.4
0.1
0.03
0
CBOD5
0.05
TSS
TN
Project AWT Goals
TP
CBOD5
5
13.2
5
15
Project AWT Goals

Wetland Reuse Site (Background)
NPDES Permit Compliance Levels
Typical Secondary Effluent
Typical Secondary Effluent

NPDES Permit Compliance Levels
Wetland Reuse Site (Background)
TSS
5
4
5
15
TN
2
1.4
3
15
TP
0.05
0.03
0.1
4
Figure I.10.3 Comparison of secondary and AWT effluent qualities and permit compliance levels to
backgroundwater quality at the Wetlands Reuse site.
EFFLUENT
TRANSFER PUMPS
DEEP BED
FILTERS
BALLASTED
FLOCCULATION
POLISHING
FILTERS
UV LIGHT
DISINFECTION
FeSO 4 /POLYMER
SCREEN
Return to
Aeration Basin
or Digester
Return to
Digester
Return to
Aeration Basin
or Digester
Figure I.10.4 Process schematic for the 10-mgd AWT facility.
The design of this facility was planned to produce a constant flow of 10 mgd of reclaimed
water. A schematic process flow diagram for the AWT design is provided in Figure I.10.4.
Four transfer pumps were planned to convey 10 mgd of ECRWWTP effluent to the
influent of the splitter box for the deep-bed denitrification filters. Methanol would be
added as required to ensure optimal denitrification efficiency in the filters. By moving the
denitrification reactor ahead of the phosphorus removal step, more effective and consistent
denitrification should be achieved at reduced methanol feed requirements. Effluent from
the denitrification filters would flow into a ballasted coagulation, flocculation, and sedimentation process for phosphorus removal. The ballasted process consists of two 5-mgd
ACTIFLO process trains with the clarification stage sized for 30-gpm/ft2. A ferric sulfate
Chapter I.10:
211
dosage of 15 mg/L and an anionic polymer dosage of 1 mg/L will be added to improve
the flocculation process. The clarified effluent from the ACTIFLO process will flow into a
polishing filter. The recommended polishing filter design is an automatic backwash (ABW)
traveling-bridge filter (2 units at 5 mgd, each at a maximum filtration rate of 4 gpm/ft2).
Traveling-bridge filters are economical and are in service at numerous wastewater treatment plants for TSS removal. The polishing filter will provide additional floc removal,
protect against ACTIFLO upset conditions, and provide additional solids removal. The
polishing filter should normally be lightly loaded with solids and should not appreciably
increase side stream flows back to the ECRWWTP. The AWT facility will use ultraviolet
(UV) light irradiation for disinfection. Since the FDEP does not have regulations addressing disinfection methods other than using chlorination, the California Administrative
Code, Title 22 (1978), has been used as guidance.
Wetland reuse system

For the Wetland Reuse Site, a 30-in.-diameter reclaimed water transmission pipe will be
extended from the ECRWWTP west to an 8-in. HDPE discharge header located on the
west side of the Wetland Reuse Site. Existing containment berms are located along the
north and west perimeters of the Wetland Reuse Site. Additional containment berms will
be constructed along the east and south boundaries of the Wetland Reuse Site. A lowhead transfer pump station will be installed in the southeast corner of the Wetland Reuse
Site to convey wetlands surface water to the Standby Wellfield. This transfer station will
draw from a collection ditch constructed on the eastern boundary of the Wetland Reuse
Site.
Summary
Development of the Wetlands-Based Indirect Potable Reuse Project required substantial
vision and effort. The city of West Palm Beach purchased the Wetland Reuse Site in order
to protect and preserve its surface water supply system and develop an appropriate reuse
system to augment the water supply. The AWT-constructed wetlands demonstration
project was implemented to evaluate the feasibility of full-scale implementation of the
Wetlands-based Indirect Potable Reuse Project, and a baseline-monitoring program was
completed for the Wetland Reuse Site. A study was performed to optimize the coagulation
process to achieve appropriate TP concentrations in the AWT effluent. Consideration of
demonstrated treatment technologies and the results of activities to optimize the removal
of phosphorus from the AWT discharge at the demonstration project indicate that the
high-quality reclaimed water facility can be designed to consistently meet permit compliance limits and operational goals for TN and TP.
A commitment to construction and operation of a high-quality reclaimed water facility
has been provided to meet strict discharge requirements. The results of modeling indicate
that maintenance of viable wetlands on the Wetland Reuse Site and augmentation of the
water supply through aquifer recharge can be achieved. The key is proper operational
management of the entire system, which includes flow to the Wetland Reuse Site, conveyance to the Standby Wellfield, and pumping from the Standby Wellfield to the M Canal.
Literature cited
California Administrative Code, Title 22, Environmental health Wastewater reclamation criteria,
State of CA Dept. Health Serv., Sanitary Eng. Sect., Berkeley, CA, 1978.
Kadlec, R.H. and Knight, R.L., Treatment Wetlands, CRC Lewis Publishers, Boca Raton, FL, 1995.
212
McDonald, M.G. and Harbaugh, A.W., A modular three-dimensional finite-difference groundwater

flow model, U.S. Geol. Survey Tech. Water Resour. Investig. Book A3, 1984.
Restrepo, J.I. and Montoya, A., MODFLOW wetland module final report, prepared for the South
FL Water Manage. Dist., 1997.
SFWMD (South FL Water Manage. Dist.), 1998 Everglades Annual Report, 1999.
Chapter I.10:
213
Response
Treating wastewater with innovative technology
South Florida can, in many ways, be regarded as a microcosm for both ailments and
innovative solutions with respect to water management. Daniel Williams outlines various
aspects of regional watershed management in Chapter II.11, and James Bays discusses a
wetland creation project in Chapter I.14. In this chapter, Larry Schwartz presents a case
study from the same area that is informative in its creative use of science and engineering
as integral components toward achieving water sensitive design.
Faced with growing water shortages due to redirection of flows to the Everglades as
part of the restorative efforts there, West Palm Beach decided to halt its practice of deepwell injection of wastewater and instead to treat it on the surface through use of wetlands.
Schwartz describes various advanced technologies that effectively reclaim the treated
water for future use. One of the ancillary motivations of this shift in treatment approaches
was the understanding that the surface wetland, upon completion, would provide benefits
to wildlife and be used as an educational and public recreational resource.
The most important lesson gleamed from this study for water sensitive designers is
the recognition that without solid and defensible scientific data, the effective design of
treatment wetlands may be problematic, if not impossible. It is absolutely critical that a
comprehensive predesign monitoring program be established for projects, especially those
implementing innovative technologies. Otherwise, it is impossible to determine the relative quality of the discharge reclaimed water following wetland treatment. Further, as
Schwartz mentions, biological assessments must also be addressed and the use of hydrological modeling applied as a design tool.
chapter I.11
Restoring urban wetland Pond

systems (Boston, Massachusetts)
Clarissa Rowe
Abstract
The Halls Pond Sanctuary, located in the Town of Brookline, Massachusetts, is an exceptional natural area within a dense urban environment. Its relationship with the adjacent
residential areas, the caring and commitment of the Friends of Halls Pond, its history as
a spark for the development of the Massachusetts Audubon Society, as well as the beginnings of the national conservation movement all contribute to its importance. Visitor
pressure, unchecked invasive vegetation, and increased development and runoff have
acted in concert to degrade the pristine nature of the sanctuary. The challenge of the Halls
Pond Sanctuary Restoration Project was to improve water quality and to expand, reintroduce, upgrade, and integrate carefully a number of habitat types into a cohesive whole,
while providing sensible opportunities for visitor enjoyment and interpretation of a rare
and valuable urban wild.
Hills Pond, located in Menotomy Rocks Park in Arlington, Massachusetts, is another
example of a wetland and pond restoration in a dense suburb of Boston. Hills Pond, three
times larger than Halls Pond, is also the collection point for storm runoff from the
surrounding watershed. In 1991, the Town of Arlington let a dredging contract that
proposed cleaning up the pond, creating a wetland swale, and placing the pond dredge
throughout the park. The challenge of the Hills Pond project was to get a stalled dredging
project completed and to redesign elements of the engineering project to better suit the
public park setting.
Introduction
Over the past 100 years, the suburbs of Boston have changed from farmlands and large
tracts of family land to densely populated communities that support Bostons economy
and position as one of the countrys major cities. The farmlands and family lands have
been replaced by dense residential development. Luckily, in these two case studies there
were individuals and town committees that recognized the land had natural value that
could enhance the lives of future generations. In Brookline, which abuts Boston to the
west, Minna Hall, a founder of both the land conservation movement and the Massachusetts Audubon Society, was concerned about protecting the pond and the surrounding
land for future generations. She offered to donate her family land, about 2 acres, to the
1-56670-562-2/02/$0.00+$1.50
215
216
town as a nature sanctuary, but her offer was refused. It was not until 1975 that the towns
citizens, along with the Conservation Commission, acquired the land in order to prevent
its threatened development and subdivision. At the end of the 19th century in the northwest town of Arlington, a small group of prominent citizens, members of the Arlington
Improvement Association, saw the value of securing public parkland for the town. As a
result, Menotomy Rocks Park was founded in 1896. Hills Pond, which abutted the original
Park, belonged to the George Hill family, which had received title to the land from a 17thcentury royal land grant. In 1884 farmer Hill transformed the former Little Island Swamp
into an irrigation pond for his farmlands. The town subsequently acquired Hills Pond in
1924.
Unfortunately, by the 1990s it became apparent that the municipal owners of both
Halls and Hills ponds had not been effective stewards of their ponds and natural areas.
Fortunately, in both cases, dedicated groups of nature lovers formed organizations to
oversee the care of the land. In the early part of the 20th century, municipal budgets
allowed a large municipal work force, which often included people with skill in arbor
care and an understanding of natural systems. In the 1980s, however, Massachusetts
passed a tax-cutting measure that allowed municipal budgets to grow by only 2.5% per
year. Consequently, money could be spent only on the essentials, such as schools, health,
and safety. By the early 1990s, many Massachusetts parks were in dire need of capital
improvements and maintenance, and both these projects had been relegated to the bottom
of their towns capital improvement plans. Park commissions were focused primarily on
providing active recreation for their towns. It fell to the conservation commissions to be
the stewards of the natural and passively used areas. In both case studies, the conservation
commissions realized the problems with their ponds and began the process to clean them
up. Because the conservation commissions budgets were minimal, cleanup of the ponds
was really accomplished by the vocal park advocates, who never stopped putting pressure
on town officials.
The author was involved in both projects. Brown and Rowe, Landscape Architects
and Planners, was the prime consultant for the design phase of the Halls Pond Restoration
Project. The author, in her capacity as president of the Friends of Menotomy Rocks Park,
served as a pro bono landscape architect for the Hills Pond project.
The pond science and engineering services for both Halls and Hills ponds were
provided by ENSRs Dr. Ken Wagner. Dr. Dennis Lowry, also of ENSR, was the wetland
scientist for the Halls Pond project.
Environmental issues at Halls Pond, Brookline, Massachusetts

Situated in a low-lying and busy urban area, Halls Pond Sanctuary, Amory Woods Conservation Sanctuary, and Amory Playground represent valuable open space for the people
and wildlife of Brookline, MA. Few other significant public open spaces are located in
this section of the town, and none with a body of water, which further adds to the
sanctuaries beauty and ecological importance (Figure I.11.1). A Plan for Halls Pond
Sanctuary, prepared under the auspices of the Brookline Conservation Commission in
1996, outlined several goals for the pond (Giezentanner and Eunson, 1996). The two
sanctuaries needed to be reconnected and the wetland between them reconstructed
(Figures I.11.2 and I.11.3). Invasive vegetation and water pollution caused by groundwater
and storm runoff needed to be brought under control to improve the water quality of the
pond. Other goals called for developing a new path system (Figure I.11.4), creating new
entrances and gateways, erecting new fences, and reintroducing native plants.
An understanding of the sources of water and contaminants is essential to improving
and managing any pond (Figure I.11.5). Generally, three primary water sources are available:
Chapter I.11:
Restoring urban wetland Pond systems (Boston, Massachusetts)
217
Figure I.11.1 Halls Pond. (Photo by Alison Richardson.)
Figure I.11.2 Edge of Halls Pond. (Photo by Charles Mayer.)
Figure I.11.3 Long view of Halls Pond. (Photo by Charles Mayer.)
direct precipitation, groundwater seepage, and surface runoff. Sources of water and contaminants for Halls Pond had been evaluated in a past study (Metcalf and Eddy, 1986)
and confirmed by later testing (ENSR, 1999). Precipitation, often a minor influence and
not especially amenable to control, did not appear to be a critical factor for Halls Pond.
The other two sources were more influential and manageable.
218
Figure I.11.4 Boardwalk at edge of Halls Pond. (Photo by Clarissa Rowe.)
Figure I.11.5 ENSR personnel testing water and sediment in Halls Pond. (Photo by Clarissa Rowe.)
Halls Pond lies within a highly urbanized watershed. The ponds drainage area covers
107 acres, which can be divided into three subbasins (Figure I.11.6). Subbasin A, which is
served by a closed drainage system, covers 89 acres and drains to the pond through a 4 ft
4 ft box culvert, which is the major drainage inlet to the pond. Subbasin B contributes
Chapter I.11:
219
Figure I.11.6 Storm drain system within Halls Pond watershed, image from A Plan for Halls Pond
Sanctuary.
10 acres of overland drainage from the area immediately surrounding the pond, including
Amory Playgrounds ball fields. Subbasin C contributes 8 acres of drainage from several
streets north of the pond. The pond drains out through a 60-in. circular metal pipe that
runs out to Amory Street and down to the Charles River. The pipe consists of varying
diameters on its route to the river. At times the Charles River backs into Halls Pond.
Groundwater tends to be a fairly stable influence and, in the case of Halls Pond,
appears capable of sustaining the water level at an acceptable level, even during a very
dry summer. The pond is shallow, but no reports have been made about the pond going
dry during summer droughts. Evidence of substantial groundwater inputs was apparent
to the ENSR team. Prior to the routing of stormwater into the pond, groundwater was
probably the main influence on water level and quality. Water quality is still affected by
the groundwater, as evidenced by the iron and manganese oxides in the pond. These are
typically the result of high levels of dissolved iron and manganese in low-oxygen groundwater being exposed to oxygen after seeping into the pond.
The water quality of surface runoff is highly variable in most cases; for Halls Pond
it appears to be the dominant influence on pond hydrology and pondwater quality. Runoff
from adjacent ball fields and tennis courts or sanctuary grounds poses some threat because
of town maintenance practices, but management is entirely feasible. The discharge of
runoff from the stormwater drainage system at Beacon Street, a busy, six-lane commuter
road to Boston, has a major impact on water level and quality.
Improvement of water quality is extremely difficult to achieve in urban situations,
and very costly. Halls Pond has been serving as a small detention system for highly
contaminated urban runoff for decades. The runoff may account for most of the undesirable features of the pond, including the decline of amphibian populations through either
220
Figure I.11.7 Invasive vegetation near outflow structure in Halls Pond. (Photo by Clarissa Rowe.)
toxic reactions or siltation of egg masses during the critical spring breeding season. The
alternative to stormwater management would be diversion of the inflow pipe to the pipe
around the pond, which is achievable in this case.
Restoration approach to Halls Pond

Significant enhancement of the habitat potential of Halls Pond requires a substantial
overhaul, involving diversion or purification of stormwater, possible dredging, and edge
treatment. There was a desire to reintroduce biological communities, including fish, reptiles, amphibians, aquatic invertebrates, and plants, after the water quality improved and
the native vegetation was reintroduced (Figure I.11.7). Although not inexpensive, the
restoration program envisioned by the town is feasible and could markedly improve
conditions in the pond. Properly implemented, such a program can minimize future
maintenance needs, fostering a stable and natural environment for habitat and passive
recreational pursuits.
Wetland between Halls Pond and Amory Woods

The concept of an expanded wetland area within the sanctuary has definite merit, but its
development needs considerable thought and discussion within the context of goals and
site limitations. The suggested area for the new wetland requires substantial grading if it
is to support wetland plants and discourage invasive, non-native nuisance species. Use
of wetland as a fringe around the pond has appeal both for filtering the pond water and
for habitat enhancement. Location of the wetland and its vegetation features will depend
Chapter I.11:
221
to a large extent on future hydrology, which will be affected by any stormwater management or diversion.
Creation or enhancement of wetlands should be viewed within the context of other,
more fixed features such as the pond, garden areas, drainage features, desirable trees, and
access points. Wetland-related goals were achievable in multiple areas because of the
shallow depth to groundwater and the amount of runoff available. Clear goals are essential, along with some priority for potentially competing objectives (Figure I.11.8).
Vegetation management is also critical in wetland areas. Cattails are present and,
although welcome, are expanding; a quantity of jewelweed is present, and the possibility
exists for phragmites to take over in the future.
The Friends of Halls Pond have been the driving forces for the project and the steward
of the sanctuary for over a decade. Once the town completes the project, the Friends group
expects to take over most of the maintenance of the sanctuary again. With the Brookline
Conservation Commission, the Friends of Halls Pond determined the program for the
sanctuary, Halls Pond, and the expanded wetlands into the Amory Woods Sanctuary.
Wetland creation represents an exciting opportunity for community involvement, and
Brown and Rowe designed and organized the effort in a way that would allow volunteers
to reduce the impact to the construction budget. Once the work that needed heavy machinery is completed, the Friends group can do some of the planting, replanting, and maintenance of the Sanctuary. The wetland fringe plants can be installed in small sizes and,
within 3 years, will reach sizes that give form to the area and will begin to provide habitat
value and filtration and aeration to the water. See Appendix A for a list of the wetland
plants that will be introduced for the new wetland planting.
Control invasive vegetation and reestablish native plants

Although the preservation of key landscape features and associated vegetation is viewed
as an important component of this project, current vegetative features require major
alteration to achieve project goals. Change focuses on two key elements alteration of
vegetation to maximize habitat and prevention of nonnative nuisance plant invasions.
Many of the larger, older willow trees need maintenance, but most can be saved even
as the wetland is expanded. Much of the understory, however, is dominated by invasive
species or plants of limited habitat value and needs substantial clearing and replanting.
Although careful design and implementation should establish a fairly permanent vegetation structure, control of existing invasive vegetation and prevention of future successful
invasions will require some form of physical or chemical management technique. Physical
techniques are more labor-intensive but are amenable to directed volunteer labor and do
not carry the stigma associated with many chemical techniques. If physical techniques are
employed, great care must be given to removing all roots so that the plants do not grow
back. It should be kept in mind, however, that chemical controls have progressed greatly
over the last decade to a point at which highly selective and narrowly applied chemicals
could be used to open target areas or to control future growth (Figure I.11.9).
Community process
The Town of Brookline was strongly committed to a lengthy and open community process
during both the planning and the design phases of the work. During the initiation of the
Sanctuary Plan, 20 public meetings about the sanctuarys future were held. During the design
phase, 12 public meetings were held. The majority of the public was in favor of the project,
but a vocal minority, which included some Brookline Town Meeting members, was opposed
to it. Their opposition focused on the wetland being a swamp, a mosquito-breeding center
Figure I.11.8 Plan for Halls Pond produced by Brown and Rowe, Inc.
222
Chapter I.11:
223
Figure I.11.9 Boardwalk in wet area at Halls Pond. (Photo by Clarissa Rowe.)
that would cause disease. Dog walkers also commented that the new sanctuary would
exclude dogs in order to protect the other fauna. Some opposition to the project resulted
from the proposal to close off a shortcut from the elite, surrounding neighborhood to a
nearby elementary school, but this opposition was usually cloaked in the swamp as a
disease-ridden hazard language instead of the inconvenience of having to walk an additional four blocks to school.
Environmental issues at Hills Pond, Menotomy Rocks Park,

Arlington, Massachusetts
The goals of this project are similar to those for Halls Pond, except that Menotomy Rocks
Park is a public park, not a nature sanctuary. Hills Pond, part of a wonderful system of
water bodies in the Alewife Reservation, is a stopping place for many species of migratory
birds. Pedestrians, dog walkers, children, fishermen, skaters, and passive nature lovers
use the town park year-round.
In the 1890s, Farmer Hill had turned a swamp on his farmland into a manmade pond
of about 3 acres (Mattheisen, 1996). The town acquired the pond in 1924, and it quickly
became a centerpiece for winter skating and summer fishing activities. By the 1980s, the
pond was in an advanced stage of eutrophication, a condition caused by the quality of
the storm runoff from the surrounding residential community and by the presence of
invasive, nonnative vegetation in the pond (Metcalf and Eddy, 1986).
Restoration approach to Hills Pond

Town entities spent considerable time determining how to solve the water problems in
the pond. Conservation Commission members pushed to have new stormwater technology solve the problems of runoff. A grease and grit Vortechnics trap (Figure I.11.10) and
daylighted stone swale were two elements of the new technology. Both elements cleaned
the storm runoff water before it reached the pond itself.
Unfortunately, park users and abutters were not consulted during the engineering
design process. When the final construction drawings were reviewed by the park users,
224
Figure I.11.10 Vortechnics Box installation at Hills Pond. (Photo by Clarissa Rowe.)
Figure I.11.11 Park users consult with Town Selectmen of Arlington. (Photo by Clarissa Rowe.)
vocal opposition surfaced (Figure I.11.11). The large stone runoff swale was to be located
in one of the only flat, sunny, usable areas of the park. The swale that would slow the
runoff before it reached the pond was about 35 ft long and 8 ft wide, and lined with large
rock boulders. Park users thought that the rocky swale would be an ugly, dangerous
addition to the heavily used park area.
Dredge Hills Pond

In the early 1990s, the town started dredging the 3-acre pond (Figure I.11.12). Dredging
the pond was the major task of the project. Significant problems developed during the
dredging project. In the winter of 1991, the project was halted and the park was closed to
the public. Neighborhood opposition intensified during the dredging construction because
Chapter I.11:
225
Figure I.11.12 Hills Pond dredge. (Photo by Clarissa Rowe.)
Figure I.11.13 Hills Pond dredge on park areas. (Photo by Clarissa Rowe.)
the town closed the park. Rainy weather badly stalled the work. Soon it was apparent
that the initial testing underestimated the depth of the dredge, and piles and piles of wet
dredge were placed all over the well-used level areas of the park. No plans were included
in the construction package for removing any of the dredge from the park (Figure I.11.13).
The engineering company had recommended comprehensive testing for the depth of the
material in the pond, but the town decided that this testing was not within its budget and
thus authorized only limited testing.
During that winter, the Arlington selectmen appointed a 93-citizen committee to study
the dredging problem (Figure I.11.14). Despite its size, this committee worked with the
town entities to redesign the project: to move the filtering wetland area into a corner of
the pond; to assist in the construction overview of the project; and to find a loam contractor
that would remove the dredge for his loam operations at little cost to the town. The Friends
226
Figure I.11.14 Hills Pond fully dredged. (Photo by Clarissa Rowe.)
of Menotomy Rocks Park was born and remains active in advocating for the park in a
climate of municipal belt tightening.
Luckily for the Hills Pond project, the pond dredge was organic and clean because
the primary drainage into the pond was from the surrounding residences and residential
streets. The only notable contaminant in the water was winter road salt, which could be
easily leached out of the soil. Groundwater was not a factor in Hills Pond. Several small
springs do appear to feed into the pond, but they do not really affect the water level. After
the dredging project, the pond had to be filled with water from a newly dug well in the
park. A pond outfall to control the pond water level was installed (Figure I.11.15). A water
line from the well to the pond was installed so that the pond could be fed and aerated
during the summer months. The Friends of Menotomy Rocks Park organized, helped pay
for, and installed the wetlands planting for the pond (Figures I.11.16, I.11.17, and I.11.18).
In the spring of 2001, only 6 years after the dredging was completed, Hills Pond was
treated again with various herbicides to control invasive weeds. Pond maintenance needs
to be an ongoing line item in any municipal maintenance budget.
Install filtering wetland

During the redesign process, the wetland was relocated into a corner of the pond itself.
The purpose of the wetland was to slow the stormwater from surrounding hills and have
the sediment in the runoff settle into a settling basin, allowing the wetland plants to
filter or change the nature of the sediment before the water entered the larger pond. An
earthen berm, which is usually just above the water level of the pond, separates the stilling
basin from the rest of the pond (Figure I.11.19). The Friends group now has years of water
data that document the success of the basin in cleaning the water before its goes into the
bigger pond. After dredging, the pond was restocked with sunfish and bass so that it
remains a popular fishing spot. The Friends group maintains the basin by weeding out
any invasive plants, like purple loosestrife, that arrive with the birds. The original planting
plan also included cattails, which are pruned back yearly as they attempt to spread into
the larger part of the pond (Figures I.11.20 and I.11.21).
Chapter I.11:
227
Figure I.11.15 Hills Pond outflow chamber. (Photo by Clarissa Rowe.)
Figure I.11.16 Friends of Menotomy Rocks Park install bioengineering logs and plants. (Photo by
Clarissa Rowe.)
228
Figures I.11.17 Planting the wetland at Hills Pond. (Photo by Clarissa Rowe.)
Figures I.11.18 Planting the wetland at Hills Pond. (Photo by Clarissa Rowe.)
Conclusion: Understanding the natural systems

The lessons learned from these two urban wetland projects are myriad. Most concern
making sure that any municipality engaging on a pond or wetland restoration has an
adequate design, construction, and maintenance budget for the work involved.
1. The municipal pond owners should test the pond sediments and surrounding soils
at the beginning of any design or engineering effort. Both towns in this case study
tried to cut costs during the investigative stage of the project. In both instances,
this approach cost them substantial money later in the project. In one project, the
pond sediment was actually an asset; it could be sold to a maker of topsoil because
it was so rich in organic content and free of contaminants. In the other project, the
Chapter I.11:
229
Figure I.11.19 Earthen berm separates the stilling basin from Hills Pond. (Photo by Clarissa Rowe.)
Figure I.11.20 Wetland plantings at Hills Pond thrive 3 years later. (Photo by Clarissa Rowe.)
pond sediment exceeded state standards for contamination and its removal was
cost-prohibitive. This finding caused a major rethinking of the project.
2. The pond projects, because they were in urban/suburban areas and were well
watched, needed the services of a pond scientist, a wetlands scientist, and a landscape architect throughout the design process. The balance of the disciplines was
necessary to cover all aspects of the projects. In Hills Pond in Arlington, the town
again tried to save money by hiring only a pond scientist and civil engineer. The
resulting design put an innovative engineering solution right in the middle of one
of the most heavily used areas of a popular public park. A landscape architect on
the team would have suggested a better location for the swale. The Town of
Brookline did not have the funds to continue the services of the landscape architect
and scientists during the construction phase. The mix of professions meant that
230
Figure I.11.21 Topographic plan for Menotomy Rocks Park. (Prepared by Tom Gonsiorowski of the Friends of Menotomy Rocks
Park.)
Chapter I.11:
231
the project costs were a lot higher than the standard design and engineering
services.
3. In both towns, the clients were neither familiar with the delicate nature of wetland
restoration and pond science, nor did they understand how the natural systems
of their ponds really worked. Both projects were done under the bailiwick of the
Departments of Public Works. Neither department had extensive experience with
pond maintenance and construction around delicate wetland areas. In both cases
the volunteer boards of the conservation commissions spearheaded the renovation
projects but then ran out of or had no funds when the project required almost fulltime involvement on the towns part.
4. A comprehensive vegetation management plan should have been developed as
part of the design process.
5. The public process is an important step in educating the general public about the
long-term health of their water resources. In Arlington, where the public process
was not well advertised, the public outcry set the project back by months. In
Brookline, the public process required an additional three years of public meetings,
and even after that lengthy process, there was still a vocal majority against the
swamp project. Because of these two public processes, both ponds now have
very active and enlarged Friends groups that act as the chief stewards for the
continued health of their water bodies. Both still have an uphill battle convincing
the financial stewards of their towns to invest in the continued maintenance of
these municipal assets. Taking care of these town ponds is a great deal more
complicated in the early 21st century than it was when Farmer Hill and a few
mules or horses created Hills Pond in the mid-19th century and which, then, [the
pond] mostly looked after itself, as one of the Friends of Menotomy Rocks Park
said recently.
Acknowledgments
Clarissa Rowe, one of the founders of Brown, Richardson and Rowe, Inc., and a resident
of Arlington, has been actively involved in providing pro bono services for the Hills Pond
project. Dr. Ken Wagner, the pond scientist from ENSR and the Engineer of Record for
the town, continued to work closely with the Friends group after ENSRs contract with
the town expired. The Friends also set up an ongoing public education program for the
park and the pond. The Friends group won the 1995 North American Lakes Management
Societys award for volunteer technical excellence, with special recognition going to
Friends leaders Benjamin Reeve, Town Manager Donald R. Marquis, and Ms. Rowe.
ENSR (155 Otis Street, Northboro, MA 01532) provided the pond science, wetland
creation, and engineering services for the Halls Pond project. Ken Wagner was the pond
scientist and Dr. Dennis Lowry was the wetland specialist. Emilie Stuart and Nina Brown
of Brown, Richardson and Rowe, Inc., were the editors for this chapter.
Literature cited
ENSR, Notice of intent for Halls Pond Sanctuary restoration, prepared for the Brookline Dept. Public
Works, 1999.
Giezentanner, B. and Eunson, D., Massachusetts Audubon Soc. Environ. Extension Serv., A plan for
Halls Pond Sanctuary, prepared for the Town of Brookline Conservation Commission and
Friends of Halls Pond, Brookline, MA, 1996.
Mattheisen, D., Menotomy Rocks Park A centennial history, prepared for The Friends of Menotomy
Rocks Park, Arlington, MA, 1996.
232
Metcalf and Eddy Engineers, Feasibility study of Hills Pond, Menotomy Rocks Park, report to the
town of Arlington, Arlington, MA, Dec. 1986.
Metcalf and Eddy Engineers, Final report to town of Brookline on diagnostic/feasibility study of
Halls Pond, Brookline, MA, Nov. 1986.
Chapter I.11:
Appendix A
Halls Pond wetland plant list
These plants will be planted during the Halls Pond Restoration program:a
Arrowhead (Sagittaria latifolia)
Atlantic white cedar (Chamaecyparis thyoides)
Arrowwood (Viburnum dentatum)
Buttonbush (Cephalanthus occidentalis)
Burreed (Sparganium americanum)
Bluejoint grass (Calamagrostis canadensis)
Blueflag iris (Iris versicolor)
Bulrush (Scirpus validus)
Black willow (Salix nigra)
Wood reedgrass (Cinna arundinacea)
Chokeberry (Aronia melanocarpa)
Cardinal flower (Lobelia cardinalis)
Cinnamon fern (Osmunda cinnamomea)
Highbush blueberry (Vaccinium corymbosum)
Mountain holly (Nemopanthus mucronatus)
Marsh fern (Thelypteris palustris)
Fowl mannagrass (Glyceria canadensis)
Marsh marigold (Caltha palustris)
Meadow sweet (Spirea latifolia)
Sensitive fern (Onoclea sensibilis)
Pickerelweed (Pontederia cordata)
Red maple (Acer rubrum)
Royal fern (Osmunda regalis)
Speckled alder (Alnus rugosa or A. incana)
Spicebush (Lindera benzoin)
Sedges (Carex crinita and C. stricta)
Sweetflag (Acorus calamus)
Sweet gale (Myrica gale)
Shadbush (Amelanchier canadensis)
Sweet Pepperbush (Clethra alnifola)
Soft rush (Juncus effusus)
Winterberry holly (Ilex verticillata)
Water lily (Nuphar luteum)
Water plantain (Alisma trivale or A. plantago-aquatica)
Wild raisin (Viburnum cassinoides)
a
Wetland plant list by Dennis Lowry of ENSR and Clarissa Rowe of Brown and Rowe, Inc.
233
234
Response
Project development through concerned citizenry
In urban settings, ponds become valued for the sense of space and solace they provide in
contrast to the landscapes surrounding them. For just as stormwater rapidly courses off
impervious pavement, urban lifestyles also seem to rush by atop such surfaces. As Clarissa
Rowe discusses, water sensitive design works best from a bottom-up process of engaging
concerned citizenry at early, and indeed all, stages of project development. The two ponds
described in this chapter suffer from the same problems as those covered by Nicholas
Pouder and Robert France in Chapter I.16, and by Thomas Benjamin in Chapter II.4. As
in those other studies, Rowe takes pains to state that ultimate solutions lie in integrating
site-specific restoration measures within a framework of more comprehensive watershed
management. The two case studies outlined in this chapter demonstrate the strength of
commitments of local stewardship and the role of conservation commissions in mobilizing
restorative actions.
Another interesting element in this story is the manner in which created wetlands
were used to help mitigate inflowing problems; i.e., construction of a berm in the inflow
embayment to create a small wetland comprising a settling basin and filtering plants for
treating the nutrient-laden water before it merges with that of the pond itself. This
approach of identifying and then targeting an area of concern is similar to that promoted
by Gail Krantzberg and Judi Barnes in Chapter II.6 with respect to the entire Great Lakes
Basin.
Finally, Rowes recounting of the what one would have hoped to be antiquated
perception of wetlands by pond abutters as being disease-ridden swamps harboring all
sorts of ills and hazards, indicates that no matter how clever our water sensitive designs
may be, public education is necessary. In other words, if the oft-repeated mantra of the
real estate profession is, location, location, location, that of the water sensitive design
profession should be, education, education, education.
chapter I.12
Water connections: Wetlands for science

instruction (Wichita, Kansas)
Robert France and Kaki Martin
Abstract
Wetlands provide wonderful opportunities for directly experiencing the workings of
nature. By engaging curiosity with exploration, created wetlands can function as outdoor
laboratories for experiential science instruction. This chapter describes one such project
in Wichita, KS, where a functional wetland is integrated physically into a playground
while at the same time fostering environmental learning. Water from a reflecting pool
surrounding the learning building is circulated through a treatment wetland for purification through nutrient removal before being discharged back into the pool. Another important element of the project is the role water plays as a physical integrator between the
architecture and the landscape components of the site development. Due to the high
visibility of the project, careful attention was paid to plant selection, planting, maintenance,
and the circumvention of grading problems during construction.
Introduction
An anecdote from a recent wetland conference is very illustrative of the problems underscoring the discipline of treatment wetland design. A well-known speaker who was in the
process of updating a major guidebook about wetland construction dodged several questions from the audience concerning the possibility of designing for multiple purposes.
Defending himself, this individual stressed that the guidebook he was developing for the
EPA was concerned with what he referred to as blue-collar wetlands, those for which,
because of obvious space limitations, it was impossible to create functions (presumably
for wildlife benefits, aesthetics, or educational opportunities) other than those intended
as engineering solutions to water quality improvement. Such an attitude, especially coming from an individual charged with the important task of writing an official government
document, is worrisome given its demonstration of limited design vision.
Unfortunately, such an attitude has been the status quo in the construction and engineering of created wetlands for decades. What is particularly alarming is that it obviously
continues to play a blinkering role in restricting vision, as witness to the government
official cited previously. It is time to bury such a pessimistic view of the potential for
innovative wetland creation. In reality, the difference between a utilitarian blue-collar
wetland and a multipurpose white-collar wetland has really much less to do about
1-56670-562-2/02/$0.00+$1.50
235
236
Figure I.12.1 Site plan for Exploration Place, Wichita, KS, showing building surrounded by the
Arkansas River and a reflecting pool connected to a treatment wetland within the playground.
absence of space as it has to do about absence of imagination (France, 2002). Today,

numerous wetlands combine functional utility with other ancillary benefits (e.g.,
Chapters I.10, I.13, and I.14). Specifically, the opportunity for creating wetlands to provide
natural laboratories for science education is immense (Kusler et al., 1998; France, 2002).
For example, the justification for constructing one such school wetland was explained as,
Natural objects inspire a childs curiosity, evoking a passion for learning that stimulates
all the senses (Hammatt, 2001).
Study site description and wetland justification

Located at the confluence of the Little Arkansas and Arkansas rivers, the 98,500 ft2 Exploration Place building dramatically occupies a 20-acre site at the rivers edge adjacent to
downtown Wichita. Exploration Place has become not only a significant addition to the
Wichita skyline but also an innovative learning center that brings together people of all
ages to a site where exploring and creativity of all kinds are encouraged.
Due to the spatial adjacency of the reflecting pool and the building and the unobtrusive
quality of the dam walls at each end of the pool (Figure I.12.1), from many vantage points
one is easily convinced that the Arkansas River flows in its natural course between the
two buildings. This feeling is most notable when crossing the interior bridge from the
Land Building, which houses the ticket booths, administrative offices, caf, and museum
store, to the River Building, which houses the exhibits. The 41,000 ft2 pool wraps the
landside edge of the River Building and meets the river at either end with concrete
retaining walls. During spring flooding or at other times following heavy rains, the water
levels in both the river and the pool rise blending the two bodies of water. Most of the
time, however, water from the river and the constructed pool do not mix but are separated
by the retaining walls on each end such that they only appear to be connected.
The construction of the pool was not complex. The depth of the pool is 3 ft at the
deepest points, with the grade sloping to a maximum depth of 6 in. at the perimeter for
safety purposes. The pool bed is underlain with an 18-in. clay lens over well-compacted
Chapter I.12:
Water connections: Wetlands for science instruction (Wichita, Kansas)
237
earth with 6 in. of sand over the top to unify the surface. The dam walls at either end do
not serve any structural purpose, acting only as retaining walls for the pool. Pools this
large and this shallow, with no inherent circulation, have a well-known tendency to
accumulate significant levels of nutrients and consequent algal blooms (Tourbier and
Westmacott, 1992). Given the importance of visual quality, it was important that the
aesthetic of the pool remain intact and consistent. As a result, water quality needed to
remain high to prevent unacceptable blooms. At first, the architectural team and the client
team considered aerators as their only method to achieve the goal.
Park space
The expansive landscape associated with Exploration Place (www.explore.org) was conceived as a programmatic extension of the inside exhibits and activities. In its physical
form, the landscape plays upon and reinforces the strong forms of the architecture. The
landscape was designed and developed with the missions of Exploration Place at the
forefront: education and scientific exploration for all ages. The designed landscape incorporated, through the spatial relationships between landforms, a compelling landscape that
incorporated the goals and desires of the client. Specific features include the Hill, a
dramatic earth mound that utilized the excavated earth from the building construction to
make a 18-ft-high point on the site for everything from star gazing to kite flying; a generous
picnic area under a grid of trees as well as the earthwork Crescent, which wraps around
the Kids Explore playground. The Crescent is an ideal place from which to view the
playground as well as the 7,458 ft2 wetland that extends out from its center.
The wetland was envisioned as another site feature that could be utilized for educational purposes but also provide a strong aesthetic component to the site, primarily
through the use of more showy water loving plants such as Iris versicolor (blue flag). A
water flow connection between the wetland and the reflecting pool (the other water system
on the site discussed earlier) was not considered in the initial design phases though
their physical proximities clearly led to the final decision to create a functioning, interconnected water system.
Wetland connections
Water is the means by which the strong architectural form of the building and the structured landscape components of the project meet. In physical terms, the broad end of the
misshapen triangular wetland excavation meets the outside of the reflecting pool arc and
is separated by only a bituminous concrete bike path (Figure I.12.2).
Reflecting pools such as that at Exploration Place are notoriously prone to nutrient
buildup and consequent proliferation of algae. Given the summer temperatures to be
expected in this region and the low N:P ratios found in many nearby water bodies, the
reflecting pool will have a high propensity to develop nuisance blooms of nitrogen-fixing
blue-green algae. Even if new nutrient loads are reduced, the pool would still be susceptible to eutrophication due to resuspension of bottom sediments. Given the shallow depth
of the reflecting pool in relation to its length of openwater fetch, empirical relationships
(France, 1997) indicate that wind-induced mixing will be a frequent event.
The other problem of reflecting pools occurs as a result of the death, settling out, and
decomposition of phytoplanktonic algae. To counter the ensuing anoxia, large fountains
are often used to oxygenate the water. In the Midwest and West, these frequently take the
form of expensive water jets that notably detract from the site aesthetics. Such water jets
will also serve to recirculate bottom sediments, thereby possibly sustaining algal bloom
conditions.
238
Figure I.12.2 Image of hydrological connections between the treatment wetland and reflecting pool.
Water from the latter enters the wetland through the right channel, swings around the central berm,
and then exits back to the pool via the large channel in the middle of the photograph. Note how
the building seems to float on the water of the reflecting pool and river. (Color version available
As a result of some brief discussion about the specifics of the site including the size
of each water body and the construction technique used to build the reflecting pool, as
well as the overall goals of the Exploration Place, it became clear that by hydrologically
connecting the two water bodies it would be possible to make a dramatic contribution in
two ways. Not only would reliance on aerators in the reflecting pool be reduced but also
a more substantial educational element to the park would be added by making the wetland
truly functional rather than being merely aesthetic. Our plan is therefore based on
bridging both physically and conceptually the bike path separating the buildingreflecting pool side and the parkland works side. There is a comprehensive literature
demonstrating the role of constructed wetlands in improving water quality (e.g., Campbell
and Ogden, 1999; France, 2002 and references therein). A secondary analysis of nearly 100
case studies (France, 2002) indicates that reductions in the range of 95% for phosphorus
and of 90% for nitrogen are possible in constructed wetlands. By circulating the water
from the reflecting pool, through the wetland, and back into the reflecting pool, water
quality will be improved and the number of aerators reduced.
Wetland design
System elements
Whereas the perimeter form of the intended wetland did not change in the revised process
of connecting the two systems, the interior of the wetland did. In order to increase the
flow length of the water, it was necessary to locate a berm down the center of the wetland.
This resulted in the desired effect of doubling the water flow length and because the berm
originates from the bike path, access was created to allow teachers or museum guides to
actually bring small groups into the wetland (Figure I.12.3).
Long linear wetlands are ideal for nutrient removal due to maximizing opportunities
for contact of water with plants and associated epiphytes. In this case, the berm prevents
flow short-circuiting and therefore goes toward ensuring efficient nutrient stripping. A
Chapter I.12:
239
Figure I.12.3 Construction of the central berm designed to increase the flow path distance for
nutrient removal in the treatment wetland. This area will also function as an avenue to bring students
into the wetland for study. (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
secondary analysis of constructed wetlands (France, 2002) revealed that distances between
60 to 120 ft are sufficient for the effective removal of most contaminants. The central berm
creates a treatment distance of 318 ft for this otherwise small wetland. This is a more than
adequate distance for reducing the amount of nutrients expected in the system.
Water enters the constructed wetland through a PVC pipe that perforates the concrete
wall of the reflecting pool. A small submersible pump is located at the far end of the
reflecting pool to aide in the movement of water from one end of the 584-ft-long reflecting
pool to the wetland outflow pipe. A pretreatment sediment forebay 3.5 ft deep serves to
dissipate the energy of the incoming flow. This allows sedimentation to be confined to a
single region rather than running the possibility of clogging wetland planting beds. The
nearby pedestrian pathway provides easy access for maintenance should dredging become
necessary in the future.
Because a mixture of flooding regimes expands wetland functions, interspersion of
shallow and deep water areas will increase the efficiency of contaminant removal and
wildlife attractiveness. Edges of the pretreatment forebay are planted with narrow-leaved
cattail (Typha angustifolia) and common 3-square bulrush (Scirpus pungens). Water from the
pretreatment forebay moves over a small earthen dam that prevents the water from leaving
the forebay until the water level rises high enough in the forebay to begin to overflow the
dam, sending the water into the shallow wet meadow zone. In this zone water is no more
than 4 in deep. Sedges and flags dominate in this zone, including sweet flag (Acorus
clamus), blue flag (Iris Versicolor), fringed sedge (Carex crinita), and fox sedge (Carex vulpinoidea). The grade drops, allowing the water to gradually become deeper water. Soft rush
(Juncus effusus) and rice cutgrass (Leersia oryzoides) are planted in this transition zone that
culminates in a second pool dominated by a stand of giant burreed (Sparganium eurycarpum). This area marks the midpoint of the wetland circuit. Continuing from the burreed
stand, the grade slowly drops again and the planting is pushed to the edges, allowing for
more open areas of water. Broad-leaved arrowhead (Sagittaria latifolia), arrow arum (Pelandra virginica), spike rush (Eleocaris paustris), and lizards tail (Saurus cernuus) cover the
edges of the outer wetland edge and the berm edge in this zone, allowing for a continued
240
Figure I.12.4 Newly planted wetland showing the importance of grading to plant survival. (Color
area of open water. The end of this linear zone culminates in the final micropool. This pool,
situated at the discharge outlet, further reduces the release of sediments and floating organic
matter back into the reflecting pool. As the case for the pretreatment sediment forebay,
the proximity of the pedestrian pathway will allow for servicing of this area should the
need arise. Soft-stemmed bulrush (Scirpus validus) provides the vegetative cover here. The
berm is planted with a dense shrub layer of button bush and two specimen trees, Quercus
bicolor and Cornus amomun, that will eventually provide some shade. The heavy cover on
the berm not only provides additional wildlife cover but also discourages access by visitors
when specific wetland activities are not occurring. And, in order to stabilize the steep
banks of the wetland, all edge areas were planted with prairie cordgrass (Spartina pectinata).
Planting list and strategy

Our criteria for screening potential wetland plants included:
Demonstrated success in the bioregion of the western plains

Attractiveness a visually diverse palette for humans
A structurally complex habitat and food source to wildlife
Demonstrated performance in improving water quality
Ease of availability
Our selected list contained 2 submerged plants and 17 emergents in addition to 5

shoreline edge or wet meadow plants, and 5 woody shrubs and tree species.
Implementation
Grading
Often, a major factor affecting wetland success is the creation of precise elevations. These
will influence the hydrologic residence time and thereby the performance of the wetland
in terms of water quality improvement. Grading is also important for the survival of
wetland plants, some of which require a narrow inundation range in order to achieve
maximal growth and, therefore, phytoremediation (Figure I.12.4).
Chapter I.12:
241
Close supervision of site work is essential for accurate adherence to design guidelines.
The best plans in the world may never bear fruition if the site work crew does not follow
plans properly or if less-than-ideal weather conditions during construction prevent the site
work from being as precise as is required. In the present case, torrential rains led to saturated
ground conditions in Wichita during the final grading of much of the site, including the
wetland area. More than once, the weather not only delayed work but also caused work that
had originally been done correctly to either become substandard or be lost all together. For
example, one corner of the wetland became silted up during one such storm event. Although
the final grading was correct and had been approved, the subsequent rain and fact that the
slopes of the wetland had yet to be planted or sufficiently protected against erosion meant
that considerable corrections needed to be made before the clay liner was laid down. Further,
weather conditions and a well-thought-out work schedule are important not just for the
moving and compacting of earth to happen efficiently and correctly but also for the installation of the clay liner. If it is too hot and dry, the liner can dry up and become cracked if water
is not almost immediately placed in the basin. Conversely, if it is too wet, the manipulation
of the clay liner can be unnecessarily difficult for work crews.
Planting and maintenance

Although landscape construction crews will have considerable experience with terrestrial
trees, their knowledge about the particulars of successful planting of wetlands is often
limited. To counter this possible lack of experience, we provided a brief outline for the
contractor undertaking the plantings.
Timing
For seeding, timing is not particularly critical. With nondormant spring planting, however,
timing of planting may be one of the most important elements in project success, such
that if project delays negate spring plantings, it is often wise to delay planting for an entire
year rather than attempt to do so one month late, which is outside the ideal success
window. For fall planting, dormant plants will have greater overwinter survival and be
ready to grow in the early spring during freshets.
Plant material
All nondormant plants should be healthy with vigorous leaf, stem, and root systems, free
of disease or substantial damage. Potted plants should have well-developed root systems
such that if they can be easily removed from their containers they should be rejected.
Storage
During transit and prior to installation, plants must be kept covered, moist, cool, and out
of the elements. If not planted on the day of delivery, all plants should be draped in moist
matting such as straw. Tarps may need to be erected during the sorting period to block
sun and drying wind in order to prevent desiccation. Staggered arrivals will allow for
planting to occur within 2 days. For dewatered plantings, it is essential that the entire
process be completed within 5 days.
Layout
Demarcated areas must be clearly laid out with identifying stakes (including upslope
locations) prior to arrival of plants so as to minimize vegetation storage time. Obviously,
a good understanding of the expected water depths will enable identification of the correct
zonation of planting areas tailored for each species.
242
Spacing
A spacing interval of about 3 ft for emergents and 1.5 ft for small flowering plants should
be adopted to allow for spread. The planting density and stock size should be increased
to compensate for expected losses in high-energy systems (the alternative in the present
situation is to reduce, or better yet, stop, the flow of water from the reflecting pool until
the plants are healthily established). Some plants such as Typha are aggressive and may
need to be physically restricted within buried in situ contained pots. Plantings should be
extended upslope at least 3 ft above the expected water level.
Fertilizer
Some experts favor fertilization for all wetland plantings; the decision should be based
on anticipated water nutrient levels, expected rate of vegetative spread, and depth of
water. Most deepwater plantings probably do not require fertilization. For others, placement of a tablet at the bottom of the planting hole rather than being applied aqueously
to the entire area will limit development of algal blooms. For nondormant plants, the
mixture should be at least 19% N, 6% P, and 12% K.
Water
There are pluses and minuses for dewatered plantings over plantings in the expected
water levels. If planting in dewatered conditions, there may be a tendency (1) to place the
materials at soil depths that are too shallow and/or (2) not to firmly plant into the soil
such that once water is introduced, the plants will float free. Peat pot plants should be
inserted at least 1 in. (but not more than 2 in.) deeper than what they were grown at in
the nursery to ensure root/rhizome coverage. In most cases, 2- to 3-in. (not more than
4-in.) depths are ideal for littoral plantings. Gradual flooding is required to protect the
liner, limit suspension of sediments, and allow plants to become acclimatized. Plantings
should be allowed to become well established for at least 6 weeks before contaminated
water is introduced into the system and access permitted for herbaceous animals. Some
plants may require water fluctuations, but limit drawdown periods until they become
well established.
Plant establishment period

In most traditional landscaping, the plant establishment period is typically one year,
during which the contractor guarantees the survival of the plant material. Such a procedure should also be insisted on with respect to wetland plantings (Dunne et al., 1998).
The one-year establishment period should extend from the time of planting through
August of the second summer (i.e., generally more than a 365-day year). At the end of
this period, areas with less than 20 to 50% aerial cover or less than 50 to 80% plant survival
shall be deemed deficient and in need of replacement plantings or other maintenance.
Sometimes it is possible to assess wetland planting success at a much earlier date (see
below). The desired goal expected in the subsequent year after wetland planting should
be the establishment of viable populations of a wide variety of species. Once the plantings
become established, it is important to recognize that succession will take over and little
or no subsequent management will be required.
Herbivory
Once emergents are established, the attractiveness of the area to geese will be reduced. In
the meantime, however, it may be necessary to install a protective lattice network of strings
tied to wooden stakes in areas where waterfowl and rodents may be present.
Chapter I.12:
243
Drawdown
In future years, a spring drawdown may be a good idea to kick start the system by
allowing increased light penetration to reach the plants.
Project tending
As quoted in the wetland design manual of France (2002): Failed wetland creation/restoration projects abound due to opaque goal aspirations, poor engineering and ecological
design, inadequate background data collection, absence of adaptive management, the
erroneous perception that project termination occurs with planting, and a cavalier process
of critical evaluation; and Wetland creation by no means ends with simply filling in a
hole with water over which plants seeds are sprinkled or vegetation is hand-planted.
Therefore, the following essential duties need to be addressed for successful implementation of any wetland creation project:
Long-term responsibilities and funding
Routine schedule for cleaning and checking inlet and outlet water control installations and pumps
Mowing and embankment inspection
Depth of sediment accumulation and maintenance schedule for dredging the sedimentation forebay
Water level and hydrological budget recalculation once emergent plants become
mature and begin to restrict water flow (their intended job for chemical cleaning)
Periodic water quality monitoring of both reflecting pool and treatment wetland
Periodic survey of plant species composition, density and viability, and overall
wetland health
Additionally, plant resources and availability were an issue that the landscape contractor needed considerable guidance with. Plant resources were researched and determined by the landscape architect. In this case, bare root plant material was shipped nextday air from the Midwest. This required coordination and precise scheduling on both ends.
Replanting
For the present case, a diminishing project budget did not allow for weekly or sometimes
even bimonthly visits during construction. Trips were planned for efficiency. In some cases
this meant not being on-site to oversee significant moments of the construction process
from beginning to end. The planting of the wetland was one such moment. Approximately
one month after the planting was done, it was obvious that there had been significant
plant loss. Although the reason for the failure is not exactly known, it is believed that the
wetland was not filled with water quickly enough to support that plant health. Of the
approximately 724 plants placed, 500 required replacement. By the time all parties agreed
to do a second planting, plant availability had dwindled on some of the original specified
species. In this case, three substitutions were made. The second planting proved to be far
more successful, and a digital camera sent daily images to the designer to ensure proper
progress.
Post-construction
Within days of the flooding of the wetland, even without a very high plant survival rate,
the wetland was active. Insects and birds of varying types were almost immediately
244
audible and visible in the wetland. Today, the wetland is a functionally healthy and
aesthetically visible component of the overall site, and has become a popular destination
for students and their teachers or parents.
Future education program

With time, this wetland, primarily designed for water quality improvement, could play
an additional role in outdoor education as part of the Exploration Place program. One
important aspect characteristic of educational wetlands is the manner in which they bring
students into a close relationship with nature. It is easiest to imagine this occurring at the
present site if there were a raised boardwalk partway along the central berm, perhaps
augmented with an overhanging observation platform. A series of small step-down platforms along the complete length of the wetland could offer students access to the water
for chemical sampling, in order to monitor how effective the system is in reducing nutrient
concentrations.
Second, although often underfunded, creation of an education program utilizing illustrations and accompanying text on interpretive signs can be a good way to connect people
with wetlands. In the present case, such outdoor education could be supported by establishing a teaching program inside the museum. Information such as brochures, slide
shows, videos and models, and nonstructural elements, such as school science projects
and guided tours, would go a long way toward promoting the goals of this project.
Although most interpretive programs do a reasonable job in conveying information about
the plants and animals present in wetlands, there is a need, especially in cases such as
this, of raising public awareness about the wider benefits that wetlands contribute to
watershed processes, in particular with respect to their function as purifying kidneys in
water quality improvement.
Literature cited
Dunne, K.P., Rodrigo, A.M., and Samanns, E., Engineering specification guidelines for wetland plant
establishment and subgrade preparation. Wetlands Res. Prog. U.S. Army Corps Engineers,
1998.
France, R.L., Landwater linkages: influences of riparian deforestation on lake thermocline depth
and possible consequences for cold stenotherms. Can. J. Fish. Aquat. Sci. 54, 1299, 1997.
France, R.L., Designing Wetlands: Principles and Practices for Landscape Architects and Land-Use Planners.
W.W. Norton. In press, 2002.
Hammatt, H. 2001. Mind games. Landscape Architecture Magazine, Jan. 2001.
Kusler, J.A. et al., Guidebook for Creating Wetland Interpretation Sites Including Wetlands & Ecotourism,
Assoc. State Wetland Manag., 1998.
Tourbier, J.T. and Westmacott, R., Lakes and Ponds, Urban Land Inst., Washington, 1992.
Chapter I.12:
245
Response
Project development of interpretive wetlands
As the project described by Robert France and Kaki Martin in this chapter illustrates, an
incredible didactic role can be played by water sensitive design. Whether approached as
small-scale demonstration projects such as that described by Nicholas Pouder and Robert
France in Chapter I.16; as incredibly beautiful projects at high-visibility sites such as those
presented by Glenn Allen in Chapter I.13, and Thomas Liptan and Robert Murase in
Chapter I.6; or as large-scale, completely functional projects with detailed interpretive
signage, such as that outlined by Catherine Berris in Chapter I.9, water is an element that
easily lends itself to instructional use for learning about how the environment functions
and how humans interact with nature. The opportunities for developing interpretive
wetlands for environmental education are enormous and something sorely needed to raise
the image that many in the public still have about such noxious swamps of ill repute
(see Chapter I.11 by Clarissa Rowe).
Another salient message put forward by France and Martin which is of importance
to water sensitive designers, is the sobering realization that no matter how good one
imagines a design to be, unless time is spent carefully supervising the contractors physical
implementations of that design upon the landscape, the best intentions can fall far short.
In this respect, project development from conception to final implementation (i.e., the
cradle-to-grave mentality endorsed by those dealing with the environmental sustainability of commercial products) becomes critically significant for measuring lasting project
success.
chapter I.13
Constructed wetlands and stormwater

management at the Northern Water
Feature (Sydney Olympic Park)
Glenn Allen
Abstract
The redevelopment of the Homebush Bay area is one of the largest urban/industrial
renewal projects in Australia to date and has resulted in a sporting, commercial, and
residential area of exceptional quality. The Homebush Bay site for the Sydney Olympics
constitutes 770 hectares (1,900 acres) of what were once originally mudflats and mangrove
wetlands. Over the last 100 years, the area had been largely taken over by landfill and
industrial operations. The master concept design, by Hargreaves Associates (with the
Government Architect Design Directorate), focused on environmental rescue recreating
the landscape for public use. One of the major gestures of that reclamation is the use and
reuse of water on the site (the Blue Move). Central to this move is the capturing of the
sites stormwater runoff, its storage and cleansing in a series of water quality control ponds
and created wetlands, and its reuse in the recycled water system of the site. Hargreaves
Northern Water Feature is a key component of this water cycle infrastructure and embodies
the water quality goals of the environmentally sustainable development.
Introduction
The redevelopment of Sydneys Homebush Bay for the Sydney 2000 Olympics is one of
the largest urban/industrial renewal projects in Australia to date, and it has resulted in a
sporting, commercial, and residential area of exceptional quality. Hargreaves Associates
master concept design for the site focused on environmental rescue to reclaim what
was an abused and abandoned industrial landscape for public use. One of the major
components of that reclamation is the capturing of the sites stormwater runoff, its storage
and cleansing in a series of water quality control ponds and created wetlands, and its
reuse in the sites recycled water system (Figures I.13.1 and I.13.2). The Northern Water
Feature is a key element of this water cycle infrastructure and embodies the water quality
goals of this environmentally sustainable development.
The 770-hectare (1,900-acre) site originally consisted of mudflats and mangrove wetlands. Over the last 100 years, however, the area was largely taken over by landfill and
1-56670-562-2/02/$0.00+$1.50
247
248
Figure I.13.1 The Northern Water feature. (Photo by John Gollings. Courtesy of Hargreaves Associates.)
Figure I.13.2 Aerial view of the Northern Water Feature with Olympic Plaza beyond. (Photo by
Olympic Coordination Authority. Courtesy of Hargreaves Associates.)
industrial operations, including the State of New South Wales Abattoir, the Newington
Landfill, and the New South Wales State Brick Works. The original mangrove and wetland
edges had all but disappeared by 1996. In addition, an indicator species of environmental
quality native to the area, the green and gold bell frog (Litoria aurea) has over the last few
decades dwindled to tiny numbers (50 individuals on the Homebush Bay site). Its habitat
Chapter I.13:
Constructed wetlands and stormwater management
249
Figure I.13.3 Master concept design The Red Move. (Courtesy of Hargreaves Associates.)
is now concentrated in this and other similarly abandoned industrial sites, and it has been
declared an endangered species.
The master concept design (1997) for the site was developed by Hargreaves Associates
in association with the Government Architect Design Directorate for the Olympic Coordination Authority (OCA). The design created for the Homebush Bay site included an
Urban Core of 300 hectares (741 acres) of major Olympic venues, exhibition, and commercial facilities united by a major act of place making. Central to the master concept design
are three key design moves that give form and coherence to the projects public spaces.
The Red Move is Olympic Plaza, the central urban space. A 9.5-hectare (23.5-acre)
open space addresses the major venues and buildings. This paved open space flows
seamlessly like a bold carpet over the heart of the public domain, uniting the site, and
accommodating huge crowds. This is now the heart of Homebush Bay, a civic place of
arrival, a place of ceremony and procession, and one of Australias greatest civic spaces
(Figure I.13.3).
The Green Move is a landscape framework of fingers and parks that stretch through
the site, connecting Olympic Plaza directly to the surrounding Millennium Parklands.
Five eastwest green fingers, each with its own distinctive character that range progressively from the urban to the natural, were developed. In addition, a large, open green
park was created as a counterpoint to Olympic Plaza, preserving significant existing stands
of eucalyptus and fig trees (Figure I.13.4).
The Blue Move is the use of water on the site, made visible as an ordering element
of the urban core. The Fig Grove at the center of the Urban Core and the Northern Water
Feature at the north end of Olympic Plaza celebrate the collection, cleansing, and reuse
of stormwater runoff while symbolically connecting the site and its systems to the wetlands, creek, and river systems beyond (Figure I.13.5).
Crucial to the Blue Move is the OCAs development of an approach to the management
of drainage and stormwater as part of an overall water cycle management strategy for
Homebush Bay. The strategy deals specifically with those aspects of the water cycle related
to the generation, treatment, and disposal of stormwater. Its objectives included the facilitation and procurement of the water cycle infrastructure; the implementation of ecologically
250
Figure I.13.4 Master concept design The Green Move. (Courtesy of Hargreaves Associates.)
Figure I.13.5 Master concept design The Blue Move. (Courtesy of Hargreaves Associates.)
sustainable guidelines and practices; the use of innovative and technically feasible
approaches; and the completion of this infrastructure within the designated time frame
and budget.
The OCA mandated that ecologically sustainable development be the underpinning
of the environmental philosophy of the Sydney 2000 Olympics as a whole, and the Homebush Bay development in particular. The environment strategy for Homebush Bay stated
that its identified goals were as follows:
The conservation of species

Ecosystems
That the remaining natural ecosystems of Homebush Bay be protected throughout
the ongoing use of the site
That the work incorporate appropriate stormwater management practices to ensure
the protection of the natural ecosystems of Homebush Bay
Chapter I.13:
251
That the new landscape works enhance the native species biodiversity of Homebush
Bay
People
That after redevelopment, Homebush Bay offer a high quality of life to those who
live or work at the site, and a highly desirable recreation destination
That Homebush Bay become recognized as a benchmark in the evolution of environmentally sensitive urban development
Conservation of resources
Water
That the new works at Homebush Bay incorporate a reduction in the demand for
potable water from Sydneys main supply
That a reclaimed water scheme be instituted at Homebush Bay
Energy
That the Homebush on-site works minimize the long-term increase in demand of
energy from sources that are nonrenewable or emit greenhouse gases in energy
generation or consumption
Construction materials
That the Homebush Bay works minimize the use of materials that deplete natural
resources or create toxic pollution in their manufacture, use or disposal
Topsoil
That importation of topsoil to the Homebush Bay site be minimized
Pollution control
Air
That the on-site works minimize negative impacts on Sydneys air quality and avoid
ozone-depleting substances
Noise
That measures be taken to minimize the impact of noise at the Homebush Bay site
Light
That the on-site works minimize the impact of night lighting on both environmental
conservation areas and residential areas
252
Water
That the on-site works result in improvement in the quality and quantity of water
entering Homebush Bay and the Parramatta River from the site
That the natural groundwater be recharged wherever possible, and that all stormwater be treated for removal of gross pollutants and nutrients
Soil and sediment

That the on-site works remediate the results of polluting activities of the past and
ensure the protection of soil and sediments within the developed area
Waste management
That all development maximize the appropriate use of recycled material and reduce
waste generation
The Homebush Bay water cycle strategys aims, based on the environment strategy,
were, then, to:
Achieve the drainage design standards necessitated by the creation of the Public
Domain
Integrate stormwater pollution control structures into the drainage system
Develop affordable and economical solutions
Visually integrate drainage structures into the built environment
Preserve and enhance landscape and heritage values
Provide a safe and habitable environment for both local species and site users
As part of this effort, Hargreaves Associates brought together a multidisciplinary
design team to develop the concept and the design and undertake the construction of the
Northern Water Feature. Critical to the success of the project was addressing a number of
significant design challenges, including treating runoff from a 100-ha catchment of potentially high-peak population density, particularly during the Olympics; reducing average
annual pollutant export from the Public Domain catchment prior to discharge into a
receiving waterway by 70 to 90%; creating new habitats for threatened and endangered
species (Litoria aurea) in an area of contaminated soils and landfill reuse; and minimizing
the environmental impact on the adjoining Haslams Creek and other sensitive neighboring
sites, such as the Royal Australian Showgrounds (Figure I.13.6).
Since the initial formulation of the drainage and stormwater management strategy,
the original concepts for stormwater infrastructure and, in particular, the water quality
control ponds to be located at the outlets of the western, eastern, and southern catchments
were refined. The team further developed the strategies and systems as the design transformed the former western catchment water quality control pond into the dominant
landscape element that is now the Northern Water Feature.
Physically, the Northern Water Feature frames the northern end of Olympic Plaza,
stepping down in arching granite terraces to meet a newly created wetland at the edge
of Haslams Creek (Figure I.13.7). Here, 10-m-high arcs of water fan down the terraces,
making visible the cleansing through the marsh of the sites stormwater runoff
(Figure I.13.8). The challenges of the stormwater management strategy have been met in
a newly created freshwater marsh and transformed into an emblem of the environmental
Chapter I.13:
253
Figure I.13.6 The water cycle management strategy for Homebush Bay. (Courtesy of Hargreaves
Associates.)
strategy of Homebush Bay. The water feature provides a series of curvilinear fingers with
a number of open water areas, protruding headlands, and long continuous stretches of
macrophytes.
The primary stormwater quality treatment features of the Northern Water Feature
include the creation of a pool and wetland system (Figure I.13.9) with a low-permeability
clay liner to keep the water separated from the contaminated fill and the groundwater
beneath. Excavation for the wetland involved the recycling and reuse of these highly
variable landfill wastes and construction fill and the installation of a deep leachate system
using a custom-developed method of installing polyethylene liners (Figure I.13.10). The
creation of the wetland included construction of a deep inlet pond with open water and
fringing macrophytes, with submerged inlets to the inlet pond; the splitting of frequent
flows up to approximately a three-month average recurrence interval (ARI) runoff and
discharge at the head of the two arms of the inlet pool (to promote water circulation in
the inlet pool) (Figure I.13.11). It also involved the construction of a deep outlet pool with
open water and fringing macrophytes as well as the rehabilitation of the existing spillway.
The design provided the provision to pump first flush to the brickpit storage pond for
reuse, to recirculate water from the outlet pond to the inlet pond, to supply makeup water
to the fountain, and to irrigate the surrounding landscape. Efficient water and water plant
management necessitated the creation of control structures that enable water levels to be
controlled during macrophyte establishment and subsequent maintenance, and the provision of a grassed overflow spillway in the event of severe blockage of the primary outlet
spillway during major storm events (Figure I.13.12).
In addition, the feature includes the creation of some of the deepest continuous
deflection separation system (CDS) units ever built to intercept gross pollutants from the
catchment. These large, nonmechanical, nonblocking, high-flow rate, screening chambers
strain out the larger pollutants from the runoff and are located 12 m below the plaza just
upstream from the storm drain outfalls into the wetland. The units are configured so as
to use the water flow of the runoff event to control the water-borne pollutants in a circular
flow within the unit, settling them out into a sump or floating to the top of the water
within the unit.
254
Figure I.13.7 Olympic Plaza and the Northern Water Feature. (Courtesy of Hargreaves Associates.)
Figure I.13.8 Twelve-m-high jets of the Northern Water Feature. (Photo by John Gollings. Courtesy
of Hargreaves Associates.)
Chapter I.13:
255
Figure I.13.9 Created Wetland Ponds, showing curvilinear fingers of macrophytes and open water
areas. (Courtesy of Hargreaves Associates.)
Edge treatment (Figure I.13.13) of the wetland pond is primarily a carefully constructed, vegetated landscape. At the edge, stepped terraces from 3 to 7 meters wide allow
progressive establishment of wetland species around the perimeter of the pond.
A shallow, 1:50 grade for the bank adjacent to the pond edge addresses issues of public
safety and liability by creating a safe, gentle transition to deeper open water (Figure I.13.14).
The slope also ensures adequate drainage to avoid still areas of water and allow effective
mosquito management. Strategically placed gabions (at and beneath the permanent water
level) direct water flows through the system (Figure I.13.15).
Soft landscape edge treatments reflect the environmentally sustainable design principles that form the basis of the wetland design and limit the reliance on expensive imported
materials. Graded zones of aquatic and terrestrial plants (Table I.13.1) stabilize the waters
edge physically and visually record the environmental processes operating in the ponds.
The fluctuating water level and changing habitat describe the moisture gradient from
water to land as an educational narrative. This narrative of environment and water cleansing
is expanded in a public art installation. In Osmosis, by Ari Purhonen (Figure I.13.16), the
color transition of the artwork (from red, through the color spectrum, to violet), perceptible
through the steel grate decking of a 112-m-long wetland viewing pier, interprets the water
and wetland systems below.
The Northern Water Feature meets its multiple objectives, including stormwater quality treatment, aesthetic, and landscape roles, and is now a key component of the water
cycle infrastructure for Homebush Bay (Figure I.13.17). It is a signal landscape and a
construction that has become a memorable urban development project that exemplifies
the OCAs goals of environmentally sustainable development.
Acknowledgments
The permission of the Olympic Coordination Authority to outline the drainage and stormwater management strategy, the master concept design, and to discuss the Northern Water
Feature and its role in the water cycle management strategy for Homebush Bay is gratefully
acknowledged. The views expressed in this chapter are those of the author and are not
necessarily the views of the Authority.
256
Figure I.13.10 The Northern Water Feature wetland plantings and fountain jets beyond. (Photo
by John Gollings. Courtesy of Hargreaves Associates.)
The Northern Water Feature design team

The design team included Hargreaves Associates, with Gavin McMillan, local landscape
architect; Anton James, local landscape architect; Shaffer Barnsley, local landscape architect; Willing and PartnersGeoff Sainty & Associates, environmental engineers; WoodwardClyde, environmental engineers; Ove Arup, civil/structural engineer; Barry Webb
& Associates, lighting consultants; Sydney Fountains/Waterforms, fountain mechanical
consultants; Fluid Flow, irrigation design consultant; and Ari Purhonen, artist.
Chapter I.13:
257
Figure I.13.11 Hydrographic model of surface elevations. (Courtesy of Hargreaves Associates.)
Figure I.13.12 Wetland edge plantings the Northern Water Feature. (Courtesy of Hargreaves
Associates.)
Table I.13.1 Macrophyte Plantings of the Northern
Water Feature
Wetland Plant Schedule
Baumea articulata Jointed twig rush
Balboschoenus fluviatilis Marsh club rush
Eleocharis acuta Common spike rush
Eleocharis Tall spike rush
Gahnia sieberiana Red-fruited saw sedge
Lepironia articulata Lepironis
Phylidrum lanuginosum Wolly frogmouth
Schoenoplectus mucronatus Schoenoplectus
Schoenopolectus validus River clubrush
Spacing
4
4
4
2
1
2
3
5
4
per
per
per
per
per
per
per
per
per
m/2
m/2
m/2
m/2
m/2
m/2
m/2
m/2
m/2
258
Figure I.13.13 The wetland and fountain. (Courtesy of Hargreaves Associates.)
Figure I.13.14 Surface roughness in the Northern Water Feature. (Courtesy of Hargreaves Associates.)
Chapter I.13:
259
Figure I.13.15 Flow patterns in the Northern Water Feature in a 20-year ARI flood. (Courtesy of
Hargreaves Associates.)
Figure I.13.16 Osmosis public art installation by Ari Purhonen on the observation pier. (Photo
by John Gollings. Courtesy of Hargreaves Associates.)
260
Figure I.13.17 The Northern Water Feature at night. (Photo by John Gollings. Courtesy of Hargreaves
Associates.)
Literature cited
AGC Woodward-Clyde Pty, Ltd., Statement of environmental effects, the Northern Water Feature,
prepared for the Olympic Coordination Authority, 1997.
Allen, G., Public domain design workshop Process, Landscape Aust., 22, 2000a.
Allen, G., The Northern Water Feature, Landscape Aust., 22, 2000b.
Allen, G. and McMillan, G., Sydney 2000, the Northern Water Feature; Sustainable Architecture White
Papers, Earthpledge Foundation, Brown, D.E., Fox, M., Pelletier, M.R., Eds., 2000.
Hargreaves Associates (with Government Architect Design Directorate), The Homebush Bay site
Master concept design, prepared for the Olympic Coordination Authority, 1997.
Olympic Coordination Authority, Homebush Bay development guidelines Vol. I Environ. strat.,
1995.
Phillips, B.C., Allen, G., and Listowski, A., Constructed wetlands Their role in the water cycle
management strategy for Homebush Bay, Proc., Hydrastorm 98 Conf., IEAust, Adelaide, 1998.
Willing & Partners Pty, Ltd., Stormwater management strategy for Homebush Bay, final report,
prepared for Sydney Water and Olympic Coordination Authority, 1995.
The State Government of New South Wales, Sydney regional environ. Plan No. 24 Homebush Bay.
The Sydney Olympics 2000 Bid Ltd., Environmental guidelines for the Summer Olympic Games,
2000.
Chapter I.13:
261
Response
Highly visible water:
Recreating a landscape for public use
As Glenn Allen demonstrates, in this chapter, a valuable role that water sensitive design
can often provide is one of environmental connection and education, especially if undertaken at sites of high visibility. For despite great advances in public education about the
multiple roles wetlands play in watershed functioning, there is still the popular perception
that such swamps are blights where creation, instead of destruction, is simply ludicrous
(see, for example, Chapter I.11 by Clarissa Rowe). As long as wetland creation remains
trapped in the construction of blue-collar functional wetlands (cf. Robert France and
Kaki Martin in Chapter I.12), they will continue to be unrecognized and regarded with
derision. The project described in this chapter illustrates that the educational benefits of
creating white-collar interpretive wetlands at sites of high visibility (Indeed, it is hard
to imagine anywhere being more visible than an Olympic Park!) will be important for the
conscious raising that should aid in establishment of interpretive wetlands at other, less
visible, locations elsewhere, such as those described by France and Martin in Chapter I.12,
by James Bays in Chapter I.14, and by Catherine Berris in Chapter I.9.
The new water features created at this site truly connect existing natural waters and
provide a conceptual and spatial integration for the overall landscape design (such a
strategy was also adopted on a smaller scale by France and Martin in Chapter I.12). The
role of landscape architecture is strongly felt in the design of this project, for not only was
water regarded from an environmental perspective such as for water quality improvement,
habitat creation for endangered species, and conservation reuse, it also was celebrated as
a beautiful and worthy element in its own right through the use of sculptures and fountains. This represents a very important direction in recent water sensitive design: the
engineered exposure and artistic acknowledgment of hydro-infrastructure.
chapter I.14
Principles and applications of wetland

park creation
James S. Bays
Introduction
Natural and constructed wetlands have been used for decades for stormwater and wastewater treatment in all climates across North America and the world. By design, treated
effluent can be carefully applied to either natural forested wetlands or shallow vegetated
constructed marshes, where biological, chemical, and physical treatment processes in the
wetland water, soils, and vegetation provide consistent and sustainable pollutant removal.
Increasingly common are treatment wetlands designed to provide multiple benefits,
including aquifer recharge, creation and maintenance of undeveloped green space
within urban areas, wildlife habitat, and educational and passive recreational opportunities for the community.
Wetland parks can be designed to blend seamlessly into a neighborhood, public
facility, or local park network, frequently serving as a separate destination to bird watchers,
nature students, and members of the public interested in passive recreational activities
such as hiking and sightseeing. Well-used wetland parks (e.g., Chapters I.8, I.9, I.12, and
I.13) are designed to provide a variety of marsh, aquatic, transitional, and terrestrial
habitats, with ample opportunity for bird photography, nature study, and environmental
education. Valued environmental functions of wetlands then become accessible and visible
to a general public that ordinarily might not take notice. Project owners, sponsors, and
operators then benefit from an appreciative public.
Successful designs for wetland park facilities result when a project design team exhibits the following attributes:
Experienced technical staff, including landscape architects, wetland scientists, civil
and water resource engineers, and public involvement specialists
Communicative and solution-oriented project leaders motivated to understand the
perspectives of the different technical disciplines employed
Enthusiastic support of project owners who become directly involved or issue clear
direction to enable the project team
Meaningful public interest and involvement predicated on objective presentations
of project siting and design issues that provide opportunities for public responses
to be incorporated before and during the final design phase
1-56670-562-2/02/$0.00+$1.50
263
264

Spreader
Swales
Buried Forcemain
from Urban Area
Interpretive
Center
Overlook
Open
Water
Picnic
Areas
Marsh
Boardwalk
Islands
Berm
Cascade
Aeration
Collector
Swale
Rip-Rap
Discharge to
Receiving Stream
Figure I.14.1 A conceptual wetland park. In addition to water distribution and collection features
and vegetated habitats, wetland parks include public parking and access, interpretive facilities, a
boardwalk and trail system, and wildlife observation facilities.
Operation and maintenance training and support that educate the user on ecosystem processes, project design objectives, and troubleshooting as well as describe
the physical attributes of the system
This chapter presents an overview of the feasibility assessment and planning, design
and construction, and operation, maintenance, and monitoring issues required to design
a wetland park using wastewater or stormwater as a water source. Successful examples
of wetland park designs are provided.
Planning and feasibility assessment

Wetland parks provide a unique public amenity. Emphasis is placed on developing passive
recreational uses, such as hiking, bird watching, and photography, or environmental
education, through signage, guided tours, or other direct involvement. Park features that
meet these needs include adequate parking, trails, an interpretive center, and an accessible
boardwalk into the wetland (Figure I.14.1); however, wetland parks are rarely created
solely for this purpose. Instead, a wetland park project is generally propelled by the need
to achieve one or more of a wide range of environmental planning objectives: flood control,
water quality improvement, water storage or aquifer recharge, fisheries enhancement,
wildlife habitat creation or restoration, and maintenance of open space. Beneficial public
Chapter I.14:
Principles and applications of wetland park creation
265
use of such projects has historically been viewed as important but ancillary to the original
design intent (Knight, 1992, 1997). As a result, constructed wetlands designed as parks
must balance site aesthetics and public-use features while meeting fundamental hydrologic, ecological, and engineering requirements.
Careful planning, design, installation, and maintenance of a wetland project determine
whether or not the wetland will perform its intended functions. Project success reflects
the level of consideration given to maintenance of the post-construction water regime and
desired vegetative community during the planning, design, and construction. Useful
general references for wetland planning and design include Kusler and Kentula (1990),
Marble (1991), Hammer (1992), WPCF (1990), USDA (1992), Mitsch and Gosselink (2000),
Kadlec and Knight (1996), NCHRP (1996), and U.S. EPA (2000).
The creation of wetland parks varies from project to project. The public relations
benefit of the restoration or creation of a diverse, productive wetland habitat should be
readily apparent, but experience indicates that an informed, enthusiastic, and balanced
perspective on the part of the wetland owner can help overcome stakeholders potential
concerns regarding liability, maintenance, costs, and performance. Initially, a feasibility
analysis that objectively examines the potential technical, regulatory, and economic benefits and constraints of the project should be performed. A positive outcome of this critical
step can lead the way to regulatory acceptance, final design, and construction.
The feasibility study should begin with the clear definition of the desired function(s)
of a planned wetland. These functional goals can be described further as a series of
objectives that include measurable criteria in the form of a performance standard. Performance standards should specify the desired future conditions for hydrology, soils, vegetation species composition and cover, and habitat features of the wetland.
Wetland technology is still in a developing phase; therefore, predicting wetland functional performance with precision requires the use of published hydrologic, water quality,
and ecological design tools and a familiarity with successful wetland design experiences.
Existing and known potential constraints to successful wetland construction and operation
should be taken into consideration when assessing whether project objectives will be met.
Conceptual wetland design criteria are then developed that vary in detail depending
on the goal of wetland construction. All require careful consideration of wetland type,
configuration, size, water source, soils, and vegetation. Table I.14.1 summarizes design
criteria typically considered when planning constructed wetland projects.
Preliminary locations for wetland projects are then developed based on an analysis
of identified alternative locations and the extent that they satisfy stated siting requirements, or criteria. Stated goals and objectives of the project should be balanced with sitespecific constraints. Criteria for locating a wetland will vary depending on whether a
wetland is being constructed to replace or restore lost ecological functions or enhance
existing wetland functions, as in a mitigation wetland, or whether a wetland is being
constructed or enhanced to provide a new ecological function, as in a constructed or
natural wetland treatment system.
The importance of the following wetland site selection criteria can vary between sites,
depending on the wetland goal:
Proximity to desired location

Availability of sufficient contiguous area
Availability of suitable long-term wetland water source
Favorable site hydrogeology
Acceptable site geotechnical constraints
Presence of existing or potential limiting land use, natural wetlands, protected
species, and historical or archaeological resources on or adjacent to site
266
Table I.14.1 Typical Wetland Design Criteria

Criterion
Purpose
Examples
Type
Identifying the dominant

vegetation types desired
will direct subsequent water
regime analysis and soils
characterization.
The physical layout of the
wetland project guides the
design of water distribution
and public access features.
Emergent marsh tidal or freshwater

Forest tidal or freshwater
Vernal pool seasonal
Riparian woodland
Configuration
Size
The area of a wetland project

can be determined by
project objectives and
subsequently modified by
land cost, availability, and
proximity to project
location.
Water source
A suitable long-term wetland

water source must be
seasonally available and
appropriate in quality.
Possible sources include
stormwater, treated
municipal and industrial
effluent, and intercepted
surface and groundwater
flows.
Location and distribution of habitats associated

with different vegetation types and water
depths
Number and orientation of component basins or
cells
Important physical features, including depth,
volume, islands, side:slopes, and others
Water quality models can be used to predict
the minimum area necessary to achieve a
specified goal for a given hydraulic loading rate
and pollutant concentration
Habitat minimum area requirements can be set
by compensatory mitigation goals (i.e., a ratio of
created, restored, or enhanced wetland area to
wetland area displaced by a project; a minimum
depth and size may also be set as a function of
species requirements
Hydrologic storage minimum area
requirements may be set by the desired storage
volume for a specified storm event or flood
return frequency
The proposed site must have the appropriate
hydrology and topography for the intended
wetland type.
Excavation, diking, and grading of the site to
create suitable topography for wetlands are
possible options; however, for wetlands not
dependent upon effluent as a water source, a
connection to a permanent, natural groundwater
or surface water source must be within reach for
successful creation.
Potential ease and cost of acquisition of ownership rights, easement, or other

controlling interest
Ease of access for construction and maintenance
Availability of sufficient construction materials and labor resources
Relevant features of these criteria that may affect project feasibility are described in
Table I.14.2. Categories of data useful for assessing site suitability and project feasibility
are listed in Table I.14.3.
Determination of the water regime and water balance is a critical component of the
feasibility planning phase of the project. To the extent possible, the water balance analysis
should establish the water needed to meet the targeted hydroperiod or to determine the
area that can be supported with the available water supply. Figure I.14.2 illustrates the
conceptual requirements of a wetland water balance that can be used to structure a
Chapter I.14:
267
Table I.14.1 Typical Wetland Design Criteria (continued)

Criterion
Purpose
Examples
Soils
Soils should be suited to

support wetland vegetation
and to support the desired
hydrology of the wetland.
The hydrologic properties of
the site soils should be
understood before
concluding the project
design.
The geotechnical properties
of the soil will influence
side:slope ratios and
submersed angle of repose.
Vegetation
Wetland vegetation should be

selected
Tolerance of inundation and
oxygen-poor, reduced
environments
Native origin
Rate of growth and biomass
accumulation
Diversity
Infiltration rates may greatly influence the

seasonal water budget, and consequently the
need for an engineered soil or synthetic liner,
which adds significantly to cost; this can be
determined by site permeability testing
Soil type the ability of roots to penetrate the
wetland soils will influence the ability of the
wetland to support the intended species
composition
Source salvaged wetland or upland topsoils
may be used to facilitate the establishment of
wetland vegetation; the results of seed bank
surveys and the potential for invasion by
undesirable species should be given full
consideration
In constructed treatment wetlands, topsoil use
should be considered as an option, but it is not
necessary as long as the exposed soils are
capable of supporting the planted vegetation.
Berms should be constructed from stable
materials and protected by erosion control
materials and methods.
Vegetation should have a tolerance of prolonged
inundation, low oxygen concentrations in the
water and soils, and rapid dense growth to
shade surface waters and reduce algal
production.
Planting stock should originate from the project
region.
Planting centers can range widely depending on
the desired rate of growth and the potential for
invasion by undesirable species; this may mean
a center spacing of 3 to 10 ft for most wetlands,
but as small as 2- to 3-ft centers for wetlands
where rapid growth is desired.
Vegetative diversity in mitigation wetlands
should be encouraged by planting multiple
species and using topsoil as mulch where
feasible.
Note: Wetland design must take into consideration type, configuration, area, water source, soils, and vegetation.
hydrologic analysis. The level of detail needed for performing a water balance analysis
varies widely between projects and depends on the level of detailed information available
and the objectives of analysis.
The area of the wetland can be determined as a function of wetland area required to
meet ecological goals established through a compensatory mitigation analysis, or in the
case of water quality improvement, through the use of empirical treatment wetland sizing
models (e.g., Kadlec and Knight, 1996). Wetland area will vary according to the flow to
be treated, the quality of the water to be treated, and the functional treatment goals. As
described earlier, the availability of sufficient land to meet wetland project objectives is a
critically important feasibility factor.
268
Table I.14.2 Wetland Site Selection Criteria

Criterion
Relevance
Proximity
Compensatory mitigation If designed to mitigate for total or partial loss

of function, the wetland may need to be constructed in the vicinity of
the original wetland (e.g., on-site vs. off-site). Sites for mitigation
projects are frequently in the same watershed as the impacted wetland
and a similar size or larger.
Stormwater treatment Wetlands designed for stormwater treatment
may need to be located at an appropriate topographic elevation in order
to maximize gravity flow.
Municipal/industrial treatment Natural and constructed wetland
treatment systems may need to be designed on or adjacent to the
location of the pollution source in order to minimize land and pumping
costs, and to control or limit public access.
Successful restorations are positively correlated with project size.
Sufficient contiguous area should be available to allow the wetland to
be constructed at one location to minimize construction, operation, and
maintenance costs and minimize nonnative species invasion.
Sufficient availability of a suitable long-term wetland water source is
critically important to the construction and viability of wetlands. The
proposed site must have the appropriate hydrology and topography
for the intended wetland type. Excavation, diking, and grading of the
site to create suitable topography for wetlands are possible options;
however, a connection to a permanent, natural groundwater or surfacewater source must be within reach for successful creation. The
constructed wetlands viability will be determined by the seasonal and
annual availability of wastewater effluent.
Site hydrogeology should be favorable for wetland construction.
Excessively drained soils may not be suitable for wetland construction
without the installation of an aquitard of clay or other materials of lowhydraulic conductivity. Shallow depths to the surface of bedrock may
also constrain wetland excavation. Suitable soils must also be present
or available for transfer to the site. Hydric soils can differ considerably
from mesic and xeric soil in organic matter content, fertility, pH,
structure, and texture. Saturation of upland soils over time will not
guarantee the development of a suitable wetland soil. Use of wetland
soils from donor in-kind wetlands is an effective technique for
establishing native vegetation in mitigation wetlands, but consideration
must be given to the potential introduction of undesirable or poorly
controllable plant propagules.
Wetland berm and substrate materials should be suitable for wetland
construction and not lead to excessive erosion, sediment loss, or
potential for failure under normal design extremes.
Human land use may constrain the suitability of a wetland construction
location. Care should be taken to locate the wetland in areas with
compatible zoning and other land uses in full recognition of the wetland
design goals. The presence of natural wetlands, protected species
habitats, and historical or archaeological resources on or adjacent to site
may pose additional significant design constraints.
Sites not currently under the ownership of the project owner will need
to be assessed for ease of acquisition of ownership rights, easement, or
other controlling interest. Wetlands are land-intensive, so land costs will
significantly affect the total project cost.
Area
Wetland water source
Soils and hydrogeology
Geotechnical attributes
Limiting land uses and

other siting constraints
Land ownership
Chapter I.14:
269
Table I.14.2 Wetland Site Selection Criteria (continued)

Criterion
Relevance
Access
Each site should be evaluated for existing and potential ease of access
for construction and future maintenance. Local land-use regulations
should be consulted to identify possible constraints to construction and
maintenance traffic.
Availability of sufficient construction materials and labor resources
should be evaluated within a regional context in order to minimize
project cost and to maintain standards of quality for materials. The
availability of skilled contractors, plant nurseries, and acceptable
wetland construction materials should be assessed.
Sufficient data should be collected from each proposed construction
site(s) to respond to the information needs of site selection criteria and
to evaluate the potential for successful wetland permitting,
construction, and operation.
Materials
Data collection
Note: Wetland goals need to conform to site constraints for greatest potential for project success.
Table I.14.3 Potential Data Requirements for Wetland Feasibility Analysis

Category
Attributes
Climate
Temperature annual and monthly averages

Precipitation annual and monthly totals
Evaporation annual and monthly totals
Classification
Water table depth and variation
Hydraulic conductivity
Cation exchange capacity
Bedrock depth and type
Aquifer depth, type, and confinement
Surface elevation contours
Drainage basins
Surface waters
Seasonal flow and inundation patterns
Cover types
Habitat types
Occurrence of nonnative species
Local and regional land uses
Proximity to recreational facilities
Property ownership
Wetland jurisdiction and extent
Protected species occurrence
Historical and archaeological sites
Specific conductance and pH
Temperature, salinity, and dissolved oxygen
Nutrients
Metals
Organic compounds
Soils and geology
Topography and hydrography
Vegetation and wildlife
Cultural
Regulatory
Soil and water quality
Note: Data needs for wetland planning will vary by site. These categories yield useful information for site
selection, wetland design, and construction.
270

General Water Balance Equations
Qc, Qsm
Qc, Qsm
ET
Qb
Qo
Qi
Qgw
Dynamic water balance
Periodic water balance
dV/dt = Qi Qo + Qc Qb Qgw + Qsm + PA ETA
Vstored = Vi Vo + Vc Vb Vgw + Vsm + Vr Ve
where
where
V
T
Qi
Qo
Qc
Qb
Qgw
Qsm
P
ET
A
=
=
=
=
=
=
=
=
=
=
=
Vstored =
Vi
=
Vo
=
Vc
=
Vb
=
Vgw
=
Vsm
=
Vr
=
Ve
=
water storage in wetland, m3

time, d
input wastewater flow rate, m3/day
output wastewater flow rate, m3/day
catchment runoff rate, m3/day
bank loss rate, m3/day
infiltration to groundwater, m3/day
snowmelt rate, m3/day
precipitation rate, m/day
evapotranspiration rate, m/day
wetland surface area, m2
change in water storage in wetland, m3

input wastewater volume, m3
output wastewater volume, m3
catchment runoff volume, m3
bank loss volume, m3
volume of infiltration to groundwater, m3
snowmelt volume, m3
rain volume, m3
evaporated volume, m3
A sample spreadsheet table is presented below:

Assumptions for example
All units in m3/day.
Example wetland is in Florida and has a synthetic liner (therefore, Qgw, Qsm, and Qb = 0).
January rainfall total = 10 cm.
January measured evapotranspiration total = 5.5 cm.
Wetted wetland surface area = 10,000 m2
Total catchment area within berms = 14,000 m2.
Runoff coefficient, C = 0.95.
1.
2.
3.
4.
5.
6.
7.
Time
period
Jan
Dec
Total
Qi
Qo
Qc
Qb
Qgw
Qsm
ET
dV/dt
100
75
1.2
3.2
1.8
27.7
27.7
Notes:
1. For existing systems, Qi and Qo are usually measured parameters. For systems in design phase, Qo can be
estimated by entering values for the other parameters and assuming steady-state (dV/dt = 0) conditions.
2. Input precipitation and ET data into appropriate cells. Data can be found on Web sites having statewide or
local climate data. Some examples of appropriate Web sites for climate data include http://www.srcc.lsu.edu/
for the southern United States and http://www.wrcc.sage.dri.edu/ for the western United States. ET calculated as 80% of Pan evap. or input actual data if available.
3. ET = ET monthly total wetted area/# days per month. P = P monthly total wetted area/# days per month.
4. Qc = runoff coefficient (catchment area wetted area) P.
Figure I.14.2 Conceptual wetland water balance approach. The detail in a wetland water balance is
frequently limited by the availability of appropriate site-specific data, but monthly time steps are
the minimum needed in order to determine if there will be adequate water to maintain seasonal
stage elevations.
Chapter I.14:
271
Success of the planting phase of the project will depend on the careful matching of
plant species to the soils of the selected site and the proposed topography and hydrology.
Species differ in their tolerance to soil waterlogging and flooding. Duration, seasonality,
and depth of flooding are also important factors in a species ability to survive wetland
condition. Variations in microsite topography provide substrate diversity for establishing
seedlings; microsite topography is closely correlated with seedling distribution and regeneration success. Soil fertility can offset some of the more stressful aspects of hydric conditions, extending the range for some marginally tolerant species. Therefore, selection of
species adapted to site conditions is critical to the overall performance of the wetland.
Critical components of any wetland feasibility study with the intent of creating a
public-use park are to inform and involve the principal stakeholders in the project and to
conduct public charettes or informational workshops describing the projects benefits and
constraints. Depending on local interest and funding, a productive approach is to charter
a public-use planning process that will bring potential users and neighbors to the process
to determine optimal and desirable uses for the project, and to provide an opportunity to
correctly inform the public on all aspects of planning, construction, and operation. The
need to give full consideration to a detailed public information program cannot be overemphasized; experience indicates that the general public is not familiar with the many
benefits associated with a well-designed constructed wetlands park but will become
enthusiastic when provided examples of similar successful projects.
Wetland design and construction

Successful wetland construction, enhancement, or restoration requires a carefully prepared
set of plans and specifications, a construction management approach, an experienced
project team, and a schedule and plan for monitoring construction progress and implementing corrective actions. The design of wetland construction projects typically requires
a wide range of technical disciplines, including civil, water resource, geotechnical, and
environmental engineers, computer-aided design and drafting technicians, wetland biologists, and landscape architects. Final designs should include a detailed water balance,
the basis for wetland sizing, a description of the hydraulic conveyance and profile, a
grading plan, a detailed planting scheme, and public-use design features.
Wetland construction requires skilled field labor, such as earthmoving, planting, and
structural construction, as well as a general contractor whose main role is to coordinate
and oversee construction, and a planting contractor, whose role is to cultivate and/or
arrange for delivery of all planting materials, conduct and coordinate planting activities,
and arrange or conduct post-construction maintenance and monitoring. The project design
engineer or owners representative should arrange to review the construction progress
and be available to resolve situations during construction different than those considered
during design.
The contractor should be willing to meet with the project owner or representative on
a routine basis during the course of the project to discuss status and review project
problems and solutions. The contractor should provide project status reports on schedule
during the course of the project.
Contractor selection is critically important to every wetland project. Contractors
should be able to demonstrate prior successful wetland construction experience. Contractor staff should include a person with background in wetland creation/restoration design
with practical wetland construction experience and familiarity with local hydrologic and
edaphic conditions. The contractor or contractors insurer should be able to secure a
performance bond equal to the cost of construction, planting, and a period of maintenance
and monitoring.
272
Nuisance and exotic plants should be controlled during wetland grading and planting.
Trash and litter should be prevented from accumulating in the wetland. Wetland vegetation should be irrigated or kept watered as needed during the first years initial dry season
if not inundated to design depths. Water control structures and culverts should be kept
free of debris and soil, and repaired if broken.
Wetland construction plans and specifications should be sufficiently detailed for bidding purposes, engineering and biological review, and verification of as-built conditions.
As a minimum, wetland construction plans should include a table of contents, a detailed
location map, a sheet key index, and a table of quantities. Individual sheets should include
a compass arrow, scale bar, date of preparation, and a record of reviewers and revision
dates. Table I.14.4 summarizes the types of information useful to include within a wetland
construction document
Wetland operation, maintenance, and monitoring

As-built drawings should be prepared and certified by the earthwork contractor or
general contractor prior to installation of planting materials and submitted for approval
and acceptance by the project engineer. Final as-built drawings, which should be prepared at the conclusion of construction, must verify design elevations, water depths, and
elevations and extent of planting zones. These should be submitted with an initial monitoring report (sometimes referred to as a time zero report) at the completion of the
project, which would include descriptions of the major wetland plant communities, densities, species, and photographs taken at a sufficient number of stations to adequately
cover the project. Original Mylar or other media should be annotated and prints certified
by a licensed surveyor. Variations from design, and their rationale, should be noted on
the plans.
Post-construction monitoring can be performed to measure and evaluate whether a
wetland has attained its intended goals. Detailed descriptions of necessary monitoring
activities can be included in the construction, design, or permitting documents. Sampling
methods, frequency, and monitoring station locations should be described in sufficient
detail to permit monitoring to be conducted by qualified individuals unfamiliar with the
project. Monitoring plans should include descriptions of methods and goals of collecting
data on water levels and plant species cover and diversity. Photographs of the wetlands
should be taken at fixed locations as part of the post-construction monitoring process.
Additional data that may be collected will depend on the goal of wetland construction.
Periodic biological surveys of vertebrate and invertebrate communities may be performed
to document wildlife habitat and ecological productivity in the wetland. Water quality
sampling may be performed to document pollutant assimilation, organic matter production and export, and sediment retention. Flood retention and groundwater recharge functions may be documented by installation of monitoring wells, and water stage and rainfall
recorders. Specialized input from biologists, hydrologists, hydrogeologists, and engineers
should be sought before designing and implementing any monitoring.
Wetland performance after construction can be determined by comparison of measured wetland conditions at selected time intervals against specific criteria. Criteria to be
measured should reflect project goals. For example, specific criteria for a forested mitigation wetland might include mean tree height, basal area, density, and frequency; percent
cover by planted and volunteer plant species; minimum, maximum, and average depths
or hydroperiod; and occurrence of selected aquatic wildlife. Specific criteria for a treatment
wetland might include target effluent concentrations and expected pollutant removal
efficiency, as well as other indications of wetland condition, such as percent cover by
planted and volunteer plant species.
Chapter I.14:
273
Table I.14.4 Typical Components of Wetland Construction Documents

Information Type
Description
Aerial
photograph
If available, construction plans should include current aerial photographs at

a scale sufficient to completely show the outline of the project work area on
one or more sheets. Locations of key landmarks, water bodies and drainage
pattern, wetlands and other restricted or protected areas (i.e., endangered or
threatened species) should be indicated. Larger-scale aerial photographs may
be used as a background for the detailed plan set if interpretive clarity is not
sacrificed.
Scale
A scale of 1 in. = 90 ft or larger (e.g., 1 in. = 50 ft) is recommended.
Topography
Wetland construction plans should be overlaid on a topographic map of

existing site elevation contours. A 1-ft contour interval is recommended as a
minimum contour interval. A smaller interval, such as 0.5 ft or less, may be
more useful in developing site plans for tidal wetlands with a broad but
shallow slope. Benchmark locations and elevations should be clearly
indicated.
Cross-sections
Typical cross-sections of all earthwork should be prepared to scale clearly

indicating all design elevations, slopes, and dimensions. The number of crosssections should be sufficient to identify typical and atypical conditions.
Geotechnical data
Locations of test borings and soil pits should be identified within the plan set
so that they may be relocated, if desired. Soil profile illustrations should be
identified and presented within the plan set and should include information
on soil chroma profile elevations, observed water elevations, depth to
underlying rock, and presence of potentially significant aquitards, such as
naturally formed layers of clay or less permeable materials.
Jurisdictional
wetlands
delineation
Jurisdictional wetland boundaries should be clearly and accurately identified

on the site topographic map as negotiated with the pertinent regulatory
agencies.
Hydrologic data
Plans should indicate existing and expected water levels, identify adjacent
water bodies, and establish major surface drainage patterns at the
construction site. All elevations should be made relative to National Geodetic
Vertical Datum (NGVD) or other datum predetermined by convention, or an
elevation conversion should be supplied. Sufficient information should be
provided to determine seasonal elevations of receiving waters.
Nontidal wetlands Site hydrological data should include seasonal high and
average water elevations determined from vegetative indicators, soil
indicators, or hydrological monitoring data for existing wetlands, if any, and
at adjacent upland sites or perennial streams, which may provide an
indication of the adjacent groundwater table elevation. A series of 0.75- to
1.0-in.-diameter monitoring wells in representative soil types within the
wetland may be installed and read over a period long enough to describe a
normal seasonal range.
Tidal wetlands Tidal wetlands should include mean high water (MHW),
mean low water (MLW), and spring high tide (EHWS), and spring low tide
(ELWS) relevant to the locale of the construction site.
Water supply during construction Provision should be made on a site-specific
basis to divert water temporarily to provide inundation to the wetland during
construction by either temporary or permanent structures, such as pumps,
irrigation pipes, siphons, diversion channels, or other method appropriate to
the site.
274
Table I.14.4 Typical Components of Wetland Construction Documents (continued)

Information Type
Description
Planting
specifications
Plans should indicate zones or areas to be planted. A planting list should be

prepared for each wetland zone that includes quantities, elevation ranges,
propagule description (e.g., bare root, container grown nursery stock, field
harvested specimens) and acceptable planting conditions. Special
considerations or requirements should be noted and described in sufficient
detail. These should include fertilizer specifications, preplanting
conditioning, geographic constraints on plant sources, performance, and
irrigation requirements. Plants should be planted at intervals sufficiently
dense to assure rapid growth of vegetative cover. Plants should be
commercially available within the project area.
Plant materials Most planned wetland projects require the introduction of
plant species for successful creation and restoration of a wetland. Only
healthy stock should be used. Local seed sources are preferable to nonlocal
sources due to acclimation to local climate and other factors. Many
commercial nurseries specialize in the production of woody and herbaceous
species suitable for wetland restoration projects. If field harvested plants are
used, the planting contractor should be able to provide proof of permission
to sample from the nursery wetland.
Plant propagules Planting stock can be found in several forms. Woody
seedlings can be either containerized or bare root. Containerized seedlings
are grown in a small tube-shaped container that produces a small, compact
root system contained in a small plug of potting medium. Bare-root seedlings
are harvested from the nursery bed mechanically, are root pruned, and should
be free of associated soil. Herbaceous stock is usually produced as
containerized seedlings. Planted stock may be priced per unit for large, highquality containerized seedlings and saplings or per bundle of 500 or 900 for
woody bare root seedlings.
Care of plant stock Care of the planting stock after delivery at the planting site
is as critical as selection of appropriate stock. Stock should be kept in cold storage
until planting and protected from sunlight and drying. This is especially
important for bare root stock. Stock may be stored under tarpaulins or in coolers.
Vegetation
maintenance
Control of exotic or nuisance plants should be required within the wetland

during and after construction. Construction documents should list nuisance
or nonnative plant species that will require control. Details on control
methods should be provided for expected nuisance species. Control of
herbivory by animals may be required and should be anticipated in the
construction and monitoring phases. Provisions should be made for irrigation
during construction with available water sources for mitigation wetlands and
effluent for constructed wetlands.
Land use
Locations of restricted areas, structures, utility lines, or other infrastructure

within or adjacent to the construction area should be indicated. Special
construction restrictions or contractor coordination requirements should be
indicated.
Erosion and
sediment control
Construction plans should indicate the location, quantities, and maintenance

of acceptable and appropriate sediment control methods and slope
stabilization techniques. Possible sediment barriers include staked haybales,
geotextile silt-screens, sod, and plant seeding. Barriers should be placed at
the construction periphery and within the wetland in such a manner as to
minimize sedimentation and erosion of wetland berms or edges.
Chapter I.14:
275
Table I.14.4 Typical Components of Wetland Construction Documents (continued)

Information Type
Description
Grading plan
A grading plan should be included with the plan set that identifies the location,
elevations, and dimensions of project earthwork. The plans should include
sufficient information on radii, turning points, and baseline offsets for the
contractor to accurately locate and build the wetland. Plans should specify
soil quality requirements, soil sources and disposal areas, and means of
transporting soil. Grading specifications should indicate the allowable
tolerance in wetland grade elevation. Constructed wetlands require strict
adherence to wetland grade specifications, while mitigation wetlands should
not be graded to a completely uniform wetland soil elevation.
Site preparation
Construction plans may include removing the top 1.5 to 2.0 ft of substrate from
the project site and stockpiling of that material to use as cover for the site to
provide a seed bank or propagule source. This measure requires careful
consideration and analysis of the existing seed bank and potential for
introduction of nuisance or nonnative species.
Water supply and

flow control
Plans should show the location of surface water inflows, whether open
channel, pumped, or piped, and all outflow weirs, hydraulic control
structures, surface water conveyance, seasonal high-water overland flow, and
pipe outflows. Profile sheets should provide invert elevations of all culverts,
structures, seasonal overflows, and weirs. Detail sheets should provide
information sufficient to construct all inflow and outflow control structures,
including typical plan and profile views, and descriptions and possible
manufacturers of all items not constructed specifically for the project.
Note: The extent of the information required will depend on the project, but this table provides helpful guidance
on any constructed wetland project.
Corrective action should be taken if monitoring indicates that performance criteria are
not being met, or if other indications are found that the wetland is not functioning as designed.
Mitigation and constructed wetlands performance can be adversely affected by inundation
less than or greater than the design requires. Flow, residence time, pollutant removal efficiency, and compliance with wetland discharge standards may be adversely affected. Wetland
vegetation may be adversely effected. Possible solutions may include changing the volume,
quality, or timing of water deliveries to the wetland, the invert elevations of water control
structures, the wetland grade elevation, and the species of vegetation to be planted. Corrective
actions for mitigation wetlands should be coordinated with permitting agencies.
Selected case histories

Four wetland projects have been selected to illustrate how wetlands designed for water
quality improvement can be designed to provide unique, high-quality user experiences
as recreational parks and environmental educational facilities (Figure I.14.3). Conceptual
details and operational histories are provided for the following projects:
Green River Natural Resource Area, Kent, WA

Sweetwater, Tucson, AZ
Victoria Wetlands, Victoria, TX
Wakodahatchee, West Palm Beach, FL
276
Green River
Kent, WA
A
Sweetwater
Tucson, AZ
Wakodahatchee
Palm Beach County, FL
Victoria, TX
Figure I.14.3 Selected wetland parks. Four wetland parks from different climatic regions within the
United States combine water quality improvement, wildlife habitat creation, water reuse, and public
use and recreation.
Table I.14.5 summarizes and compares the multiple design objectives of the four
selected wetland case histories.
Chapter I.14:
277
Table I.14.5 Selected Wetland Planning Goals and Objectives
Functional
Objective
Control flooding
Recharge local
groundwater
Green River
Natural
Enhancement
Area, Kent, WA
Reduce flooding
by storing a
significant
fraction of Mill
Creek spring
flow
Not a design
criterion
Restore native
wildlife habitat
Create emergent
and transitional
marshes and
open ponds
with native
plant species
Enhance native
fisheries
Create summer
habitat for
migrating
native salmon
Maintain open,
green space in a
rapidly
urbanizing area
south of Seattle
Polishing
marshes
constructed in
abandoned
wastewater
lagoons
Preserve open
space
Reuse existing
facilities and
materials
Create public-use
and recreational
facilities
All-access hiking
and bicycle
trails; shaded
viewing towers;
public-access
parking
Sweetwater,
Tucson, AZ
Victoria
Wetlands,
Victoria, TX
Wakodahatchee,
Palm Beach
County, FL
Not a design
criterion
Not a design
criterion
Not a design
criterion
Discharge about
one million
gal/day of
effluent to
recharge basins
Create emergent
marshes and
wooded
riparian
habitats with
native plant
species and
exclude exotic
plants
Nonnative fish
excluded from
project
Not a design
criterion
Maintain
existing pond
infiltration to
shallow
groundwater
Create emergent
marshes with
native plant
species and
exclude exotic
plants
Create green
space in
marginal lands
Replaced existing
wastewater
ponds
All-access trails,
shaded gazebo,
informational
kiosk, and
marsh
overlooks;
public-access
parking
Create emergent
marshes with
native plant
species
Native fish
community has
become
established
Create a native
wetland
ecosystem as a
buffer
Recycled lumber
used in
boardwalk and
interpretive
center;
operational
windmill used
to supply water
to interpretive
center
Accessible trails,
boardwalk,
informational
kiosk, birdviewing blind,
and marshviewing tower;
public-access
parking
Not a design
criterion
Maintain green
space in densely
populated
eastern Palm
Beach County
Constructed
from existing
percolation
ponds
All-access
boardwalk,
shaded gazebos,
informational
kiosk, and
multiple
wetland
overlooks;
public-access
parking
278
Table I.14.5 Selected Wetland Planning Goals and Objectives (continued)
Functional
Objective
Improve
discharge water
quality
Green River
Natural
Enhancement
Area, Kent, WA
Reduce pollutant
load from runoff
from developed
watershed
Sweetwater,
Tucson, AZ
Reduce
suspended
solids and
nutrients before
application to
recharge basins
Victoria
Wetlands,
Victoria, TX
Wakodahatchee,
Palm Beach
County, FL
Treat nutrients
and pollutants
in biologically
treated
industrial
wastewater to
meet discharge
quality
Treat nutrients
and pollutants
in secondary
municipal
effluent to meet
discharge
quality limits
Note: These well-utilized public-use and recreational facilities balance functional objectives with site aesthetics
and public use.
Green River Natural Resource Area, Kent, Washington

Introduction
The Green River Natural Resource Area (GRNRA) is an innovative stormwater management and habitat enhancement project in Kent, WA. The project integrates the needs of
people and nature in a rapidly urbanizing landscape. The Green River Natural Resources
Area project transformed an abandoned sewage lagoon system into a combined stormwater detention and enhanced wetland facility that provides a rich diversity of wildlife
habitat. The 304-acre site is one of the last remaining open tracts of land in the Kent Valley
and incorporates state-of-the-art techniques of wetland creation and enhancement, urban
wildlife management, and stormwater treatment. With its adjacent public park and trail
system, the site is one of the largest man-made, multi-use wildlife refuges in the United
States. Useful background information, maps, and site details may be found on the city
of Kent Web site (City of Kent, 2000) and in CH2M HILL (1999).
Background
The project area is in the northwestern portion of the Kent Valley and is bounded generally
by the Green River to the west, South 212th Street to the north, 64th Avenue to the east,
and the Puget Power pedestrian/bike trail to the south (Figure I.14.4). Kent used the five
original lagoon cells for sewage treatment from 1969 to 1973, when construction of Metros
Cross-Valley Interceptor diverted all sewage flows to the Renton treatment plant. The
lagoons then became a fairly stable pond system, maintained by a relatively impermeable
clay liner. The northern cell was overgrown with vegetation; the two central cells were
normally filled with water but became partially dry in the fall, creating mudflat habitat;
and the two large southern cells (each 16.5 acres in size) contained up to 3 ft of water.
Before construction, the site contained a number of wetland and upland habitat types.
Over 200 species of mammals and birds were observed at or near the lagoons, which
served as a nesting, feeding, and brooding area for many species that use the Green River
corridor and the nearby eastwest powerline corridor as travel routes.
Planning for the new facility began in 1979 and focused on the sites potential as a
regional stormwater detention facility that could help to alleviate increases in stormwater
flows in Mill Creek. The City of Kent Public Works Department has funded the construction and continued operation of the facility since 1996. Grants from the King County 1993
Regional Conservation Futures Acquisition Program and Metros Regional Shoreline
Improvement Fund for $500,000 each helped purchase additional land, bringing the
Chapter I.14:
INTERSTATE
GRNRA
S 212 St.
GRNRA
Lake
Youngs
W. Valley Hwy.
516
n R er
iv
International Blvd.
405
G r ee
1st Ave. S
518
509
279
SE 272nd St.
INTERSTATE
ss
5
King County
S 228 St.
SR
7
16
er
Kent
GRNRA
W. Valley Hwy.
ell
Rd
S 212 St.
G ree n Riv
Washington
167
Ru
INTERSTATE
SR
18
51
Figure I.14.4 Location of the Green River Natural Resources Area, Kent, WA. Located southeast of
Seattle in Kent, WA, the GRNRA has two public gates that allow parking and pedestrian access to
the site along Russell Road, and another may be accessed along the south bike path by bike or foot.
City of Kent Public Works, 220 Fourth Ave S., Kent, WA 98032. Operations, 5821 S 240th St.,
253-856-5600.
projects total size to 304 acres. Total construction cost for the entire 304-acre property,
including wetlands, upland restoration areas, public recreational and interpretive facilities,
and land purchase was estimated at $10.6 million.
Beneficial features
Flood control. The expected storm flows in the Mill Creek drainage basin in a 100year flood event are 560 cubic feet per second (cfs). The GRNRA provides sufficient storage
to reduce flows in the lower portion of the Mill Creek to 180 cfs a 68% reduction. The
project is therefore a critical element in preventing flooding in downtown Kent and nearby
industrial areas, where certain low-lying streets, parking lots, and warehouses used to be
inundated nearly every year during high-flow events in the spring and winter months.
Water quality improvement. The water quality of Mill Creek, which lies in a highly
urbanized area, was heavily impacted by untreated stormwater. The GRNRA was configured to protect high-quality habitat at the site while substantially improving water quality
in the lower reaches of Mill Creek. Stormwater runoff entering the site passes through an
extensive treatment system, including two presettling ponds and a 20-acre constructed
wetland, reducing sediment loads and urban pollutants from upstream (Figure I.14.5).
Water then enters the main lagoon for additional treatment prior to draining back into
Mill Creek. In addition, provisions have been made to supplement flows during the
critical low-flow periods in summer using three groundwater extraction wells. This flow
280

Open Water
Marsh
Wet Meadow/Scrub-shrub
Mud Flat
Upland Meadow
Anderson
Shrubland
Park
Early Seral Forest
North
Young Forest
Berm Shrubland
Parking
Flow
Lot
Bicycle/Foot Path
er
Lagoon
Discharge
Channel
G
re
en
Ri
v
Bypass
Channel
South
Parking Lot
Future
Interpretive
Center
Overflow
Channel
Lagoon
Open
Water
Presettling
Pond
Constructed
Wetlands
Native
Plant
Nursery
Figure I.14.5 Plan view, Green River Natural Resources Area. The GRNRA comprises over 300 acres
of natural and created upland and wetland habitats, including treatment marshes, a storage and
polishing pond, final polishing wetlands, and numerous natural habitats and islands.
augmentation has the potential to improve oxygen levels and temperature conditions in
lower Mill Creek.
Wildlife habitat restoration. The projects 304 acres provide habitat to an estimated
165 bird and 53 mammal species and serves as a nesting, feeding, and brooding area for
many species that use the Green River corridor as a travel route. The creation of a large
emergent marsh/open-water wetland, in addition to increases in scrub-shrub, forested
wetland, and improved upland habitats, is expected to increase the diversity of wildlife
habitat, particularly for waterfowl. Habitat value will be maintained through formal
preservation status, active management, and specific limits on public access. The GRNA
is already considered a hotspot by local birders.
Fisheries enhancement. The design of the GRNRA specifically addressed elements
identified as critical limiting factors in a comprehensive fisheries evaluation of the entire
watershed. For example, the stormwater detention pond was designed to provide critical
habitat where coho fingerlings can overwinter without being swept out of the watershed
in a storm event, as could happen in the main channel of Mill Creek. In addition, the
detention pond was sized to accommodate the full fisheries potential of the watershed.
Preservation of open space. The Green River Valley is a highly urbanized area, with
much of the land surface covered by asphalt or rooftops. Through funding obtained from
King County, the city was able to purchase over 200 acres of fallow farmland adjacent to
Chapter I.14:
281
the abandoned sewage lagoons and incorporate this land into the GRNRA project site,
thereby saving one of the last open areas in the region from development as an industrial
park.
Reuse of existing facilities and excavated materials. A main feature of the project, and
an impetus behind its development, was the reuse of a 65-acre sewage lagoon that was
abandoned in 1973. Since that time, the five-cell lagoon had become a fairly stable pond
system, with the two largest cells containing up to 3 ft of water year-round and two other
cells shifting seasonally from water to mud flat habitat. The final design reused the existing
embankments and followed existing land shapes to reduce project costs and minimize
construction time, thereby limiting disruption to the species already in residence. A total
of 600,000 yd3 of material were removed to create the current wetland configuration and
then reused to create on-site uplands.
Public education and recreation. With its nature walks, three wildlife viewing towers,
and a bike path along the west and south sides of the site, the facility provides extensive
recreational as well as educational opportunities (Figure I.14.6). The facility also provides
ongoing opportunities for hands-on involvement in the management and maintenance of
the facility for example, in constructing habitat, maintaining vegetation, and monitoring
water quality and wildlife populations. Volunteers have built and installed bird boxes,
planted native plant species, and conducted bird counts. A 5-acre native plant nursery,
constructed on the southwestern corner of the site, will propagate the thousands of native
plants for placement within the GRNRA and other natural areas within the city and will
provide a home for plant salvaging projects. The nursery will also serve as an education
and training facility for youth and volunteers in Kent. In addition, the sites master plan
allows for construction of an Environmental Interpretive Center, which will be a focal
point for citizens involvement and education on environmental issues throughout the
Green-Duwamish watershed.
Sweetwater Wetlands, Tucson, Arizona

Introduction
The Sweetwater Wetlands and Recharge Project (Figure I.14.7) is a 24.3-ha (60-acre) operational facility built by Tucson Water to combine functional elements such as effluent
treatment, recharge, and research with a natural park setting that offers educational and
wildlife viewing opportunities to the community (Karpiscak et al., 1993). The wetlands
were developed with significant assistance and input from citizens, students from elementary through high school grade levels, and numerous environmental and community
organizations.
Background
The Roger Road Wastewater Treatment Facility produces around 42,000 acre-ft of secondary-treated effluent annually. Most of this is discharged to the Santa Cruz River. In May
1994, the Arizona Department of Environmental Quality (ADEQ) filed a suit against the
city of Tucson, alleging violations of state monitoring and reporting requirements. The
city and ADEQ negotiated a settlement that, among other things, committed the city to
designing and building an experimental wetland/recharge facility to treat backwash filter
water from the citys Reclaimed Water Treatment Plant. The wetland would be constructed
within an area including an existing sludge pond. The facility would be designed to
include wildlife habitat and educational as well as recreational amenities.
282
Figure I.14.6 Public-use facilities, Green River Natural Resource Enhancement Area. Three wildlife
viewing towers are connected by a graveled trail accessible by foot, bicycle, and wheelchair.
The citys reclaimed water treatment plants filters are periodically cleaned by backwashing. The backwash water then is recycled through the countys treatment plant for
reprocessing, at an annual cost of about $100,000. Instead of being reprocessed by the
plant, the backwash water now is treated in the Sweetwater Wetlands.
The backwash water is first conveyed to two relatively small (1.2-acre) but densely
vegetated treatment wetlands for settling of solids before entering the wetland ponds.
Additional solids removal takes place through microbiological transformations in two
wetland ponds totaling 17 acres downstream of the settling wetlands (Figure I.14.8). The
backwash water is treated to meet or exceed secondary standards. The wetland effluent
then flows to two recharge basins totaling 6 acres. About 300 acre-ft of backwash water
will be treated annually for recharge.
Chapter I.14:
Graham Cochise
Pima
Pinal
Pima
283
77
INTERSTATE
Sweetwater
Wetlands
10
Arizona
Saquaro
National
Park-East
86
INTERSTATE
INTERSTATE
19
Pima
County
10
Tucson
83
286
Figure I.14.7 Location of the Sweetwater Wetlands, Tucson, AZ. Located northwest of Tucson on
Rogers Road, the Sweetwater Wetlands are accessible by car and bicycle year-round.
Figure I.14.8 Plan view, Sweetwater Wetlands. The Sweetwater Wetlands include 30 acres of treatment marsh, polishing, and habitat wetlands and 50 acres of recharge basins.
Estimated total construction cost was about $1.7 million, with about $600,000 earmarked for public-use amenities. The facility design was finalized in late 1995. Construction was completed in early autumn 1997.
284
Beneficial features
Water quality improvement. Backwash water from the citys Reclaimed Water Treatment Plant, as well as secondary effluent from Pima Countys Wastewater Treatment
Facility, are introduced into four densely vegetated settling ponds covering 0.7 ha (1.8
acre). The water flows from the ponds into two 3.0 ha (7.5 acre) free-water-surface wetland
cells where natural biological processes further treat the water. The water then is released
into four recharge basins (5.7 ha [14 acre]) where it filters through the ground. The
recharged water is later recovered and reused to irrigate parks, golf courses, schoolyards,
and street medians in the Tucson area. Figure I.14.8 provides an overview of the wetland
configuration.
The wetland system was designed to reduce suspended solids to less than 8 nephelometric treatment units (NTU) and BOD to less than 7 mg/L. A second alternative consisted
of a treatment wetland system to treat backwash water from the filtration plant. The
average suspended solid load is 275 mg/L, and the target load is less than 20 mg/L.
Available data indicate that Sweetwater is meeting these goals.
Wildlife habitat restoration. Since April 1997, the emergent plants have grown rapidly
and the density has increased greatly. Additionally, new shoots have grown back after
each winter. A wildlife study funded by the Arizona Game and Fish Department (AGFD)
and further observations by the Tucson Audubon Society have documented over 120
species of birds at the Sweetwater Wetlands since the spring of 1997. Prior to construction
of the Wetlands, background surveys noted less than 20 bird species as being present at
the 24.3-ha (60-acre) site. The AGFD study has documented not only seasonal species
changes but also a shift in populations with time as the Wetlands develops. Studies also
have documented the nesting of species such as black-necked stilts and black-bellied
whistling ducks. Mammals observed at the site have ranged from field mice to bobcats.
Public education and recreation. This project included an extensive public information and participation task to ensure public input. The citys consultant, along with a
citizens advisory committee appointed by the city council, developed the types of educational and recreational amenities appropriate to this unusual facility. In addition, three
public open house meetings were held to solicit input from the general public. As a result
of this process, the wetland became an urban park.
The Sweetwater Wetlands includes extensive passive public-use elements to provide
the visitor with a variety of experiences relating to water management and ecological
issues (Figure I.14.9). A public parking area is designed to accommodate cars as well as
school buses. All the public-use facilities are wheelchair accessible. A six-sided kiosk is
the initial focal point for visitors. A ramada provides shade, seating, and a place for groups
to meet. Restrooms are available near the kiosk, served by their own subsurface flow
treatment wetland. Extensive landscaping of the area includes the marsh of the wetland,
a transitional marsh around the edge of the wetland cells, and an upland zone outside
and above the transitional community. A looping trail system provides numerous viewpoints and interpretive signage to inform the visitor. Interpretive signage is provided at
the kiosk. Native arid landscape plantings are incorporated throughout the facility. Since
the public opening in early 1998, Sweetwater has attracted more than 40,000 human
visitors, as well as nearly 200 species of birds, mammals, reptiles, and insects.
Mosquito control research has been extensive and detailed at this facility; bat houses
are included as a natural control for adult mosquitoes, and native fishes and invertebrates
are encouraged. A commonly used predator of mosquito larvae, the mosquitofish (Gambusia affinis), has not been implemented at this wetland project, unlike other areas, because
Chapter I.14:
285
Figure I.14.9 Public-use facilities, Sweetwater Wetlands. Shaded gazebo overlooks, paved trails and
bridges, and informational signage provide access to a variety of habitats for visitors to the Sweetwater Wetlands.
286
itt
De
or
ct
Vi
ia
59
183
on
cks
Ja
ia
tor
Vic
172
234
Victoria
59
616
87
77
Texas
Victoria
County
Victoria
Figure I.14.10 Location of the Victoria Wetlands, Victoria TX. The Victoria Wetlands are located at
2695 Bloomington Rd., N., Victoria, TX 77905, Contact Amy Hodges at (361) 572-2137 (amy.e.hodges@
usa.dupont.com) or John Snyder, DuPont Wetland Educator, (361) 572-1153 ([email protected]),
http://www.lwv.org/where/protecting/webwalk/drill_down7.html.
of the strong desire of the local environmental community to limit the distribution of
species non-endemic to the region. Natural aquatic predators of the mosquitoes, including
arthropods and dragonfly and damselfly nymphs, although numerous, have not been able
to keep up with the mosquito larvae population, and active measures for its control have
been implemented. These have included innovative aerial distribution of safe mosquito
larvicides using a remote-controlled helicopter.
DuPont Victoria Wetlands, Victoria, Texas

Introduction
At DuPonts nylon intermediaries manufacturing facility in Victoria, TX, a 53-acre treatment wetland park and wildlife habitat area has been designed and constructed (DuPont,
2001). Designed to provide final polishing of biologically treated process wastewater, the
wetland also provides a high-quality wildlife habitat and a scientific educational facility
for public schools (Figure I.14.10).
The wetland was designed with three stages that focus on water quality improvement
at the front end (Stages 1 and 2) and habitat and public use in the final stage. Features
include five parallel cells and transverse deep zones for efficient solids removal in Stage 1;
two parallel cells with more sinuous embankments and deep zones for effective hydraulic
efficiency and habitat/aesthetic values in Stage 2; and a single, sinuous cell with large
open-water areas, mixed marsh, and shallow littoral shelf areas for habitat and public use
in Stage 3 (Figure I.14.11). Total project cost is estimated to be $3 million.
During the project design, DuPont staff and their consultants worked to address
questions from the Education Advisory Group, the Texas Natural Resources Conservation
Chapter I.14:
Figure I.14.11 Plan view, Victoria Wetlands. The Victoria Wetlands comprise over 53 acres of created treatment marshes, polishing
wetlands, habitat marshes, preserved upland mesquite bosque, and numerous natural habitats and islands.
287
288
Figure I.14.12 Public-use facilities, Victoria Wetlands. A large, open-air interpretive facility, bird
blinds, shaded overlooks, and numerous foot trails support an intensive wetland training program.
Council, and citizens. The third-stage wetland area includes an outdoor/open laboratory
building for science education, a blind for bird watching, a wheelchair-accessible path
and overlook for the general public, and a boardwalk through marsh vegetation
(Figure I.14.12).
Beneficial features
Water quality improvement. Available monitoring data indicate measurable treatment by the wetlands of nitrate in the polished effluent from the manufacturing facility.
Wetland discharge permit criteria, established through the National Pollutant Discharge
Elimination System (NPDES) permit, have not been exceeded since the wetland commenced operation (Cole et al., 2001).
Chapter I.14:
289
Figure I.14.12 (continued)
Wildlife habitat restoration. The wetland wildlife community includes a total of 195
bird species have been documented at the wetland, along with a full complement of frogs,
reptiles, amphibians, and mammals. The average bird population density of 10/acre
(25/hectare) is reported to be consistent with that found in other treatment wetlands (Cole
et al., 2001). The remote location, dense but diverse habitat, and secure facility provide an
important sanctuary to wildlife year-round.
Twenty species of common wetland plants were initially installed in the wetlands,
including cattail (Typha sp.), giant bulrush (Scirpus californicus), hardstem bulrush (S.
acutus), tall Olneys bulrush (S. americanus), short Olneys bulrush (S. pungens), pickerelweed (Pontederia cordata), and softstem bulrush (S. validus) (Cole et al., 2001). Natural
290
introduction of plant species by wildlife and wind and water movement has increased
the total species present to 48 (Cole et al., 2001).
The nutria (Myocastor coypus), a nonnative aquatic rodent that is increasing in abundance throughout the southeastern United States, has become an important management
concern. This aquatic herbivore has consumed considerable portions of the standing
vegetation crop within this wetland project; current management options include reducing
water levels to reduce the suitable habitat available, trapping the animal, and humane
methods of removal. This example of animal herbivory points to the need to develop an
adaptive operation and maintenance plans that allows unconventional management issues
to be addressed.
Fisheries enhancement. Fish populations representative of normal riverine fish communities are present in the wetlands and have remained diverse and stable since operation
began in 1997. Mosquitofish and bluegill (Lepomis macrochirus) are dominant within this
community.
Preservation of open space. A lesser benefit in this rural setting than the other wetlands described in this paper, the open, prairie-like aspect of the wetlands creates an
attractive setting and is consistent with the Victoria Plants general interest in making
beneficial uses of its property.
Reuse of existing facilities. The reuse of an old windmill as a pump for water to the
interpretive center, as well as the careful site planning necessary to conserve a mesquite
bosque and existing stands as habitat islands, reflect the care taken to ensure that the
ecological effect of constructing the wetland remained positive.
Public education and recreation. DuPont has entered into a cooperative arrangement
with the local school system and the regional education center to hire a full-time teacher
who implements a curriculum program for use by elementary, middle, and high school
students. Local teachers within a 150-mile radius of Victoria have integrated this educational facility into their required science curriculum at the various age levels. The DupontVictoria site is typically booked for visiting students a year in advance. The site has been
estimated to host 2,000 to 2,500 students per year. As of April 2001, more than 9,000 4thgrade through 12th-grade students have participated in the wetland educational program.
This extensive overall program reaches well into the community and involves representatives from at least 25 outside groups ranging from the Texas Parks and Wildlife Department and the U.S. Natural Resources Conservation Service to the Audubon Society and
Texas A&M University.
Wakodahatchee Wetlands, Palm Beach County, Florida

Introduction
In 1996, the Palm Beach County Water Utilities Department (PBCWUD) in Florida converted nine existing wastewater percolation ponds into constructed treatment wetlands
at a former regional wastewater utility located in densely populated, suburban Delray
Beach (Figure I.14.13). Appropriately, the wetlands were named the Wakodahatchee, or
Seminole Indian language for created waters. Wakodahatchee covers a total of 56 acres
of created wetlands populated with native plant and animal species created from existing
wastewater percolation ponds. Opened to the public in 1996, this park-like setting provides
not only a significant environmental resource but is highly valued by regional nature
lovers and wildlife photographers (Bays et al., 2000).
Chapter I.14:
291
Woolbright Road
Wakodahatchee
Wetlands
SRWRF
Wakodahatchee
Wetlands
N
Approximate
Scale in Feet
0
809
1900
95
3800
Palm Beach
County
Wakodahatchee
Wetlands
806
I-95
Military Trail
Jog Rd.
Lake
Okeechobee
Florida Turnpike
Canal L-30
Lake
Ida
Atlantic Ave.
Figure I.14.13 Location of the Wakodahatchee Wetlands, Palm Beach County, FL. The Wakodahatchee Wetlands are located at 13026 Jog Road in suburban Delray Beach, FL, on the east side of
Jog Road, between Woolbright Road and Atlantic Avenue. The wetland is open to the general public
from sunrise to sunset, seven days a week. Tours of the Wakodahatchee are conducted on the second
Tuesday of every month at 9 a.m., and on the third Wednesday of the month at 4 p.m. during winter
months. The tour lasts approximately one hour, and the tour group size is limited to 20. Space on
the tour must be reserved in advance. For reservations or questions about the wetland, please phone
(561) 641-3429.
Background
Originally conceived in 1992 as a demonstration of the beneficial reuse of reclaimed water,
the Wakodahatchee Wetlands are an important component of the 30-million-gal per day
(mgd) Southern Region Water Reclamation Facility (SRWRF). The SRWRF treats the effluent to a secondary quality after which it is disposed of by deep injection well (DIW), a
common practice in Florida. Prior to the construction of the SRWRF facility, however,
wastewaters were disposed of through percolation ponds and a deep injection well at the
System No. 3 Regional Utility since the 1970s. As part of the recent Phase II expansion of
the SRWRF, the construction permit required the conversion of the System 3 percolation
ponds to treatment wetlands as an effluent reuse demonstration.
The percolation ponds at the former System 3 Regional Utility were modified between
November 1995 and October 1996 at a cost of $2.85 million into eight emergent marsh
wetlands, interspersed with broad areas of open water and islands of wildlife habitat.
Wetland area totals 39 acres, with individual wetland cells ranging from 2.3 acres to 10.9
acres. Most of the marsh area is designed to operate at an average depth of 0.5 ft, but may
be operated normally at depths to 1.5 ft. Total volume of water within the wetland is ~20
million gal at the normal pool elevation. The average length to width ratio is ~3:1. The
total site hydraulic loading rate varies from 0.9 cm/day at flows of 0.5 mgd, to 4.6 cm/day
for flows of 2 mgd. Nominal detention time varies from a maximum of 40 days at 0.5 mgd
of flow to 10 days at 2 mgd. A total of 28 deep zones and 8 habitat islands is included.
Snags and perching posts were added for wildlife use.
Secondary effluent is pumped from the SRWRF to a splitter box, which distributes
the flow to six parallel ponds (Figure I. 14.14). Effluent from ponds B, C, D, E, and F flows
292
Inflow
Splitter Box
Outlet to Wet
Well and Pump
to Deep Well
Transitional Marsh/Wetland
Open Water
Mixed Emergent Marsh
Bulrush
Habitat Island
Forest/Upland Grasses
Water Control Structure
Inlet Distribution Tee
Flow Direction
Scale in Feet
150
300
Figure I.14.14 Plan view, Wakodahatchee Wetlands. The Wakodahatchee comprises over 50 acres of
emergent marshes, swamp forests, deep ponds, islands, and transitional and upland habitats.
to a collector channel and then to terminal collector pond I. The sixth parallel pond (AG)
discharges in series to pond H, which in turn discharges to pond I. All wetland effluent
is then pumped to an on-site DIW for disposal during the first year of operational monitoring. Pending state water quality regulatory approval, the treated wetland effluent will
be discharged to the adjacent L-30 Canal for surficial aquifer recharge.
Beneficial features
Water quality improvement. The Wakodahatchee were designed to reduce PBCWUD
reliance on DIW disposal of effluent by using treatment wetlands to improve water quality
of the effluent to where it could be discharged to the regional surface water canal system.
Available performance data for the period from November 1996 through September 1998
indicate that, at an average flow rate of 1.4 mgd, the wetland reduces total nitrogen in the
SRWRF secondary effluent from 24.5 mg/L to 8 mg/L, and total phosphorus from 2.0
mg/L to 1.0 mg/L. These data match reasonably well the expected wetland performance
estimated using the first-order, area-based treatment model (Kadlec and Knight, 1996).
Hydrologic storage and recharge. A stated design objective of the Wakodahatchee
Wetlands was to recharge local waster supplies through infiltration to the surficial aquifer.
Available hydrologic records for the wetland indicate that, of an average inflow rate of
1.4 mgd, about 0.5 mgd of effluent was lost to infiltration and evaporation over the period
from November 1996 through September 1998.
Wildlife habitat restoration. The Wakodahatchee were designed to create significant
wildlife habitat accessible to the public in a park-like setting. About 70% of the wetland
area is vegetated by native emergent, forested, and transitional wetland species designed
to emulate native south Florida wetland plant communities. Emergent marsh zones are
Chapter I.14:
293
composed of softstem and giant bulrush, duck-potato (Sagittaria lancifolia), arrowhead (Sagittaria latifolia), spikerush (Eleocharis cellulosa), fireflag (Thalia geniculata), and pickerelweed.
Herbaceous species planted at the upper edge of the marsh zone include sawgrass (Cladium
jamaicense), Fakhahatchee grass (Tripsacum dactyloides), and Gulf muhlygrass (Muhlenbergia
sp.). Forested species also planted at marsh edge include cypress (Taxodium distichum),
pond apple (Annona glabra), Carolina willow (Salix caroliniana), red maple (Acer rubrum),
and buttonbush (Cephalanthus occidentalis). As expected, duckweed (Lemna minor and
related species) has become naturally ubiquitous throughout the wetland.
Cattails were not planted and their anticipated colonization is being controlled. Native
upland plant species have been planted on the site berms and include dahoon holly, sabal
palm, saw palmetto, cocoplum, live oak, mahogany, and slash pine. Melaleuca and Brazilian pepper were removed from the site prior to construction.
Monitoring data collected by the Palm Beach County Department of Environmental
Resource Management indicate that the Wakodahatchee Wetlands have attracted an abundant variety of wildlife, including turtles, frogs, otters, alligators, and birds. More than
176 different species of birds have been spotted at the site. These species thrive in the
various wetland zones found at the Wakodahatchee Wetland. Open-water ponds areas
attract water fowl and diving birds; emergent marsh areas are habitat for rails, moorhens,
and sparrows; shallow littoral shelves are utilized by herons and egrets; islands with
shrubs and snags serve as nesting, roosting, and basking sites; and the forested wetlands
areas are rapidly growing into swamp forests of pond apple and bald cypress. Alligators
nested in 2000 and 2001; their presence is a point of great interest to site users and an
interesting proof of the self-contained, sustainable ecosystems that can develop within
these types of wetland projects.
Fisheries enhancement. The Wakodahatchee Wetlands were stocked naturally and
deliberately with mosquitofish to serve as a prey base for feeding wildlife and as a natural
predator deterrent for mosquito larvae. Mosquitofish populations are abundant, on the
order of 10 to 15 species per m2. This provides an abundant supply of food for wading
birds, such as herons and egrets, as well as aerial diving birds, such as the least tern and
the belted kingfisher, while helping to maintain mosquito populations in check.
Preservation of open space. The suburban setting and the high level of public use by
the Wakodahatchee Wetlands argue strongly for the value found in preserving the open
space that had been conserved by the former percolation ponds. Rather than yet another
dense suburban development, the neighbors of the Wakodahatchee Wetlands have a fully
functional wetland wildlife park that is expected to stay as a park for as long as the public
and the county deem it desirable.
Reuse of existing facilities and excavated materials. The creative reuse of the percolation ponds has proven to be one of the most popular public works projects implemented
by the county. Operating costs are nominal, and public perception of this project is positive.
Public education and recreation. A mile-long boardwalk was constructed with gazebos and informative signs for easy public access and educational use (Figure I.14.15).
Parking spaces are provided at the adjacent county complex. The county held a Wakodahatchee Wetlands day in 2000 and 2001, each time attracting over 1,000 people per day.
Annual visitor totals to the Wakodahatchee are on the order of 2,000 people annually.
294
Figure I.14.15 Public-use facilities, Wakodahatchee Wetlands. A 0.75-mile-long boardwalk, three

gazebos, educational signs, and 50-vehicle parking lot provide ready access to this productive
wetland ecosystem.
Conclusion
Constructed wetland parks are valuable and rewarding endeavors to those who participate
in the planning, design, construction, monitoring, and use. If properly designed, public
safety and welfare are maintained while the natural environment is enhanced. By reviewing the examples provided and by following the general guidance provided here, more
constructed wetland parks should appear feasible to landscape architects, engineers, scientists, and sponsoring agencies. If past experience holds true, public use of these facilities
will continue to grow and exceed expectations.
Chapter I.14:
295
Figure I.14.15 (continued)
Literature cited
Bays, J. et al., The Wakodahatchee Wetlands, Palm Beach County, Florida, Proc. 73rd Meet. Water
Environ. Fed., Alexandria, VA, 2000.
CH2M HILL, The Green River natural resource area: A study of harmony in diversity, Seattle, WA, 1999.
City of Kent, WA, http://www.ci.kent.wa.us/Public Works/Special Programs/grnra.htm, 2000.
Cole, J. et al., Wetland health plan, Indus. Wastewater, July/Aug., 35, 2001.
DuPont, Victoria Wetlands Web site, http://www.dupont.com, 2001.
Hammer, D., Creating Freshwater Wetlands, Lewis Publishers, Boca Raton, FL, 1992.
Kadlec, R.L. and Knight, R.H., Treatment Wetlands, Lewis Publishers, Boca Raton, FL, 1996.
Karpiscak, M.M. et al., Treating municipal effluent using constructed wetlands technology in the
Sonoran Desert, in K.D. Schmidt, Ed., Proc. Symp. Effluent Use Manage., Tucson, AZ, 1993.
Knight, R.H., 1992. Ancillary benefits and potential problems with the use of wetlands for nonpoint
source pollution control, Ecol. Eng., 1, 97, 1992.
Knight, R.H., 1997. Wildlife habitat and public use benefits of treatment wetlands, Water Sci. Tech.,
35, 35, 1997.
Kusler, J. and Kentula, M., Eds., Wetland Creation and Restoration, Island Press, Washington, DC, 1990.
Marble, A.D., A Guide to Wetland Functional Design, Lewis Publishers, Boca Raton, FL, 1991.
Mitsch, W.J. and Gosselink, J., Wetlands, 3rd ed., John Wiley & Sons, New York, 2000.
National Council on Highway Research and Planning (NCHRP), Guidelines for the development
of wetland replacement areas, NR379, Washington, DC, 1996.
Schueler, T., Design of stormwater wetland systems: Guidelines for creating diverse and effective
stormwater wetland systems in the Mid-Atlantic region, Dept. Environ. Prog., Metropolitan
Washington Coun. Gov., Washington, DC.
U.S. Dept. Agric. (USDA), Soil Conservation Service, Wetland restoration, enhancement, or creation,
Part 650, in Engineering Field Handbook, Washington, DC, 1992.
U.S. EPA, Guiding principles for constructed treatment wetlands: Providing for water quality and
wildlife habitat, EPA 843-B-00-003.
http://epa.gov/owow/wetlands/constructed/guide.html, 2000.
Water Pollution Control Fed. (WPCF), Natural Treatment Systems, Manual of Practice 16, Alexandria,
VA, 1990.
White, T., Planning, design, and construction of the Kent Green River Natural Resources Area, Kent,
WA, submitted to Urban Wetlands, 2000.
296
Response
Designing wetlands for multiple benefits
A personal anecdote described in the book Designing Wetlands: Principles and Practices for
Landscape Architects and Land-Use Planners (France, 2002) aptly describes what James Bays
advocates in this chapter:
Recently, I attended a conference at which one of the speakers (who
was in the process of updating a major guidebook about wetland
construction) dodged several questions from the audience concerning the possibility of designing for multiple purposes. Defending
himself, this individual stressed that the guidebook he was working
on for the EPA was concerned with blue-collar wetlands, those
for which, because of obvious space limitations (i.e., suburban settings), it was therefore impossible to create functions (presumably
aesthetics and wildlife benefits) other than those intended engineering solutions to either waste- or storm-water management.
As I continued, and as Bays clearly demonstrates in his chapter, The difference between
a utilitarian blue-collar wetland and a multipurpose white-collar wetland has really much
less to do about absence of space than it has to do about absence of imagination.
Bays chapter sets the stage for a major paradigm shift wherein created wetlands
become created wetland parks such as those case studies he outlines as well as those
described by Diana Balmori in Chapter I.8, Catherine Berris in Chapter I.9, Robert France
and Kaki Martin in Chapter I.12, and Glenn Allen in Chapter I.13. As Bays identifies, not
only do such wetland parks meet their primary functional objectives such as flood control,
water quality improvement, and water storage or aquifer recharge, they also provide
fisheries enhancement, wildlife habitat creation, and open space parkland. These latter
attributes, once regarded as mere ancillary benefits, are now becoming recognized as key
elements in wetland water sensitive design, as important in their own way to judgment
of overall project success. In other words, because the ecological and aesthetic acumen of
the public has increased so markedly over the last decade, the old-fashioned square-box,
single-purpose constructed wetland will no longer be acceptable.
Bays stresses the need for wetland projects to be carefully planned in terms of hydrology, soils, and vegetation, all in relation to a series of site selection criteria. Importantly,
and in support of Wendi Goldsmith in Chapter I.17, and Robert France and Philip Craul
in Chapter I.7, for example, Bays calls for the need of water sensitive designs to be brought
to fore through the combined efforts of teams of individuals rather than by single egos.
In his mind, successful wetland park creation requires the expertise of civil, water resource,
geotechnical and environmental engineers, in addition to computer and drafting professionals, wetland biologists, and landscape architects.
chapter I.15
Applications of low-impact development

techniques (Maryland)
Michael L. Clar
Abstract
Low-impact development (LID) is an innovative technology to control stormwater quantity/quality impacts at the source using microscale management practices distributed and
integrated throughout the landscape. This technology, developed by Prince Georges
County, MD, to address perceived problems in conventional approaches to stormwater
management (SWM) within the county and the state, makes multifunctional use of the
urban landscape allowing one to design a hydrologically functional site (PGC, 1997). This
approach results in an ecologically based approach to stormwater management that is
usually more aesthetically pleasing, precludes impacts to receiving waters, and is generally
less costly to construct and maintain than conventional end-of-pipe systems
This chapter provides a chronological overview of a number of projects that have been
instrumental in the development of LID technology. The focus of the chapter is on case
studies and demonstration projects. Although the emphasis of these projects is the state
of Maryland, projects and case studies from other regions of the country are also included.
Introduction
The range of projects and case studies included in this chapter help to demonstrate the
tremendous range and versatility of LID technology in addressing environmental issues
resulting from land development activities. Chapter I.5 introduces the fundamental concepts and principles associated with LID. This chapter complements that earlier chapter
by documenting a number of LID applications and case studies, which include
1. LID as a water quality control technique for infill development
2. LID as a water quality retrofit for existing urban areas
3. LID as a comprehensive strategy to replicate predevelopment hydrologic functions
for a developed site
4. LID as a winwin strategy to provide improved environmental performance and
reduced site development costs
5. LID as a volume control method to provide downstream peak discharge protection
for major storm events
1-56670-562-2/02/$0.00+$1.50
297
298
Figure I.15.1 Bioretention schematic representation (MDE, 2000). (Color version available at www.
gsd.harvard.edu/watercolors. Password: lentic-lotic.)
6. LID as an improved approach to protect water supply water reservoirs

7. LID as an approach to address total impervious area (TIA) limitations
Maryland case studies

Case study 1: First steps Prince Georges County bioretention design manual
One of the first steps in the development and application of LID techniques in Maryland
consisted of the development of the Prince Georges County Design manual (ETA, 1993).
The bioretention design manual provided one of the first integrated, landscape-based
micromanagement tools for stormwater management, as shown in Figures I.15.1 and
I.15.2, that made the development of the LID concept possible.
This project was conceived by Larry Coffman of Prince Georges County, MD, and
directed by Michael Clar of Ecosite, Inc., and began with modest objectives. These included
a feasibility study to evaluate if an upland living filter could be used to achieve the water
quality goals of SWM (i.e., control the first flow of surface runoff) and be used on small
sites to replace the very costly and ineffective best management practices (BMPs) such as
the oil-grit separator and other similar devices.
The feasibility study was conducted by an interdisciplinary team with training and
expertise in water resources engineering, hydraulics, hydrology, geohydrology, soil science, landscape architecture, and biology. It demonstrated that bioretention could be used
Chapter I.15:
Applications of low-impact development techniques (Maryland)
299
Figure I.15.2 Bioretention cell located in a cul-de-sac illustrates multifunctional and landscape integrated features of this microscale practice. (Color version available at www.gsd.harvard.edu/watercolors.
to meet the existing water quality control criteria. In addition, it also revealed that many
additional benefits could be anticipated, including improved aesthetics, significantly lower
construction and maintenance costs, as well as the potential to reduce peak discharges by
increasing the time of concentration and reducing the effective curve number, CN.
Based on the positive results of the feasibility study, a number of case studies were
developed and documented for various uses of the bioretention concept and a Bioretention Design Manual was prepared.
Case study 2: Infill development with LID Beltway Plaza Expansion

The Beltway Plaza Expansion project demonstrated the application of bioretention techniques
for an infill development project. Infill development projects are very commonplace in existing urban areas. Often, these sites have an existing storm drainage infrastructure designed
to convey the 2- or 10-year storms but typically lack either stormwater management or water
quality control. This was the case at the Beltway Plaza located in Prince Georges County,
MD. It was determined that this site would receive a waiver for peak discharge control due
to the existing storm drainage system but would require water quality control.
A schematic of the site is shown in Figure I.15.3. The entire parking lot for the expansion received water quality control through the use of bioretention cells, as shown in
Figure I.15.4. This site proved to be a significant milestone for the use of LID technology
for a number of reasons. First, it demonstrated that the bioretention concept could be used
to control a large parking lot area by breaking the area up into a number of small drainage
catchments: the microscale management concept. Second, it provided crucial information
on the durability of this practice. The bioretention cells have been in operation since 1993
and are holding up very well. Third, it provided a site that could be monitored to document
the pollutant removal performance of bioretention cells as described in the next paragraph.
Fourth, it provided actual data that verify the very low and simple maintenance requirements and costs of these practices, which are approximately $200 per site/year.
300
Figure I.15.3 Beltway Plaza Bioretention Site illustrates how the impervious areas from this shopping center expansion were effectively disconnected and treated by designing the landscape
islands as bioretention cells. (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
Figure I.15.4 Beltway Plaza bioretention cells. The reported maintenance costs of bioretention cells
is low, averaging approximately $200/site. (Color version available at www.gsd.harvard.edu/
The pollutant removal performance of the bioretention cells at the Beltway Plaza
location was field monitored and, as shown in Table I.15.1, verified that it not only met
the water quality control criteria, but actually ranked as the most effective pollutant
Chapter I.15:
301
Table I.15.1 Pollutant Removal Performance of Bioretention

Practices (% Removal Rates)
Upper zone
Middle zone
Lower zone
Cu
Pb
Zn
TKN
NH4
NO3
TN
90
93
93
93
99
99
87
98
99
0
73
81
37
60
68
54
86
79
97
194
23
29
0
43
Note: The monitoring data for the Beltway Plaza site showed that bioretention cells exhibit very high pollutant removal rates for most
parameters.
Source: Davis et al. (1998).
Figure I.15.5 Schematic layout of bioretention cells at Largo Office parking lot illustrates the use of
bioretention practice as an urban retrofit for an existing parking lot. (Color version available at
removal BMP available. In addition, the success of this site suggested that bioretention
could be an effective retrofit BMP for existing urban areas.
Case study 3: Parking lot retrofit with bioretention Prince Georges

County Offices
The parking lot of the Prince Georges County Office Complex in Largo, MD had been
designed with conventional end-of-pipe SWM controls. The Prince Georges County government conducted a bioretention retrofit project to demonstrate that the bioretention cell
was an effective way to retrofit existing urban areas to provide water quality benefits.
Figure I.15.5 provides a site schematic map, while Figure I.15.6 shows the bioretention cell
in operation shortly after completion of construction.
302
Figure I.15.6 Operation of bioretention cell during a storm curb cuts intercept and direct runoff
into the bioretention cell for storage and treatment. (Color version available at www.gsd.harvard.
This demonstration site also served as an important milestone in the development of

LID technology in Maryland for a number of reasons. First, it demonstrated that existing
impervious areas could be cost-effectively retrofitted with bioretention cells to provide
water quality control. Second, it demonstrated and documented the use of various materials for the construction of the cells. Third, it provided another site that was monitored
to document the pollutant removal performance of bioretention cells in a retrofit setting.
Case study 4: LID comes of age Prince Georges County LID Design Manual
The bioretention case studies described above together with considerable technical analysis
of improved approaches for addressing the stormwater management impacts of development
activities culminated in the development of the Prince Georges LID design manual in 1997
(PGC, 1997). This manual is a milestone in LID technology for several reasons. First, it
provided a complete and systemic approach for integrating the major elements of LID design.
These include (1) using conservation, avoidance, and minimization, (2) reducing impervious
areas, (3) disconnecting unavoidable impervious areas, (4) maintaining the predevelopment
time of concentration, (5) selecting applicable control measures to approximate predevelopment hydrology, and (6) implementing public outreach and institutional activities.
The LID design manual describes and documents each step in the computational process
for maintaining predevelopment hydrology. In addition a series of detailed case studies is
provided in the appendices of the manual; the case studies demonstrate the application of
the LID process for single-family residential, townhouse residential, and a commercial project.
Case study 5: Detailed LID comparisons with conventional SWM design

In 1997 and 1998, Prince Georges County conducted a series of studies to compare the costs
of using LID design with conventional SWM design practices (Greenhorne & OMara, Inc.,
1998). Three case studies that compare LID with conventional design were developed with
Chapter I.15:
303
Table I.15.2 Construction Cost Comparisons for Patuxent Riding

Element
Grading, roadway R/W
Storm drain
SWM*
Bioretention, rain barrels
Total cost
Units
Unit cost
Costs
Conventional Low Impact
$569,698
$255,721
$260, 858
$0
$1,086,277
74
$14,679
$426,575
$132,558
$10,530
$252,124
$821,787
81
$10,146
Percent of Total Cost

Conventional Low Impact
52
24
24
100
52
16
1
31
100
* The SWM cost for LID is the SWM fee-in-lieu.

Note: The use of LID compared with conventional subdivision development provided an increase in
lot yield and a corresponding decrease in unit costs.
an emphasis on mitigating hydrologic impacts. The case study approach is valuable because
of the highly site-specific and interrelated outcomes of LID techniques, which cannot be
sufficiently evaluated if they have been applied only to idealized or theoretical circumstances.
The three case studies included the following:
1. Patuxent Riding (residential lots)
2. Pennsylvania Riding (residential townhouses)
3. Great Eastern Shopping Center (commercial)
The results of the Patuxent Riding case study are presented in Table I.15.2 and summarized next.
Costs
The LID design provided seven additional lots while achieving a significant cost decrease in
the development of the Patuxent Riding site. Cost comparisons are summarized in
Table I.15.2. These cost figures demonstrate significant savings in grading and roadway costs
and dramatic reductions in the costs for storm drains and SWM. Although the total roadway
length was increased due to the addition of a cul-de-sac, savings in roadway costs resulted
from decreases in both road width and road thickness. The storm drain costs were reduced
by nearly 50% on account of a greater proportion of the conveyance made available on the
surface in grassy swales instead of in underground pipes. With the elimination of the need
for two SWM ponds, the SWM cost was reduced to the fee-in-lieu. The savings in SWM costs
were approximately equal to the cost of bioretention and rain barrels.
The amount of woods preserved in both the conventional development and LID met
the Tree Conservation Plan (TCP) requirements of Prince Georges County. Because disturbance of the wooded area at Patuxent riding is slight, no cost difference was reported
between LID and conventional development in terms of meeting TCP requirements.
Benefits
In comparison with conventional development, the LID for Patuxent Riding retained more
water on-site, reducing runoff volume and increasing base flow in receiving streams. LID
eliminated the need for an off-site easement for the SWM Pond #1 outfall, and the cost
of maintaining SWM facilities was likewise eliminated. A major benefit to the developer
was the significantly reduced cost per unit. The seven additional lots were added while
the total construction cost was reduced by approximately $200,000. This case study was
significant because it clearly demonstrated that LID can be a winwin strategy for land
304
LID Site Planning

Define development
envelope
Reduce/Minimize total
site impervious areas
Disconnect impervious
areas
Modify/Increase
drainage flow paths
LID Hydrologic
Analysis
LID Public
Outreach Program
Define public outreach
program objectives
Identify target audience
Develop outreach
materials
Distribute outreach
materials
The LID
Approach
LID Erosion and

Sediment Control
Planning
Scheduling of operations
Soil erosion control
Sediment control
Maintenance
Delineate watershed and microwatershed areas

Define design storm
Define modeling techniques
Evaluate predevelopment
conditions and develop baseline
measures
Evaluate site planning benefits
and compare to baseline
Evaluate Integrated
management practices (IMPs)
Evaluate supplemental needs
LID Integrated
Management Practices
Define hydrologic control
Evaluate site constraints
Screen the IMPs
Evaluate most likely IMPs
Select IMPs
Incorporate additional
controls if necessary
Figure I.15.7 Major components of the LID approach include 10-site planning, LID hydrologic analysis, selection, and design of LID integrated management practices, erosion, and sediment control
for LID, as well as public outreach initiatives.
developers and environmental protection. Previously, it had always been presumed that
providing more or higher environmental protection was associated with higher development costs. LID, however, demonstrates that better technology can provide a winwin
strategy, better environmental protection, and reduced development costs.
Case study 6: LID goes national National design manual

Under a research grant from the U.S. EPA, the contents of the Prince Georges LID design
manual were expanded and modified into a national LID guidance document titled, Lowimpact development design strategies: An integrated design approach (U.S. EPA, 2000).
The national manual organized the LID design process into five major groupings, which
summarized in Figure I.15.7: (1) site planning, (2) hydrologic analysis, (3) selection and
design of integrated management practices (IMPs), (4) erosion and sediment control
requirements for LID, and (5) public outreach activities. The manual provides more
detailed guidance on the incorporation of LID technology into the site planning process
Chapter I.15:
305
than is found in the Prince Georges County design manual. Also, it provides more detail
on a range of microscale landscape-based practices known as IMPs.
Case study 7: Pembroke subdivision Total LID site design also provides
downstream peak discharge control
The Pembroke Subdivision is a residential development located in northern Frederick
County, MD, and is the first LID subdivision permitted in this county. Designed as an LID
subdivision from the start, it exhibits and benefits from the winwin attributes of the LID
process, which include:
1. The use of LID allowed the site design to eliminate the use of two SWM ponds,
which had been envisioned in an earlier concept plan for the site. This represents
a roughly $200,000 reduction in infrastructure costs.
2. In place of the SWM ponds, 2.5 acres of undisturbed open space and wetlands
have been preserved. Again, a considerable savings in wetlands mitigation impacts
was realized.
3. Also, the site plan gained two additional lots, increasing the 43-acre site yield from
68 to 70 lots. This added roughly $100,000 in additional value to the project.
4. Extensive use of site fingerprinting techniques allowed the site design to preserve
approximately 50% of the site in undisturbed wooded condition. This design
feature was very beneficial to maintaining the predevelopment CN.
5. Approximately 3000 linear ft of roads were converted from an urban road section
to a rural road. This design feature replaced curb and gutter with grass swales, a
$60,000 savings in construction costs. Also, the rural road section reduces the
paving width from 36 to 30 ft, a 17% reduction in paving costs.
In order to satisfy county criteria for adequate downstream conveyance, a downstream
impact analysis was conducted. The analysis examined the ability of an LID site design
to maintain predevelopment peak discharge conditions for a range of storms, including
the 1-, 2-, 10-, 50-, and 100-year storms. Many public works personnel perceive innovative
SWM techniques such as LID that are capable of addressing water quality issues but
insufficient to provide downstream peak discharge control for the larger flood flows
associated with the 10-, 50-, or 100-year storms.
Initially, the site LID hydrologic analysis was based on the 1-year storm (2.5-in. rainfall),
which is Frederick Countys criteria for water quality control. The downstream analysis
revealed that the 1-year storm design was not sufficient to maintain predevelopment peak
discharges for the 10-, 50-, and 100-year storms. An incremental iterative procedure was then
used to determine the additional control requirements that would be required to provide the
required downstream control. The analysis showed that by increasing the design storm to
the 2-year storm (3.0 in. of rainfall), the required downstream protection for the complete
range of flood events (10-, 50-, and 100-year storms) was achieved.
These results have great significance for future SWM policy and design criteria. They
clearly illustrate the tremendous advantages achieved by the volume control approach
incorporated in the LID technology. Volume control allows us to replicate the predevelopment rainfall runoff processes and helps to demonstrate that the peak discharge detention approaches provided with SWM ponds in conventional SWM designs are really
hydrologically flawed concepts. The flaws associated with the peak discharge detention
approach are numerous and include the following problems:
306

1. Peak discharge control does not typically address the maintenance of groundwater
recharge.
2. Peak discharge approaches alter the frequency and duration of flood flows, resulting in stream channel degradation.
3. Peak discharge approaches can actually aggravate downstream flooding conditions due to the superpositioning of runoff hydrographs.
4. Peak discharge approaches, particularly the use of regional facilities, provide no
protection for streams above the regional facilities.
Case study 8: Maryland 2000 SWM design manual

The State of Maryland Department of the Environment (MDE) is in the process of updating
its SWM regulations and requirements. One of the key elements of the update has been
the development of a new design manual, referred to as the MDE 2000 SWM design
manual (MDE, 2000). Although this statewide manual does not entirely embrace the total
LID design process, it does incorporate some elements of the LID process. The manual
encourages the use of the better site planning techniques as described by the Center for
Watershed Protection (CWP, 1998). Many of the better site design techniques are similar
to the measures described in the LID design manual to reduce the total amount of impervious area (TIA) and to disconnect as much of the remaining TIA as possible. The manual
also includes a requirement for maintaining the groundwater recharge, which is consistent
with the LID focus on replicating predevelopment hydrology.
Another feature of the MDE manual that is very similar to the LID design manual is
the incorporation of a series of stormwater credits. The stormwater credits consist of a
series of nonstructural practices that can play a significant role in reducing water quality
impacts, as well as reducing the generation of stormwater runoff from the site. Six of these
credits are described in the manual and are summarized in Table I.15.3. Figure I.15.8
provides a schematic representation of the natural area conservation credit, while
Figure I.15.9 provides a view of a subdivision built using this credit.
LID case studies outside Maryland

The focus of this paper is on LID-related activities in Maryland, but it is important to note
that the LID concept is being recognized as a viable alternative to the traditional
approaches to SWM management and is being rapidly implemented in many jurisdictions
throughout the United States. In addition, some applications are beginning to surface in
other countries. A few of these developments are briefly identified here.
Case study 9: LID for optimum water quality protection of water supply
reservoir High Point, NC
Due to its proximity to a proposed regional water supply reservoir, the city of High Point,
NC, is faced with the implementation of very stringent water quality controls related to
nutrients control (i.e., phosphorus) and limitations on TIA. As part of a watershed-wide
assessment and the development of a comprehensive stormwater management plan (Tetra
Tech, Inc., 2000), an evaluation of the benefits of using LID technology was conducted.
The evaluation revealed that the use of LID, particularly the incorporation of bioretention techniques, could optimize the removal of phosphorus by approximately 50% over
conventional pond-based BMPs. The bioretention cells can achieve phosphorus removal
Chapter I.15:
307
Table I.15.3 Summary of MDE (2000) Stormwater Credits

Stormwater Credit
Natural area
conservation
Disconnection of
rooftop runoff
Water Quality
Volume
Groundwater (gw)
Recharge Volume
Area draining to
natural area can be
subtracted from
water quality comps
Reduces runoff factor
(Rv) used in water
quality comps
Can be used
w/percent area
method to meet gw
recharge criteria
Can be used
w/percent area
method to meet gw
recharge criteria
Can be used
w/percent area
method to meet gw
recharge criteria
Can be used to reduce
gw recharge criteria
A predevelopment
forest/meadow
value can be used for
CN
Increased flow path =
longer Tc; can reduce
CN
Can be used
individually to fully
satisfy gw recharge
criteria
Fully satisfies gw
recharge criteria

CN
Disconnection of
nonrooftop runoff
Reduces runoff factor

(Rv) used in water
quality comps
Sheet flow to buffers
Contributing site area

can be subtracted
from water quality
comps
Can be used to satisfy
water quality volume
criteria
Open channel use
Environmentally
sensitive
development
Automatically meets
the water quality
volume criteria
Peak Discharge

CN
Can reduce CN

longer Tc; no CN
credit
Note: The new Maryland stormwater management design manual incorporates a number of stormwater credits
for using LID techniques.
levels ranging from 75 to 90% compared with the reported levels for SWM ponds, which
range from 40 to 50%.
The LID evaluation also reinforced another advantage of the LID technology with
respect to the TIA limitation requirement: a number of jurisdictions have begun to place
TIA limitations on a watershed scale as a surrogate for water quality control. This approach
is based on the TIA threshold concept reported in a number of publications (CWP, 1994).
For a specific site, however, the LID concept can provide a winwin strategy that optimizes
water quality objectives while allowing higher impervious cover for a given site. This dual
strategy is accomplished in two ways. First, the LID design methodology provides procedures and techniques to hydraulically disconnect impervious areas so that, for example,
a site with 70% impervious cover will be hydrologically equivalent to a site with 40 to 50%
impervious cover. The second part of this strategy results from the fact that the LID micromanagement practices can be incorporated into elements of the landscape, providing a dual
function for site features and thus preclude the need to dedicate and disturb (clear, grub, etc.)
8 to 10% of the total site for an SWM pond. Figures I.15.10 and I.15.11 compare the site design
features for a commercial site designed with conventional controls and LID techniques.
Case study 10: LID for commercial applications Florida Aquarium,

Tampa Bay, FL
The Florida Aquarium site is an 11.5-acre asphalt and concrete parking area that serves
approximately 700,000 visitors per year. An innovative stormwater management system
308
Figure I.15.8 Schematic representation of natural area conservation credit. The MDE 2000 manual
provides guidance on how to incorporate natural area conservation into site design and obtain a
stormwater management credit (MDE, 2000).
Figure I.15. 9 Photograph of a subdivision using the natural area conservation credit. (Color version
Chapter I.15:
309
Figure I.15.10 Randleman Lake Study commercial site using conventional stormwater management
to meet design criteria.
was designed using an SWM treatment train approach that incorporates many elements
of LID design (Rushton, 1999). Runoff is controlled using the following BMPs:
1.
2.
3.
4.
End-of-island bioretention cells

Bioretention swales located around the parking perimeter
Permeable paving
Bioretention strips between parking stalls
This project has incorporated a long-term monitoring program that is providing valuable performance data on the hydrologic and water quality performance of this facility.
Figure I.15.12 provides a site schematic plan, while Figure I.15.13 shows some of the
monitoring equipment installation.
A total of 16 storm events were monitored at the Florida Aquarium site during 1998.
The Southwest Florida Water Management District measured rainfall and flow from each
of the subcatchments in the parking area and collected water quality samples on a flowweighted basis. Comparisons between pavement areas controlled by the BMPs and uncontrolled asphalt areas were made for peak runoff rate, runoff volume, runoff coefficients,
and water quality. Sediment cores from swales also were collected and analyzed.
The parking areas controlled by the BMPs showed a significant reduction in runoff
volume and peak runoff rate. Table I.15.4 shows pollutant reductions for three pavement
types; reduction is compared to pollutant concentrations in runoff from a basin without
a swale.
310
Figure I.15.11 Randleman Lake Study commercial site using LID design conventional approach to
meet design criteria. Note that bioretention islands and bioswales replace ponds, and large blocks
of undisturbed forest are retained on-site.
Case study 11: University of Virginia bioretention demonstration project

The University of Virginia, Charlottesville, VA, has initiated a long-term study of the
performance of a bioretention cell, as shown in Figure I.15.14. This study differs from the
two bioretention studies conducted in Maryland that monitored a single storm event (3 in.
of rainfall). The UVA study is providing performance data based on an annual hydrologic
budget. Initial, first-year results indicate that, based on an annual budget, analysis of the
performance of the bioretention cells will exceed all expectations. First-year results are as
follows (Yu et al., 1999): 86% for TSS, 90% for TP, 97% for COD, and 67% for oil and grease.
Case study 12: Start at the source manual San Francisco Bay Area
The San Francisco Bay Area Stormwater Management Agencies Association has published
a guidance manual titled Start at the source: Residential site planning and design guidance manual for stormwater quality protection (BASMAA, 1997). This document provides guidance for incorporating landscaped-based micromanagement BMPs at the site
level. Much of the guidance provided in this manual is very similar to the LID design
manual developed for Prince Georges County, MD.
Other case studies

A number of additional case studies are being conducted throughout the country. Space
and time limitations preclude more detailed treatment in this paper and will therefore be
reported at a future date. Some of these projects are briefly listed here:
Chapter I.15:
311
Figure I.15.12 The Florida Aquarium LID site features a treatment train approach that includes porous
pavement, bioretention strips, and ponds. (Color version available at www.gsd.harvard.edu/
1. The Maryland State Highway Administration (SHA) has recently begun an initiative that is evaluating the use of LID practices for urban and rural roads applications. The ultimate goals of this project is to develop standards and specifications
for these applications, as well as the installation and monitoring of selected practices.
2. The U.S. EPA is nearing completion of a feasibility study that examines the potential for applying LID technology to retrofit existing urban areas.
3. Rockdale County, GA, has begun a series of demonstrations that will evaluate the
application of LID technology in Rockdale County.
4. Jordan Cove Project, CT: a paired watershed study is being conducted in a joint
effort between the U.S. EPA and the University of CT. The objective of this study
is to compare the impacts at a watershed scale of conventional vs. innovative SWM
control practices. The innovative control practices include many of the LID techniques.
5. LID demonstration projects in the early stages of development are being reported
in MA, MN, SD, ID, DC, and Seattle, WA.
6. At the international level, LID initiatives are being reported in Canada, France,
Germany, Japan, New Zealand, and Taiwan.
312
Figure I.15.13 Florida Aquarium site monitoring stations. Porous pavement and bioretention strips
help achieve a 40% reduction in runoff. (Color version available at www.gsd.harvard.edu/ watercolors. Password: lentic-lotic.)
Table I.15.4 Load Efficiency of Pollutants Expressed as Percent Reduction for Three
Types of Pavement at the Florida Aquarium Site
Constituents
Ammonia
Nitrate
Total nitrogen
Ortho-phosphorus
Total phosphorus
Suspended solids
Copper
Iron
Lead
Manganese
Zinc
Percent Pollutant Reduction*

Asphalt w/Swale
Cement/Swale Permeable w/Swale
45
44
9
180
94
46
23
52
59
40
46
73
41
9
180
62
78
72
84
78
68
62
85
66
42
74
3
91
8
92
85
92
75
* The basins with swales were compared with a basin without a swale to determine the amount
of reduction in pollutant loads possible using these small alterations. Notice that the efficiencies or phosphorus are negative, indicating an increase in phosphorus load in the basins with
a swale. This emphasized the importance of source control as well as runoff volume reduction.
Summary
This chapter provides a brief summary of a number of projects that illustrate the development and application of the LID technology for stormwater management. The LID
technology as documented in the Prince Georges County and the U.S. EPA national
manuals provides a significant advance in the art and technology of stormwater management. This technological breakthrough is derived from a number of sources. The first is
the attempt to define and replicate essential predevelopment hydrologic functions. The
Chapter I.15:
313
Figure I.15.14 The UVA Bioretention Demonstration Project is providing data on the performance of
bioretention cells on an annual basis. (Color version available at www.gsd.harvard.edu/watercolors.
second is the use of an integrated interdisciplinary approach that differs considerably from
the single-parameter, hydraulic-based approaches of the past. The third is derived from
the microscale or at the source control concept.
Stormwater management issues and alternatives can differ substantially on a watershed-by-watershed basis. Most watersheds exhibit a wide variety of interrelated stormwater management problems and issues. Some watersheds, or portions of watersheds, are
dominated by older, established urban areas that were developed prior to the advent of
stormwater management technology. The needs of these watersheds have largely not been
adequately addressed, if they have been addressed at all in the past. Other watersheds,
or portions thereof, are predominantly suburban and may be subject to various control
policies that have developed over time. Still other watersheds may be predominantly rural
and can present a mix of agricultural and developing watershed issues. This wide variety
of conditions suggests the need for stormwater management policies and tools that can
address the wide range of issues at the watershed level.
The projects and case studies presented in this paper try to provide a representative
cross-section of these types of issues. Case studies are presented that address the following
conditions:
1.
2.
3.
4.
Water quality control for infill development

Water quality retrofits for existing urban areas
LID technology to replicate predevelopment hydrologic functions
Volume-based LID technology to provide real downstream peak discharge control
for flood events
5. LID control for optimum water quality control in water supply reservoir watersheds
6. LID technology to address the TIA issue
314
The author recognizes that the projects and case studies provided in this paper represent only a small sample of all the LID projects that exist. The paper does not make any
representation that these are the best available examples of LID. They are included primarily on the basis of two criteria: (1) the author is familiar with and has some documentation for these projects; and (2) the projects are helpful in describing the range of applications that can be achieved with LID technology. Finally, the author invites contact from
readers who are knowledgeable of other LID projects, so information can be shared and
the LID database improved.
Literature cited
Bay Area Stormwater Management Agencies Association (BASMAA), Start at the source: Residential
site planning and design guidance manual for stormwater quality protection, prepared by Tom
Richman and Associates, Palo Alto, CA, 1997.
Davis, A. et al., Optimization of bioretention design for water quality and hydrologic characteristics,
report 01-04-31032, Final report to Prince Georges County, MD, 1998.
Center for Watershed Protection (CWP), The importance of imperviousness, Watershed Protection
Tech., 1, 1994.
Center for Watershed Protection (CWP), Better Site Design: A Handbook for Changing Development
Rules in Your Community, Center for Watershed Protection, Ellicott City, MD, 1998.
Engineering Technologies Associates, Inc. (ETA), Design manual for use of bioretention in stormwater
management, prepared for the Dept. Environ. Resour. (DER), Prince Georges County, MD,
1993.
Greenhorne & OMara, Inc. Case studies in low-impact development: stormwater management
comparison and economics evaluation, prepared by Greenhorne & OMara, Inc, Greenbelt,
MD, for the DER, Prince Georges County, MD, 1998.
Maryland Department of the Environment (MDE), 2000 Maryland stormwater design manual, Vols. I
& II, prepared by the Center for Watershed Protection and the MDE, Water Management
Administration, Baltimore, MD, 2000.
Prince Georges County (PGC), MD, Low-impact development design manual, DER, Prince Georges
County, MD, 1997.
Rushton, B., Low-impact parking lot design reduces runoff and pollutant loads, annual report #1,
Southwest FL Watershed Manage. Dist., Brooksville, FL, 1999.
Tetra Tech, Inc., City of High Point, NC, Deep River 1 watershed assessment and stormwater plan,
prepared by Tetra Tech, Inc., Research Triangle Park, NC, 2000.
U.S. EPA, Low-impact development design strategies: An integrated design approach, prepared by
Tetra Tech, Inc., Fairfax, VA, for DER, Prince Georges County, MD, funding provided by
the U.S. EPA, Washington, DC, 2000.
Yu, S.L., Zhang, X., Earles, A., and Sievers, M., Field testing of ultra-urban BMPs, Proc. 26th Ann.
Water Resour. Plan. Manage. Conf., Wilson, E., Ed., ASCE, Tempe, AZ, 1999.
Chapter I.15:
315
Response
Values of demonstration projects and case studies
of stormwater source management
This chapter by Michael Clar clearly demonstrates that low-impact development represents a significant advance in the technology and art of stormwater management. As
introduced in Chapter I.5 by Larry Coffman and adapted by Richard Pinkham and Timothy Collins in Chapter I.4, Robert France and Philip Craul in Chapter I.7, and Thomas
Liptan and Robert Murase in Chapter I.6, low-impact development presents an integrated
and interdisciplinary approach to water sensitive design that differs considerably from
the single-parameter hydraulic approaches of previous stormwater management.
Clar makes the point that because the stormwater management issues differ from site
to site due to watershed specificity, low-impact development approaches must be flexible
in order for them to be adapted to any situation. This is why demonstration projects such
as those described in this chapter offer convincing evidence toward the eventual general
acceptance of these new stormwater technologies. The fact that one of the case studies
presented by Clar was a retrofit located in the parking lot right beneath an environmental
directors office window goes far to limiting charges of hypocrisy that developers often
(and justifiably) voice toward environmental regulators; i.e., the wisdom and strength of
practicing what you preach cannot be overestimated. Another interesting case study
shows the strategy of breaking up the flow paths of large parking lots into small, manageable drainage sub-catchments where bioretention swales can be easily fitted.
This chapter adds to the discussion by presenting evidence obtained from comparative
economic analyses that low-impact development approaches can actually have lower
construction and maintenance costs while, at the same time, offering improved aesthetics
and other multifunctional uses of the urban landscape. The concept of stormwater credits
for uncoupling runoff from impervious surfaces is a direction worth pursuing further.
chapter I.16
Restoring and protecting a small, urban

lake (Boston, Massachusetts)
Nicholas Pouder and Robert France
Abstract
Chandler Pond, a small urban lake in Boston, has experienced cultural eutrophication due
to watershed inputs of phosphorus. By the late 1990s, the problem had become so severe
that the bottom sediments of the lake, where much of the phosphorus was stored and
frequently resuspended to promote algal growth, were removed by dredging. Although
such dredging immediately improved water quality, it was determined that an underlying
cause of the problem the continued inputs of new aqueous phosphorus and nutrientladen sediments to the lake through stormwater runoff and shoreline erosion had not
been fully addressed. The present project focused on raising public awareness about the
possibilities of mitigating future eutrophication through on-site methods. Specifically, a
demonstration project restored a section of the eroding shoreline through bioengineering
techniques, in and around which were placed wetland plants to stabilize the soil, cleanse
inflowing stormwater, and provide aesthetic and recreational opportunities. A plan to
address inputs from the watershed is also being prepared.
Introduction
Urban lakes face a host of environmental problems arising from their location within
densely populated cities (Tourbier and Westmacott, 1992). Fertilizer application to residential lawns and community golf courses, as well as frequent deposition of dog feces,
ensure that a large pool of nutrients is available within (sub)urban watersheds. Given the
high degree of impervious surfaces in such settings (Chapter I.3), stormwater runoff is a
serious problem in terms of water quantity (ASLA, 1996; Chapter I.1). When this runoff
moves across vegetated surfaces such as lawns and golf courses, phosphorus from fertilizer
and dog waste become mobilized, and together with suspended sediments, can be transported to urban waters where they pose a serious threat to water quality (Terrene Institute,
1994). The resulting increased algae and vascular macrophyte growth contribute to the
bottom sediments of these lakes upon plant death, eventually decreasing water depth as
the nutrient-laden sediments accumulate. Wind-driven resuspension of bottom sediments
over shallow-water lakes as well as extensive shoreline erosion can both continually
supply nutrients to the water column and support yet further primary production, thereby
1-56670-562-2/02/$0.00+$1.50
317
318
exacerbating an already serious water quality problem. The problem of eutrophication of

urban lakes has three obvious solutions:
1. One might think that the most obvious and successful procedure for restoring
eutrophied urban lakes would be to simply reduce or stop the input of all anthropogenic phosphorus. As an ultimate strategy, however, such an option is limited
due to the large source of nutrients contained within the sediments now situated
at the bottom of such lakes. Due to the accumulation of nutrients previously
transported from the watershed to the lakes, and the potential of those nutrients
to serve as a permanent and renewable source of enrichment due to wind resuspension, it is often never enough to simply halt the input of new sediments or
nutrients. This becomes very much a case of the sins of fathers and mothers
being exacted upon the offspring.
2. In such circumstances, it becomes necessary to restore urban lakes by actively
removing past nutrient inputs. Deepening lakes in a sense setting back the clock
of their degradation caused by urbanization by dredging bottom sediments is
a very costly, yet becoming all-too-necessary, technique to rejuvenate lakes to a
more pristine past. Many examples of dredging rehabilitation exist in urban settings such as Boston (e.g., Barnett, 1999).
3. It would be wrong, however, to think that once the symptom of lake eutrophication
is temporarily fixed, that the work of environmental managers is over and the
problem(s) alleviated or mitigated. Dredging urban lakes with no attention paid
to watershed causality or implementation of an effective plan to reduce future
nutrient inputs is like treating cancer with a bandage (France, 1999). Only through
the use of retrofitted BMPs (or best management practices) (e.g., Schueler et al.,
1992; Chapters I.4, I.5, and I.15) and altered community development rules (Center
for Watershed Protection, 1999), as part of a program of comprehensive watershed
planning and management (Center for Watershed Protection, 1998; Chapters II.8
and II.9), can one really say that any urban lake is truly restored (see Chapters II.4
and I.9 for descriptions of two other Boston-area urban lakes).
Due to the presence of large neighborhood populations and obvious shortages of green
spaces, urban lakes are frequently characterized by extensive shoreline erosion brought
about by shoreline strolling and dogs running in and out of the water (Chapter II.4). These
shorelines are not only aesthetically unappealing, but they can also contribute to lake
eutrophication in two important ways:
Upon eroding into the water, shoreline soils break apart and release their burdens
of nutrients for uptake by planktonic algae and nuisance aquatic plants.
Due to bankside instability, nearshore aquatic plants are often unable to become
established, thereby reducing the capability of this first line of eutrophication
defense in the phyto-uptake of high nutrient stormwater runoff before the algae
can assimilate it.
The purpose of our intervention in an urban lake was to create a demonstration project
that would raise neighborhood awareness about the opportunities present for cost-effective in-lake management of eutrophication. By restoring a section of the severely degraded
shoreline, we hoped to reduce the potential for future, much more expensive, approaches
to lake restoration such as periodic dredging.
Chapter I.16:
Restoring and protecting a small, urban lake (Boston, Massachusetts)
319
Figure I.16.1 Chandler Pond before lake restoration. Note abundant aquatic plants.
Project history
Chandler Pond (Figure I.16.1), located at Gallagher Memorial Park in the Brighton section
of Boston, presents many of the attractions and problems typical to contemporary urban
wetland-ponds in microcosm. It is a 10.8-acre man-made pond, constructed in the 19th
century for ice production; over the years it has become heavily contaminated as a result
of sediment input from its highly urbanized 185-acre watershed, 45% of which is residential and 55% recreational (parks and golf course).
The ponds original average depth was approximately 3 ft, yet over subsequent
decades sedimentation reduced it to 1.9 ft. As a result, poor water quality and excessive
growth of nuisance aquatic macrophytes degraded the ponds habitat, recreational, and
aesthetic values. As a first step to correct the problem, the Boston Parks and Recreation
Department (BPRD) implemented a project to restore the pond to its original depth. In
1999 the pond was drained (Figure I.16.2) and subsequently dredged (Figure I.16.3; Heron,
1998). The removal of 25,000 yd3 of nutrient-rich sediment provided immediate improvement of overall pond conditions by increasing water depth and reducing the availability
of nutrients to aquatic macrophytes, which had become a nuisance prior to dredging. Yet
the underlying cause of sedimentation had not been remedied.
A special condition of the Boston Conservation Commissions (BCC) approval for the
dredging project encouraged the BPRD to address potential upstream sources of sedimentation, and nutrients which may lead to future siltation and eutrophication of Chandler Pond (BCC, 1998). In response, the BPRD conducted a preliminary assessment of
the site, which identified shoreline erosion as a significant problem contributing to siltation of the pond, excessive nutrient growth, and degraded wildlife habitat (McLaughlin,
2000) (Figure I.16.4). Excessive pedestrian use and overgrazing by resident waterfowl were
determined to be potential causes of this shoreline erosion.
To address the on-site erosion, the BPRD, under the guidance of Timothy Smith,
applied for and received grant funding from the Massachusetts Department of Environmental Management, Lake and Pond Grant Program and retained the Pouder Design
Group to prepare a shoreline master plan and demonstration shoreline stabilization
project. Stated goals of the project were to:
320
Figure I.16.2 Chandler Pond during draining in summer 1999.
Improve water quality

Enhance habitat for wildlife
Reduce on-site inputs of sediment, urban runoff, nutrients, and other contaminants
into the pond
Improve passive recreational opportunities
Recreate a naturalized riparian wetland in an urban setting
Develop a project that would serve as a model for other urban wilds restoration
projects in the city
The project team began their work by assessing the severity of shoreline degradation.
Native and naturalized vegetation were observed on approximately one-third of the
5000-ft. shoreline. A dense monotypic stand of cattails 10 to 50 ft wide existed along the
ponds western edge, and several small patches of cattails were scattered around the
Chapter I.16:
321
Figure I.16.3 Dredging operations at Chandler Pond in summer 1999.
remaining perimeter. Smaller patches of purple loosestrife and Japanese knotweed were
distributed around the pond. A portion of the northeastern and northwestern shoreline
was densely vegetated with woody species.
The remainder of the shoreline was either covered with turf or unvegetated. With the
exception of several clumps of vegetation and areas of turf, virtually the entire southern
shoreline was barren of vegetation (Figure I.16.5). These vegetated clumps typically contained a single tree with a shrub layer and/or emergent or submergent herbaceous species.
Lawn areas bordering the pond were mown to the edge of water, limiting riparian habitat
and reducing buffering capacity.
Some erosion was also evident along the steep slopes of the ponds privately controlled
northern shoreline. With varying success, homeowners had implemented haphazard
methods of slope stabilization including walls and vegetation. Yet it was immediately
apparent that the most unstable and severely degraded portion of the shoreline was located
along the ponds publicly accessible southern side. Areas of bank erosion coincided with
the areas of highest observed pedestrian use. Park visitors in this portion of the site were
frequently seen feeding waterfowl or fishing, and erosion was most severe in favorite
fishing and waterfowl feeding areas. The erosion caused unstable slopes and an undercut
shoreline (Figure I.16.4). Gullies formed in some areas.
Pedestrian circulation at the site consisted of a wide asphalt sidewalk directly adjacent
to Lake Shore Road and an eroded desire path parallel to the shoreline approximately
midway between the sidewalk and the ponds southern shoreline. A pedestrian/maintenance vehicle path was located within the park to the west of the pond.
Despite its degraded condition, the pond provided habitat for numerous species of
birds, reptiles, amphibians, fish, and small mammals (Figures I.16.6 and I.16.7). BPRD
staff and consultants, as well as area residents, reported sightings of numerous species.
Canada geese are a dominant species with dozens of resident birds, well above BPRDs
estimate of an acceptable population. The high population of geese at the site causes
excessive erosion as noted previously. Nutrient loading from goose droppings is also a
concern. Furthermore, goose droppings diminished the passive recreational potential of
lawn areas.
322
Figure I.16.4 Shoreline regions of Chandler Pond exhibiting substantial erosion.
Planning and design recommendations

To address the on-site causes of erosion and sedimentation, a long-range Shoreline Master
Plan was prepared by the Pouder Design Group with input from BPRD, W.D.N.R.G.
Limnetics, and the Chandler Pond Preservation Society (CPPS), a nonprofit community
action group comprised of interested citizens and area residents. To guide future park
improvements, conceptual and generalized management guidelines, infrastructure improvements, bank stabilization, and vegetative enhancements were proposed (Figure I.16.8). One
such recommendation was the immediate implementation of a demonstration project
Chapter I.16:
323
Figure I.16.5 Shoreline habitat suffering from excessive pedestrian use and waterfowl grazing.
Figure I.16.6 Searching for turtles following draining.
(discussed next), which could address the most serious areas of bank degradation, as
funding permitted.
Other recommendations of the long-range plan included:
Reduce the frequency of mowing of a band 10 ft wide along the pond edge while
continuing the frequency of mowing for the remainder of turf to (a) provide
enhanced filtration of runoff, (b) discourage geese from using turf areas, thereby
reducing the amount of goose droppings on lawn areas, (c) reduce nutrient loading
(d) lower maintenance costs, and (e) allow for the self-sowing of riparian species.
Retain areas of mown turf for passive recreation.
Stabilize all portions of the eroded bank using biological (plant) controls; provide
coir fiber rolls in areas of severe erosion or bank degradation.
324
Figure I.16.7 One of the dozens of turtles rescued following draining and preceding dredging.
Create new wetland areas at stormwater inlets to act as sediment traps.

Control access to the water, thereby limiting destruction of riparian areas by excessive foot traffic.
Remove the existing sidewalk and desire path, replacing them with a single
Americans with Disabilities Act (ADA)-compliant walkway.
Link the new walkway to two ADA-compliant fishing structures.
Restrict and discourage access to prime bird breeding habitat.
Create a boardwalk and bird-viewing stand to allow continued bird-watching
opportunities in this area while at the same time limiting disturbances of nesting
birds.
Retain existing and provide additional subsurface structure for wildlife habitat.
Develop, in conjunction with the CPPS, ecological educational and signage programs directed at watershed residents and park uses.
Monitor and photo-document treated and untreated areas both pre- and postproject for chemical, physical, and biological indicators of sediment and water
quality.
Shoreline restoration demonstration project

Installation of the pilot/demonstration program was intended to assess the effectiveness
of low-cost shoreline stabilization methods. The most severely degraded areas within the
park along the ponds publicly accessible southern shoreline were selected for this purpose. A primary goal of the demonstration project was to identify which of these costeffective restoration treatments could be used to stabilize other degraded shoreline areas,
both at Chandler Pond and at other locations.
The demonstration project proposed the preservation of native existing vegetation;
hand removal of invasive vegetation; and introduction of native species in deep emergent,
shallow emergent, wet meadow/pond edge, and upland habitats. Selected plant species
were typical of natural freshwater marshes in eastern Massachusetts and were acquired
from a Massachusetts source.
325
Figure I.16.8 Management options for the restoration of Chandler Pond.
Chapter I.16:
326

Other goals of the demonstration project were to:
Enhance needed feeding, brooding, and protective habitat for a range of native
insects, birds, herpitiles, and fish.
Stabilize bank soils and physically trap sediment and nutrients, making these
unavailable to nonnative, invasive, nuisance species.
Chemically retain sediment, nutrients, and other pollutants within plant tissue,
further limiting their availability to nonnative, invasive, nuisance species; and
promote the self-sowing of other native wetland plants.
Preserve water and locate new plantings to retain important view corridors. Use
of low plantings in foreground areas would frame views of and across the pond.
Community outreach
Community involvement was considered a key factor in the successful outcome of the
project. The CPPS sponsored an evening information session at which the consultant
presented plans of the demonstration project, explained the process, and solicited feedback. CPPS members were strongly committed to the success of the project and provided
volunteer labor for watering of newly installed plants, removal of invasive vegetation,
and informal surveillance of the project to minimize vandalism. Additionally, the CPPS
including regular articles about the project in its newsletter.
Implementation
To evaluate alternative methods of bank stabilization, three combinations of plants and
biologs were proposed. These were:
A combination of biologs and vegetation
Biologs only
Vegetation only
Eleven restoration areas (Figure I.16.9) of native vegetation consisting of 26 different
species (Table I.16.1) were proposed along unvegetated portions of the southern shoreline.
Plantings (Table I.16.2) were proposed to begin about 2 ft above the waterline and extend
into the pond to a water depth of 2 ft in a zone 6 to 10 ft wide. Planting was also designed
to deter travel by waterfowl between the land and water, thus reducing the impacts of
waterfowl excrement. Equally important was the desire to provide an attractive visual
display of colorful flowers and foliage throughout the growing season and to both enhance
and preserve views of the lake from the adjacent shoreline (Figure I.16.10).
Planting depth zones (Figure I.16.11) were based on inundation tolerance. Deep emergent marsh species were installed in average pool depths of 12 to 24 in. Shallow emergent marsh species were planted in average pool depths of 3 to 12 in. Wet meadow edge
plants were planted directly in the saturated biolog and from an elevation of +6 to 3 in.
Shoreline fringe species were planted at elevations greater than +6 in. The shoreline fringe
plants required regular watering during their grow-in period. Approximately 2750 plants
were installed.
Installation was conducted by a habitat restoration specialist. Planting commenced in
early June and was completed in 2 weeks. Plants and biologs were delivered to the site in
the early morning. Plants were stored in a cool, moist location to reduce desiccation. Installation began in late morning. Eleven areas were selected for intervention (Figure I.16.9); these
Chapter I.16:
Figure I.16.9 Planting plans for shoreline restoration of Chandler Pond. See Table I.16.2 for planting instructions to
installation personnel. Plant abbreviations are described in Table I.16.1.
327
Scientific Name
Aster novae-angliae
Alnus rugosa
Cornus amomun
Calamagrostis canadense
Chelone glabra
Carex lacustris
Carex lurida
Cornus sericea
Carex vulpinoidea
Dulichium arundinaceum
Eupatorium dubium
Iris versicolor
Juncus effusus
Lobelia cardinalis
Leersia oryzoides
Pontederia cordata
Peltandra virginica
Rosa palustris
Rhododendron viscosum
Scirpus acutus
Sambucus canadensis
Sagittaria latifolia
Scirpus pungens
Sparganium americanum
Scirpus validus
Veronica noveboracensis
Code
AN
AR
CA
CC
CG
CLA
CLU
CS
CV
DA
ED
IV
JE
LC
LO
PC
PV
RP
RV
SA
SC
SL
SP
SPA
SV
VN
Wet meadow/edge
Shoreline
Shoreline
Shallow emergent
Wet meadow/edge
Shallow emergent
Wet meadow/edge
Shoreline
Wet meadow/edge
Shallow emergent
Wet meadow/edge
Shallow emergent
Shallow emergent
Wet meadow/edge
Shallow emergent
Deep emergent
Deep emergent
Shoreline
Shoreline
Deep emergent
Shoreline
Deep emergent
Shallow emergent
Shallow emergent
Deep emergent
Wet meadow/edge
Zone
FACW
FACW+
FACW
FACW+
OBL
OBL
OBL
FACW+
FACW+
OBL
FAC
OBL
OBL
FACW+
OBL
OBL
OBL
OBL
OBL
OBL
FACW
OBL
OBL
OBL
OBL
FACW+
Wetland Indicator
12" o.c.
As noted
As noted
12" o.c.
12" o.c.
12" o.c.
12" o.c.
As noted
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
As noted
As noted
12" o.c.
As noted
12" o.c.
12" o.c.
12" o.c.
12" o.c.
12" o.c.
Spacing
150
8
10
100
100
150
150
12
50
150
50
250
200
50
150
150
150
5
10
100
5
150
150
250
100
100
Quantity
328
New England aster

Alder
Silky dogwood
Blue joint grass
Turtlehead
Lake sedge
Lurid sedge
Red twig dogwood
Fox sedge
Three-way sedge
Joe pye weed
Blue flag iris
Soft rush
Cardinal flower
Rice cutgrass
Pickerel weed
Arrow arum
Swamp rose
Swamp azalea
Hard stem bulrush
Elderberry
Duck potato
Three square bulrush
Burreed
Soft stem bulrush
New York ironweed
Common Name
Table I.16.1 Planting Schedule, Chandler Pond Riparian Restoration Project
Chapter I.16:
329
Table I.16.2 Notes on Planting for Construction Personnel

Remove all purple loosestrife and Japanese knotweed, including all roots, encountered in planting
areas.
Grade eroded and undercut shoreline slopes to provide smooth transition between upland and
subsurface areas or as directed by BPRD.
Use of power equipment for excavation and grading is prohibited.
Work areas are shown in their approximate locations. Verify work areas with BPRD prior to
planting.
Use no fertilizers unless directed to do so by BPRD.
Plant materials provided by BPRD shall be planted on day of delivery or shall be stored and
maintained in healthy condition by the contractor. Store plants in a secure, shaded, out of wind,
and well-ventilated site.
The contractor shall ensure that soil conditions at time of planting are sufficiently moist to
maintain plant health.
Plants shall be carefully handled to prevent injury and desiccation. Plants installed above the
waterline shall be watered by the contractor.
Prior to planting, contractor shall stake out all planting areas for approval by BPRD. Plants shall
be installed upright. All planting procedures, tools, and methods subject to approval by BPRD.
Plugs shall be planted in the soil at least one (1) inch deeper than grown in the nursery and to a
depth which ensures that the root structure is at least one inch but not more than two in. Below
the soil surface.
Fencing shall be installed at each planting area immediately following planting.
The wetland plant material shall be evaluated for acceptance thirty (30) days after the following
have been completed or satisfied:
1. All specified fencing has been installed and accepted adjacent to the planting areas.
2. Slopes have been graded to the satisfaction of the BPRD.
3. The plant material has been installed to the satisfaction of the BPRD.
consisted of one area stabilized only with a biolog, six areas that received a combination
of biolog and plants, and four areas at which only plants were used.
The contractor installed biologs by excavating a shallow trench parallel to the shoreline
(Figure I.16.12). Excavated soil was used to backfill on the upland side of the biolog. Some
biologs were set directly against bank; others were set as much as 24 in. out from bank,
thereby creating a shallow ponded area on the shore side of the biolog (Figure I.16.13).
Hardwood stakes were driven into the substrate on both the waterside and the shore side
of each biolog.
Fencing
Installation of temporary fencing (Figure I.16.14) was also proposed to enclose newly
planted areas and reduce herbivory from waterfowl and muskrats. The fence consisted of
2-in. square wood stakes and biodegradable twine strung in 10-ft cells at several levels
both below and above the water surface (Figure I.16.15). The fence was intended to remain
in place for a 1-year period, until plantings had sufficiently rooted and were no longer
vulnerable to herbivory. Within days of installation, however, it became apparent that
unleashed dogs were breaking through the goose exclusion fence into at least one planting
area. Waterfowl subsequently entered the area through the broken fence and damaged a
high percentage of newly installed plants. To educate dog owners about the impact of
their actions, laminated color signs describing the project were subsequently installed
(Figure I.16.16). The signs contained a brief description of the project and included color
before and after images of a stable shoreline.
330
Figure I.16.10 Vegetation a few months after shoreline planting.
Monitoring
Although the long-term success of the project cannot be determined for several years,
some preliminary observations were apparent. First, the use of biodegradable twine for
the fence echoed the naturalistic approach of the project; however, a more durable fence
should be used in urban areas where dogs may be present, especially on the upland side
of planting areas. Similarly, signage should be installed concurrently with plantings to
alert dog owners to the potential damage that their pets might cause. Flexibility with plant
Chapter I.16:
331
Figure I.16.11 Schematic representations of planting depth zones.
species selection was also required. A late-season crop failure at the supply nursery
required the substitution of several species of plants. Last, arrangements for secure storage
and watering of plants must be provided if they are not installed on the day of delivery.
In terms of the lake itself, a year following shoreline treatment, most of the restoration
areas exhibit strong growth of the species that were planted. One localized problem that
developed is the aggressive colonization of purple loosestrife along most of the shoreline,
possibly a result of the water drawdown that preceded bottom sediment dredging (Smith,
1999). However, a more serious and systematic problem not expected to be anywhere near
332
Figure I.16.12 Installation of biologs.
completely addressed by our small-scale demonstration project concerns the bloom of

algae that has returned to the pond to a degree comparable to what it was prior to
dredging. This has led to feelings of anxiety and confusion by pond-side residents (Smith,
1999), leading them to question the possible waste of hundreds of thousands of dollars
spent on the dredging, especially in light of the consequent inconvenience that ensued
during the entire process. This result underscores the need for a much more comprehensive
watershed approach to ensure true sustainability of restoration projects.
Future scaling up to the watershed

The primary aim of the present project was to protect the investment of recent dredging
in Chandler Pond. The polluted stormwater continues to flow into Chandler Pond unaltered, so the expectation must be that the sedimentation process, augmented by continued
shoreline erosion from those areas outside of the present demonstration project, is already
under way again. Unless steps are taken to reduce the nutrient loading and sedimentation
of the pond, the dredging will prove to be no more than a temporary solution. With current
understanding of shoreline bioengineering and wetland treatment techniques in a framework of watershed management (Chapters I.17, II.4, and I.14), there is no excuse for letting
Chandler Ponds water quality deteriorate again. At the start, it is recognized that the
steps needed to achieve a stable and acceptable water quality in the pond are manyfold,
and that a comprehensive review of the Chandler Pond watershed and a community
liaison/education program would be an important part of the strategy.
Chapter I.16:
Figure I.16.13 Schematic representation of biolog installation.
333
334
Figure I.16.14 Schematic representation of waterfowl fence.
In this respect, it is always important to place site-specific restorative designs in both

a larger landscape and a larger management context, in this case dealing with where the
two meet the watershed. Watersheds are important to urban communities because they
help to foster an embodiment of sense of place in a developed landscape and because
their waters aid in providing much-needed solace in daily lives. Importantly from an
environmental perspective, watersheds are the most appropriate units in which to effectively manage water resources in the urban landscape (Chapters II.8 and II.9). After years
of deleterious human activity in its watershed, Chandler Pond has undergone a vastly
expensive dredging operation to rejuvenate its health. It is egregiously nave, however, to
think that once the symptom is thus fixed, the work is over and the problem is mitigated
or alleviated. (The recently returned blooms of algae indicate that the lake is already
becoming stressed again.) A cardinal need exists to put Chandler Pond back into its
landscape and to develop an effective and comprehensive watershed management plan
that will prevent, or at least reduce, the severity of the problem(s) that necessitated the
dredging in the first place (Table I.16.3). Chandler Pond desperately needs a watershed
plan similar to those recently completed for nearby Halls Pond in Brookline (Chapter I.11)
or Fresh Pond in Cambridge (Chapter II.4). This is the only working solution to protect
the environmental integrity of the temporarily restored Chandler Pond, lest we face
another, by then perhaps prohibitively expensive, redredging several decades in the future.
Toward this end, a preliminary report (France, 1999) provides a first attempt at itemizing
the requisite elements needed to be addressed in development of such a master watershed
plan for Chandler Pond (Table I.16.3), some of the dangers to be avoided in the process,
and a preliminary estimated budget needed to accomplish this task.
Chapter I.16:
335
Figure I.16.15 Waterfowl exclusion fencing and netting in restoration areas.
To tackle issues of urban watershed renewal and repair necessitates a comprehensive

vision that is best approached in small steps in order to ensure the greatest likelihood of
success (EPA, 1997). The BPRD is also working with W.D.N.R.G. Limnetics to compile a
long-term, watershed-based management plan, a part of which is intended to reduce and
mitigate sediment and nutrient sources before stormwater enters Chandler Pond.
Major elements of the post-dredging watershed plan include both structural and
nonstructural corrective elements, many of which are addressed through the shoreline
336
CHANDLER POND
SHORELINE
R E S TO R AT I O N
Before
After
This project will reduce shoreline erosion and

provide habitat for fish and wildlife. The
temporary fence is needed to prevent geese
from damaging young plants. Please help
us by keeping your dog out of the planting
areas and on a leash.
Thomas M. Menino, Mayor
Justine M. Liff, Commissioner
THIS PROJECT WAS MADE POSSIBLE IN PART BY A
GRANT FROM THE LAKE AND POND GRANT PROGRAM
ADMINISTERED BY THE MASSACHUSETTS DEPARTMENT
OF ENVIRONMENTAL MANAGEMENT.
Figure I.16.16 Instructional signage.
restoration demonstration project described in this chapter. The present project involving
shoreline bioengineering and littoral wetland replanting is an integral part of the larger
watershed plan, particularly addressing the elements A.1.a and B.1.a outlined in the master
plan proposal (France, 1999) shown in Table I.16.4. Other components in the restoration
of Chandler Pond that have been previously investigated include model estimates for the
hydrologic water budget (Fugro, 1996; Element A.3.c in Table I.16.4), design of a wetland
created for the inflow embayment (France et al., 2000; Element A.1.b in Table I.16.4), design
and implementation of a long-term chemical monitoring program (Smith, 1999; Element
A.3.b in Table I.16.4), design of an upstream created wetland and the potential for daylighting the inflow stream from the golf course (France et al., 2000; Elements C.1.b and c
in Table I.16.4), and redesign of the riparian park (Boston Parks and Recreation, 1997, 1999;
Elements B.3.a and b in Table I.16.4). This latter redesign is important, for it reinforces that
the shoreline restoration undertaken for Chandler Pond, similar to that planned for nearby
Fresh Pond (Chapter II.4), is part of a larger effort to help reconnect people with urban
lakes through establishment of water sensitive riparian parks.
Chapter I.16:
337
Table I.16.3 Overall General Goals for Developing a Watershed Plan for Chandler Pond
Goals for Natural Resources
Restore the ponds one-time natural character and water quality
Benefit wildlife in the urban setting
Preserve or replace riparian (shoreline) tree cover
Foster a new, more natural assemblage of native fauna through active stocking and maintenance
Reduce or eliminate invasive, nonnative plants
Goals for Resident and Visitor Uses
Enhance enjoyment and opportunity to observe nature
Foster a sense of place for the community
Limit recreational uses to those that are compatible with natural resource objectives
Entail minimal maintenance
Goals for Education
Promote educational uses of Chandler Pond
Foster stewardship
Contribute to general environmental advocacy
Source: Adapted from Center for Watershed Protection (1998).
Table I.16.4 Specific Elements Needed to Be Addressed in a Master Plan for Chandler Pond
A. Within-Pond
1. Floral habitat
a. Littoral fringe wetland restoration design and planting scheme
b. Inflow embayment wetland creation design, construction plan, and planting scheme
2. Faunal biology
a. Species assemblage desirability study and stocking program implementation
b. Long-term monitoring plan development
3. Physical-chemical
a. Assessment of requisite elements to be surveyed
b. Long-term monitoring plan development
c. Design outline for developing a water budget model for the pond
B. Around-Pond
1. Shoreline vegetation
a. Bankside bioengineering feasibility study, restoration design, and installation plan
b. Tree-bush planting scheme design
c. Nearshore grass maintenance management plan
2. Shoreline hard structures
a. Fishing platform installation design and construction plans
b. Examination of possibilities for stormwater inflow pipes repair and maintenance
3. Park
a. Pedestrian pathway restoration design and construction plans
b. Naturalized park vegetation planting scheme design
C. Upstream and Inflows
1. Point source
a. Long-term monitoring plan development from the golf course to the pond
b. Examination of potential for daylighting buried inflow stream with design and
construction plan
c. Inflow wetland creation design, construction plan and planting scheme
2. Nonpoint sources
a. Development of a stormwater management plan for the pond based on a comprehensive
monitoring plan
338
Table I.16.4 Specific Elements Needed to Be Addressed in a Master Plan for Chandler Pond
D. Surrounding Watershed
1. Land-use planning
a. Determination of impervious cover model
b. Examination of land use management techniques and future development scenario
predictions
2. Land conservation
a. Prioritization of future water sensitive development and protection/infiltration sites
3. Aquatic Buffers
a. Examination of protective buffer strip creation potential, design, and planting schemes
4. Better site design
a. Development of guidance directives for reducing deleterious effects on stormwater
tailored to site specifics in Chandler Pond watershed
5. Erosion and sediment control
a. Development of guidance directives for reducing erosion effects tailored to site specifics
in Chandler Pond watershed
b. Prioritized site exploration assessment of potential for erosion control BMP (best
management practices) retrofits to existing urban framework with detailed budget
6. Stormwater BMPs
a. Development of guidance directives for reducing stormwater inputs tailored to site
specifics in Chandler Pond watershed
b. Prioritized site exploration assessment of potential for stormwater bioretention BMP
retrofits to existing urban framework with detailed budget
7. Nonstormwater discharges
a. Investigation of structural and nonstructural controls for limiting lawn and wastewater
discharges and uncoupling illicit connections with accompanying management plan and
detailed implementation strategy and budget
8. Watershed Stewardship Programs
a. Outline for program of fostering choices for public and private stewardship to sustain
watershed management
Source: Adapted from France (1999).
Acknowledgments
This project could not have been undertaken without the dedicated effort and insightful
wisdom of Timothy Smith of the Boston Parks Department, and the concerned citizenry
who are dedicated to Chandler Pond. This chapter is dedicated to the environmental
vigilance of Dr. Mark Chandler. The authors also thank Tim Smith for providing an update
on the present status of the pond.
Literature cited
American Society of Landscape Artists (ASLA), Integrating stormwater into the urban fabric, Portland State Univ., 1996.
Barnett, D., Dredging and landscaping at Auburn Lake. Mount Auburn Cemetery Newsletter, 1999.
Boston Conservation Commission (BCC), Order of Conditions for the Conventional Dredging of
Chandler Pond, Brighton, MA, 1998.
Boston Parks and Recreation Department (BPRD), Chandler Pond, Brighton Park improvement plan,
1997.
BPRD, Urban Wilds Program, Chandler Pond shoreline restoration pilot project, 1999.
Center for Watershed Protection (CWP), Rapid Watershed Planning Handbook. A Comprehensive Guide
for Managing Urbanizing Watersheds, Ellicott City, MD, 1998.
Chapter I.16:
339
CWP, Better Site Design: A Handbook for Changing Development Rules in Your Community, Ellicott City,
MD, 1999.
EPA, Top 10 watershed lessons learned. United States Environmental Protection Agency Report,
1997.
France, R., Proposal for developing a master plan for Chandler Pond. W.D.N.R.G. Limnetics. Unpubl.
Report., 1999.
France, R. et al., Proposal for environmental and recreational improvement to Chandler Pond
through removal of nutrients from golf course and urban runoff by creation of an inflowforebay treatment wetland complex. Harvard Design School Student Project, 2000.
Fugro Associates, Hydrologic evaluation and preliminary field investigation of Chandler Pond,
Boston, MA, 1996.
Heron, Let the digging begin, Chandler Pond Preserv. Soc. Newsletter, no. 7, 1998.
McLaughlin, B.J., Request to amend order of conditions, DEP File No. 6-790, 2000.
Pouder Design Group, Chandler pond riparian restoration project, 2000.
Schueler, T., Kumble, P., and Heraty, M., A current assessment of urban best management practices,
techniques for reducing non-point source pollution in the coastal zone, U.S. EPA, Washington, DC, 1992.
Smith, T., Chem monitoring plan for Chandler Pond, personal communication, 1999.
Terrene Institute, Urbanization and water quality, a guide to protecting the urban environment,
Tourbier, J.T. and Westmacott, R., Lakes and Ponds, The Urban Land Inst., Washington, DC, 1992.
340
Response
Buying time by bioengineering
Sometimes the efforts of environmentalists can seem frustratingly like the labors of Sisyphus when it comes mitigating, much less getting ahead of, the degradation surrounding
us. One of the most important messages in the EPAs Top 10 Watershed Lessons Learned is
the need for, and reassurance in, small iterative steps forward. In other words, if we wait
to raise all the expertise and funding necessary to completely restore or recover a body
of water, it may be too late to do so in any sensible fashion given the relentless rate at
which degradation proceeds. This chapter by Nicholas Pouder and Robert France demonstrates the utility in undertaking a small-scale bioengineering project in a severely
polluted urban lake.
A major purpose of the shoreline restoration project described in this chapter was to
raise public consciousness about proactive, and largely inexpensive, means that can, and
indeed should, be engaged in order to help sustain the quality of the water in the lake
following the expensive draining and dredging of contaminated bottom sediments. Great
educational power exists in such demonstration projects, and it goes far beyond whatever
physical benefits might ensue to the particular sites upon which they are enacted. Water
sensitive design projects have the greatest likelihood of success with attention directed to
sociological issues such as community outreach. This may be especially true when using
techniques that are a little out of the box in terms of current practice, as, for example,
the various integrated management options involved in low-impact development (Michael
Clar in Chapter I.15) or the use of permeable pavement (Bruce Ferguson in Chapter I.1)
and multifunction treatment wetlands (James Bays in Chapter I.14).
Another important issue that Pouder and France raise concerns the need to scale up
site-specific projects to the watershed scale. Specifically, in terms of bioengineering, this
is something that Wendi Goldsmith conceptually elaborates upon in Chapter I.17 and
Thomas Benjamin practically demonstrates in Chapter II.4. Pouder and Frances listing of
salient elements that need to be addressed in formulating a master plan for the lake in
this study are not specific to only that location; instead, they offer a useful set of guidelines
to aid water sensitive planning and design at any urban location.
chapter I.17
Integrating ecology, geomorphology,

and bioengineering for
watershed-friendly design
Wendi Goldsmith
Abstract
Within watersheds, cumulative impacts from development have caused widespread physical instability, degradation of water quality, and deterioration of habitats. Research findings support that the protection, enhancement, and restoration of many natural landforms
and ecosystem functions can provide the most cost-effective solutions for watershed
management. Although reasonable consensus has been reached regarding measures to
protect existing intact resources, current examples of project work reveal that engineers,
biologists, landscape architects, geologists, and other professionals adopt divergent
approaches and standards for resource enhancement and restoration. This chapter
describes a process of project goal setting (based on ecological assessment), site analysis
(based on geomorphology, hydrology, and pedology), and treatment design (based on
bioengineering principles and rigorous review from a geotechnical and/or hydraulic
engineering perspective as appropriate) that can be applied on a variety of scales to achieve
practical watershed-friendly design. A brief overview of the underlying science and engineering concepts offers a rationale for adopting the procedure. Briefly presented case
studies offer examples of how the interdisciplinary process plays out in various situations.
Although integration with large-scale watershed management problems are addressed,
smaller site-scale design topics are emphasized.
The case for sustainability

Engineers, architects, landscape architects, and other professionals are called upon to plan
and design projects connected to new development or in response to problems stemming
from old development. Standards of professional practice and responsiveness to clients
needs dictate that cost-effectiveness, safety and usability, constructability, and predictability of implementation and performance are the chief guiding principles. Another pervasive
influence is defensive design, which often takes the shape of overdesign as a substitute
for careful analysis and planning, or mechanistic design, which seeks to wrest a site or
structure into one final immutable condition instead of addressing inevitable and sometimes
1-56670-562-2/02/$0.00+$1.50
341
342
desirable processes of change and evolution. For project types ranging from suburban
residential subdivisions to urban infill commercial structures, from transportation corridors to flood control facilities, work products generated by design professionals answer
to influences such as professional liability, scope of work under contract, schedule and
budget, and conventions that are often outdated and, by definition, not forward-looking.
One topic that gets lip service but seldom enters meaningfully into the design or maintenance of projects is sustainability. While the more mundane and practical factors contributing to the design process clearly cannot be neglected, society has an overriding, though
infrequently pursued, interest in ensuring the sustainability of our natural resources,
infrastructure, and quality of life. Designers have an interest in embracing sustainability
as an outcome (e.g., Chapters II.3 and II.11), as it can position them to be leaders in their
fields, proactively identifying and addressing needs beyond the bounds of a site or construction schedule, generally saving money and creating goodwill for their clients.
It has become evident, moreover, that healthy communities are those that recognize
the need for a balance between socioeconomic vitality and stewardship of the environment.
This phenomenon is not recent, because cultures ranging from the Mayans to ancient
Rome rose and fell on the productivity and ultimate demise of natural systems. A metaphor that is occasionally used to illustrate the importance of a balanced approach is the
three-legged stool. The legs represent strong social, economic, and ecological elements.
When each is in place and providing full support, the stool is stable. Failure to maintain
one or more of the elements and the stool falls over. Since World War II, numerous failures
have been made relating to community development in the U.S. Sprawl, for example,
serves few of our needs for social interaction, is wasteful of economic resources, and is
environmentally destructive. Sustainability of this pattern of growth is highly unlikely,
particularly because of its lack of consideration of the value of natural systems and
processes. One blatant example is the water supply for the city of New York, located 100
miles away in the Catskill Mountains, which for nearly a century provided clean water
for millions of city residents. In recent years, the suburban fringe has extended to the
Catskills, where water quality is beginning to decline in proportion to the new shopping
malls and residential subdivisions that are replacing local forest cover. The EPA estimated
the cost of compensating for the change in land use with a water treatment facility at $6
billion (Hu, 2000). Sustainability is cheaper than symptom treatment in a reactive mode
and also is a better cause for the community to rally around and participate in, plus it
creates quality of life that can be equitably shared.
Various helpful efforts have been made to define sustainability, though none have
achieved widespread recognition and adoption in the United States. One system that a
growing number of countries in Europe have adopted is the Natural Step (Robert, 1997).
Karl-Henrik Robert, a Swedish oncologist, developed this method of evaluating human
impact on the environment. He recognizes the finite nature of the system interconnecting
the earths crust and the atmosphere, with only sunlight as an input. All human activities,
he believes, must serve to maintain or manage the equilibrium relationships of the system
in order to achieve sustainability and essential human health. Everything that we produce
or construct must conform to four guidelines that Robert refers to as system conditions: (1)
substances from the earths crust must not systematically increase in nature; (2) substances
produced by society must not systematically increase in nature; (3) the physical basis for
the productivity and diversity of nature must not be systematically deteriorated; and (4)
we must be efficient enough to meet basic human needs. The Natural Step involves taking
one radical step toward reorienting societys priorities to meet the four system conditions.
Roberts seminal paper expounding on the principals of the Natural Step received extensive peer review and wide acclaim among numerous leaders in the scientific community.
Achieving sustainability requires an all-pervasive effort: while many of the aspects outlined
Chapter I.17:
Integrating ecology, geomorphology, and bioengineering
343
in the Natural Step require national policy initiatives, restructuring of markets, and
changes in personal consumption habits, many issues apply to site-scale design choices.
No place or action is separate from the pursuit of sustainability.
The elements of the Natural Step have as a foundation the basic characteristics of a
stable landscape ecosystem, including (1) water is retained and processed via vegetation
and soils, (2) nutrients are recycled, (3) landforms and soils are not eroded or degraded,
and (4) biological populations are diverse and able to reproduce successfully. Without
stability in the form and function of our landscape-scale systems, sustainability cannot be
achieved. Historically, all human land uses from agriculture to urbanization have caused
significant modification and related impacts to the landscape and its ecological functions.
The nature, extent, and timing of the response by the natural system may vary due to a
number of factors. In some cases, the effects may be tolerable over an extended time period
(the decline in air quality since the dawn of industrialization), though in many, they can
be catastrophic (erosion and flooding caused by increased peak flow of stormwater runoff
from paved/roofed areas). In all cases, decline and degradation result, and sooner or later
we recognize the need to manage the system to compensate, mitigate, or repair. In those
cases where the solutions to the perceived problems are based on a comprehensive evaluation of the causes, effects, and relationships of the system as a whole, instead of focusing
on the symptom, sustainability is best achieved. For instance, improving the efficiency of
manufacturing technology and using cleaner energy sources have a more beneficial effect
than refining combustion methods and scrubbing emissions from smokestacks. Similarly,
treatments that allow rainfall to be captured on-site, as opposed to directed on an accelerated basis toward rivers, is more effective than efforts to deepen and widen rivers for
flood conveyance. All design teams would be well served by the inclusion of scientists
and natural resource managers trained and experienced at identifying effective means of
achieving sustainability, instead of allowing the design process to be driven by the prevalent set of influences.
Interdisciplinary design
In most professional design scenarios, architects or engineers are placed in charge of the
project design process. The work they perform closely follows the original scope prepared
by the client, generally represented by a person of fundamentally the same training and
background. Thus, any essential bias or limitation in the original understanding or
approach to performing project work is not likely to be noticed or comfortably broached
by the designer in charge. Refinements to the schedule, exact methods employed, and
even budget are addressed as needed, but the basic premise is seldom reevaluated (which
is most unfortunate when sustainability is not the basic premise). Various scientists are
often engaged to address the matter of obtaining necessary environmental permits. Rarely
are these scientists included with the timing or collaborative effort needed to help guide
the design itself. In fact, most scientists whose experience dwells in preparation of permit
applications and supporting studies do not possess expertise in planning, design, or
resource management that would help them to play a meaningful role in guiding project
design. Whereas most environmental permitting processes are reasonably effective at
obstructing or correcting gravely problematic proposals, they seldom serve much practical
use toward promoting optimal stewardship practices. Often the scientific studies produced
are not reflected or incorporated in any way in the final design, nor are they necessarily
studying the most important issues in terms of guiding the design toward sustainability.
In short, the presence of scientists, engineers, and other designers on a project team does
not in itself signify that the design process incorporates interdisciplinary collaboration.
344
When well executed, interdisciplinary design can be effective at achieving richly

sustainable land-use practices, as demonstrated by a growing number of case studies (see
also Chapter I.7). True interdisciplinary design is characterized by a creative, inclusive,
objective, and iterative process. The nature and scale of the process can be adapted based
on the size and schedule of the project. The cost of interdisciplinary design is frequently
higher than its conventional counterpart due to the inclusion of suitable team members
of varied disciplines, and a review and coordination process that allows iterative evaluation, incorporation, and refinement of appropriate design elements; however, this
approach need not be considered an expendable luxury that is only suited for special
projects with generous schedules and budgets. Time and time again, interdisciplinary
design produces efficiencies in meeting permitting and implementation schedules,
improvements in construction and maintenance costs, reductions in off-site impacts, and
enhancements in aesthetics and perceived quality in general. After having completed a
few successful projects, no other approach will be acceptable; as testimony are the numerous design firms, municipalities, and even entire countries that have wholeheartedly
embraced interdisciplinary design in support of sustainability.
Though there is not one single recipe for success at interdisciplinary design, several
features can aid in the formulation of such an approach:
Tailors assessment methods and design approach to site conditions

Responds to community participation
Emphasizes collaboration and facilitation rather than hierarchy and direction
Welcomes brainstorming and out-of-the-box ideas
Considers factors beyond the construction phase
Evaluates system-wide issues including past and future potential conditions
Views problems within their broad environmental, social, and economic context
The three-step process

Once the interdisciplinary approach has been accepted for use on a project and appropriate
team members have been included, a procedure must be adopted. The procedure we favor
at The Bioengineering Group, Inc. (TBG), is a three-step process: identify ecological constraints and opportunities for every project; base design on stable landforms, healthy soils,
and balanced hydrology; ensure excellence in design and execution through sound engineering and landscape architectural practice, emphasizing green solutions.
The following outline is intended to provide a framework for demystifying the science
and the art of applying interdisciplinary design. The entire method is an evolving tool
drawing on well-documented scientific principles applied in complex, multifunctional
applications. Depending on the scale, the context, and the intent, this approach can be
followed for achieving excellent results with some fairly common design scenarios, but it
has also been useful when applied in a highly experimental mode. Each point is worthy
of its own in-depth article, and many exist in the literature, though few focus on interdisciplinary restoration applications (Barrett, 1997; Larson and Goldsmith, 1997; Goldsmith,
1998; U.S. Army Corps of Engineers, 1995).
Setting project goals and objectives based on ecological assessment

Very often, ecology and biology are seen as sciences that pertain to rural, wild, or generally
nonurban areas, and little attention is paid to these subjects in the context of urban settings.
Also, they are seen as extraneous themes that can be addressed only when time, money,
and vocal supporters are abundant. Little effort has been made to recognize the practical
Chapter I.17:
345
values connected with protecting, maintaining, and improving the conditions and functions of our flora and fauna. Landowners have lacked both motivation and guiding
principles for identifying and prioritizing sensible measures and have failed to integrate
them into common development activities at the site scale. For the architect, engineer, or
builder, the challenge of interdisciplinary design is to address the big picture view of a
clients problems and their potential solutions, even when the client has not specifically
asked for them. In practice, the evaluation of the biological and ecological components is
generally broken down into the following topics.
Site potential
Prioritize restoring to past conditions, either partially or fully.
Consider mitigating upstream or regional problems.
Evaluate how site design could contribute partially to a potential future condition
(e.g., creating salmon spawning habitat even though fish passage problems will
not be solved by this project).
Limiting factors
The usual suspects include water quality impacts, lack of suitable in-stream conditions, overcompacted soils, invasive species, etc.
The list should include those factors that are intractable as well as those that can
be managed or corrected.
Account for both natural and anthropogenic factors.
Food web
Recognize linkage between terrestrial and aquatic habitat.

Provide for suitable primary production (plants).
Address needs of consumers, both invertebrates and vertebrates.
Address needs of predators, both invertebrates and vertebrates.
Establish realistic plans for the needed range of species, not only charismatic megafauna or game species.
Physical complexity
Generate microtopography and suitably varied soil conditions.
Favor natural drainage patterns including complex planform and section at all
scales.
Enhance edge habitat through integration of complex shapes at all scales, strive to
create numerous and varied nooks and crannies as physical niches.
Biodiversity
Plan a diverse seeding, planting, and maintenance approach.
Do not envision a stocking program, but plan to have numerous critters tend to
their own needs if the physical/chemical basics are there.
Focus on early successional plant species in most cases.
Aim for the widest number of species possible, instead of targeting just a few.
346
Biogeochemical cycles
Address the interrelationships between plants, soils, and water to provide appropriate redox conditions.
Establish the potential for robust N, P, and K cycles.
Use the landscape for managing soil contamination and/or water quality impacts
whenever possible.
Guiding design alternatives based on sound application of earth science

Every site will have its own unique land-use history, and each region or landform will
possess characteristic conditions. However, one thing remains constant: landscapes have
evolved their features in response to climate over time to achieve a state of dynamic
equilibrium and/or slow, controlled change. Almost equally constant is the fact that
forestry, agriculture, and urbanization practices have altered and destabilized these powerful yet sensitive equilibrium relationships. Failing to recognize, respect, and/or harness
landforming processes will inevitably lead to project design that does not achieve sustainability. The chief repercussions include causing offsite impacts, requiring repair or reconstruction, and demanding high ongoing maintenance efforts. Careful selection of site
design elements can reduce initial construction costs in some, though not all, situations,
but this approach consistently offers improved long-term costs. A broadly useful list of
considerations follows:
Watershed hydrology
Use equilibrium hydrology (i.e., undeveloped, precolonial conditions) as a benchmark for achieving sustainable function.
Balance the water budget in terms of evapotranspiration, infiltration, and runoff.
Recognize that analysis of existing versus post-construction conditions will not
yield a sustainable outcome, because the site is likely already degraded.
This method allows watershed-scale problems to be addressed remarkably well at
the site scale and on a piecemeal basis.
Stream pattern and process

Perform at least a basic fluvial geomorphic assessment in order to ascertain the
suitability of the existing conditions.
Endeavor to protect, enhance, or radically restore stable stream cross-section, profile, and planform geometry whenever possible.
Respect stream form and function as it pertains to low flow, or typical daily
conditions, at least as much as for less frequent recurrence intervals.
Do not inadvertently or purposefully introduce measures that cause disturbance
of stream pattern and process.
Do not assume that all stream/river channels should be treated as sacred spaces
not to be disturbed. Many have been trashed and can be much improved.
Apply principles of fluvial geomorphology to guide the design of suitable channel
restoration measures.
Use qualitative methods to develop and/or refine restoration schemes, but never
omit quantitative analysis for perennial streams.
Chapter I.17:
347
Sediment transport
Recognize that all streams and rivers serve to carry sediment, not just water,
generated within their watersheds.
Carry out design and construction in recognition that in-stream conditions are not
fixed or rigid and will be reworked by sediment transport processes.
Identify sources of sediment and characterize sediment yield of the basin.
Perform, at minimum, a rough characterization of bed load, suspended load, and
wash load for all perennial streams.
Analyze critical values for sediment transport hydraulics for larger streams.
Flow conveyance
Seek to promote appropriate base flow conditions.

Recognize and/or manage channel-forming discharges.
Be prepared to address regulatory needs to prevent adverse flood impacts.
Explore nontraditional measures to address flow conveyance such as increasing
channel roughness to retard downstream flooding.
Consider how channel modification or instability may have exacerbated flooding
problems (such as downcutting reducing access to floodplain storage volume and
translocating problems downstream).
Role of soils and vegetation

Assess the contribution of existing soils and/or vegetative cover or reinforcing
roots on physical stability of slopes and streambanks.
Recognize that runoff and erosion are highly sensitive to variation in soil character
and plant cover.
Assess the actual or proposed site-specific soil and plant cover conditions in maintaining infiltration and/or evapotranspiration.
Understand that soils and structures will require maintenance for ongoing infiltration functions unless vegetation promotes permeability through perturbation
by roots and burrowing organisms.
Seek to preserve or reestablish healthy soil conditions and self-maintaining plant
cover.
Formulating and implementing elements of planning

and design for sustainability
The most elegant design solutions are successful at achieving multifunctionality with
efficiency and value. Not unlike poetry, they condense multiple meanings and great
significance into compact formats. In pursuing sustainability, planning must be applied
to protect existing resources; engineering, to validate and ensure performance; and landscape architecture, to deliver functionality rather than decoration. Having set project goals
and objectives based on sustainable ecological function, and steered the selection and
development of features and treatments based on earth science, the design process will
likely be aimed in a wise direction. The actual design process may be either simple and
abbreviated for smaller projects, or quite elaborate with many iterative steps and repeated
348
review and coordination by various disciplines. We have found that the following guidelines apply to a wide array of situations and designers styles.
Bioengineering
Incorporate vegetation into site design for its functional values.
Apply woody and herbaceous vegetation for streambank protection and slope
stabilization.
Apply vegetation for management of contaminated soil and water via treatment
wetlands, phytoremediation, and riparian buffer enhancements.
View bioengineering as an extension of, and a complement to, conventional engineering.
Combine bioengineering with conventional engineering where appropriate.
Geotechnical engineering
Rely on landform analysis to drive site features and use bioengineering first, then
validate or refine design based on geotechnical analysis.
Use conventional measures such as soil benching to address soil stability below
the zone of influence of vegetation.
Avoid over-compaction of soils, relying on assessment of root contributions to soil
shear resistance instead.
Incorporate geofabrics to complement, not hinder plant and soil health.
Hydraulic engineering
Develop channel form and bank treatment characteristics first, and validate or
refine using quantitative hydraulic analysis.
Consider hydraulics at current, built-out, and managed conditions (such as with
feasible upstream detention features).
Adapt the hydraulics to allow geomorphically sound, vegetatively rich treatments,
rather than ruling out vegetation if the first analysis does not validate it.
Use upstream/downstream or reference reach channel characteristics and roughness to calibrate any modeling.
Civil/environmental engineering
Incorporate a wide and uncommon array of stormwater best management practices
in all site designs toward achieving sustainable hydrology.
Avoid the use of conventional catch-basin-and-culvert systems whenever possible.
Do not assume that end-of-pipe water treatments can adequately deliver suitable
management of water quantity and quality.
Consider the use of phytoremediation for the cleanup or containment of soil and
groundwater contamination.
Landscape architecture
Select native species, including grass seed or sod, with rare and purposeful exceptions.
Incorporate plant communities, instead of isolated species.
Chapter I.17:
349
Address self-propagation instead of planned obsolescence.

Stockpile, store with care, and reapply topsoil to conserve nutrients, organic matter,
and propagules.
Minimize reliance on mowing and mulching, favoring rougher and looser styles
of plant cover that are compatible with self-propagating plant communities.
Accommodate human access to open space and water elements, even if they did
not receive attention before restorative work.
Regional/urban planning
Prevent sprawl.
Reuse brownfields.
Protect and manage greenway corridors connected with sensitive and functional
resource areas.
Demand the greenest of green techniques in the most densely developed settings
instead of considering them too far gone.
Sound construction practices

Ensure that public safety is addressed in all designs, but refer to natural conditions
as an ideal standard for site design.
Ensure that access and staging and realistic scheduling are possible for any proposed sitework.
Specify if access shall be conducted in a sensitive manner, such as by barge or
using low ground-pressure equipment, or even only pedestrians and hand tools.
Accommodate special planting forms or quantities of unusual species by arranging
to contract grow in advance.
Assume that the contractor will be unfamiliar even reluctant to execute the
design as intended.
Provide adequate quality assurance procedures including construction oversight.
Anticipate use of common materials and equipment, even if you specify uncommon applications.
Account for required maintenance during the establishment period, and beyond,
through contractor clauses or separate arrangements.
This three-step design approach has been used on more than 100 projects conducted
by TBG, ranging from residential-scale projects to river restoration efforts involving
20 miles of heavily used large channels. Three sample case studies are briefly presented
to outline the primary ecological goals, the basic earth science principles applied, and the
particular engineering approach or construction measures selected.
Merrick Brook, Scotland, CT

Project highlights
Streambank stabilization using bioengineering features

Application of stable channel geometry principles
Vernal pool creation
Enhancement of habitat for brook trout and brown trout
350
Figure I.17.1 Merrick Brook, Scotland, CT. Newly realigned channel with bank stabilization using
root wads and native plantings, plus in-stream enhancements using boulder weirs.
Merrick Brook is part of the Talbot Wildlife Management Area and provides an
important fisheries habitat, which includes reproducing populations of native brook trout
Salvelinus fontinalis, introduced brown trout Salmo trutta, and other aquatic and terrestrial
wildlife. The persistent erosion along one extreme channel bend caused degradation of
the fisheries habitat directly at and immediately downstream of the bend. The resulting
downstream sediment loading adversely impacted the benthic aquatic organisms. The CT
DEP retained TBG to design a streambank stabilization project using bioengineering
techniques and fluvial geomorphic assessment principles to protect and enhance the
aquatic habitat.
Restoration design
The restoration design (Figure I.17.1) established a channel with a more suitable radius of
curvature for the given channel bankfull width and discharge than what presently existed,
thus reducing concentrated energy losses against the eroding meander bank. The design
curve radius is consistent with measurements of stable meander bends upstream and
downstream of the project area.
A riffle pool sequence was also established throughout the constructed channel to
create the suitable energy gradient and efficient flow and sediment transport conveyance.
Weir structures (cross weir vanes) were used to provide a stable transition at the head
and tail of the newly constructed meander bend, as well as long-term grade control and
energy dissipation. The weirs provide additional aeration of low flows to improve water
quality, and an excavated area downstream of the weir provides scour pools for aquatic
organisms in the transition to the riffle.
The meander bend was stabilized with root wads, live fascines, and transplants of
native shrubs and sods containing rooted brush species to provide immediate erosion
protection and long-term vegetative component that can shade the water and lessen water
temperature rises during the summer months.
Chapter I.17:
351
Hearthstone Quarry Brook, Chicopee, MA

Project highlights
Integration of bioengineering and conventional engineering structures

Stable channel geometry principles
Detailed pre- and post-project monitoring
Riparian function restoration
Urban stream resource enhancement
The Hearthstone Quarry Brook (HQB) runs through a dense residential area and
receives stormwater runoff from area roads, highways, and houses. With a planned road
expansion that would route even more stormwater flow to the brook, the city had to
address the problems of severe channel incision, erosion, and loss of wetland resources.
With grant-writing assistance from TBG, the city received an EPA s. 319 grant to fund a
project that would demonstrate the utilization of bioengineering techniques for channel
stabilization, nonpoint source pollution abatement, and habitat enhancement.
Site assessment and bioengineering design

TBG conducted a channel assessment to measure critical morphologic parameters and
determine the suitable design channel dimension and channel material sizes. Staff calculated shear stress conditions using field data and used modeling to help determine the
location of bioengineering applications and suitable materials. Due to the high degree of
shade and steep banks at the site, vegetation selection was critical. A historical analysis
of channel conditions was conducted as well, to assess rates of degradation, evaluate
potential resource functions, and refine design objectives.
Monitoring and construction oversight

TBG oversaw installation of the bioengineering applications, planting, and placement of
coarse woody debris. In addition, TBG staff provided technical assistance in accessing the
site and designing dewatering techniques during construction. Due to the poor access,
the confined incised valley, adjacent residential property, and saturated clay soils TBG had
to closely supervise construction and make on-site modifications to the engineering
designs. TBG staff designed several natural log dam structures to provide grade control
and aquatic habitat and integrated vegetation into several gabion structures to increase
riparian functions (Figure I.17.2). In order to evaluate the unique combination of techniques used at the site, TBG designed a monitoring scheme to assess pre- and postconstruction channel stability, habitat values, and vegetation performance.
Salem Salt Marsh, Salem, MA

Project highlights
Design/build responsibility complex for degraded saltmarsh

Phytostabilization of residually contaminated organic soils
Brownfields design and remediation
Reduction of invasive species (Phragmites australis)
Enhanced tidal flushing process and created stable landforms
352
Figure I.17.2 Hearthstone Quarry Brook, Chicopee, MA. After several years of development, a dense
cover of riparian vegetation borders a stable channel supporting appropriate organisms.
Water quality and habitat improvement

Background
The historic stockpiling of industrial waste within the saltmarsh resulted in serious degradation to the otherwise healthy tidal system. At the time of property transfer, the former
owner was required to perform a cleanup of the marsh and restore it to its original
condition. Following the removal of the waste, TBG was selected to restore the saltmarsh.
Objectives and constraints

The projects objectives and methodology were tightly framed to restore the sites original
saltmarsh vegetation and to reduce the vitality of a number of dense pockets of Common
Reed (Phragmites australis). Residual levels of lead and other metals necessitated an aggressive and effective stabilization strategy to address public concern. Phytostabilization measures using native saltmarsh plant species allowed engineering controls and habitat
improvements to be achieved simultaneously. The main responsibilities of TBG were (1)
to survey the site to determine appropriate elevations and locations for each of the specified saltmarsh species, (2) to develop a planting plan based on the survey, (3) to direct
and supervise regrading activities intended to improve tidal circulation throughout the
restoration zone, (4) to procure and propagate plant stock within the narrow geographic
parameters specified, and (5) to coordinate and perform plant installation and phragmites
reduction activities. In all, nearly 70,000 individual saltmarsh plants were planted. Despite
the practical matter of being able to proceed only during low-mid-tide cycles, this 2+acre
site was planted on schedule and in compliance with project specifications.
Chapter I.17:
353
Figure I.17.3 Salem Salt Marsh, Salem, MA. Even the most degraded and impacted urban sites can
be restored using sound design and implementation strategies to achieve beauty and function.
Results
Monitoring visits performed by TBG in 1997 showed a near-100% success rate in growth
for the saltmarsh species. Phragmites reduction and phytostabilization of contaminated
soils has also been found successful. The beautiful landscape now serves as an educational
resource for local college and high school students (Figure I.17.3).
Literature cited
Barrett, K.R., Introduction to ecological engineering for water resources: The benefits of collaborating
with nature, presented at the Ann. Conf. New England Water Environ. Assoc, 1997.
Goldsmith, W., CPESC, Soil reinforcement by river plants: Progress results, Proc. ASCE Conf. Wetlands
Eng. River Restoration, 1998.
Hu, W., U.S. says New York City may have to spend $6 billion on filtration, New York Times, June
1, 2000.
Larson, M. and Goldsmith, W., Incised channel stabilization and enhancement integrating geomorphology and bioengineering, Proc. Conf. Manage. Landscapes Disturbed by Channel Incision,
Wang, S.S.Y., Ed., 1997.
Robert, K.H., The Natural Step: A Framework for Achieving Sustainability in Our Organizations, Pegasus
Communications, Cambridge, MA, 1997.
U.S. Army Corps of Engineers (CECW-E/CECW-O), Environmental engineering initiatives for water
management, Civil Works Eng. Tech. Lett., TL 1110-2-362, http://www.usace.army.mil/inet/usace-docs/eng-tech-ltrs/etl1110-2-362/entire.pdf, 1995.
354
Response
Sustainability through interdisciplinarity
This chapter by Wendi Goldsmith provides an extremely useful mix of hopefully quixotic
aspirations with grounded reality therapy. Goldsmiths likening of the multifunctionality
of the best water sensitive design projects to the multiple meanings residing in the best
of poetry is perceptive. Her clarion call for interdiscipline plurality in the effective design
and lasting existence of projects is especially apt if those projects occur at high-visibility
sites such as those described by Robert France and Philip Craul in Chapter I.7, Diana
Balmori in Chapter I.8, Catherine Berris in Chapter I.9, and Glenn Allen in Chapter I.13.
A strong point made in this chapter is the need to include scientists in the actual design
process, not merely having them only participating in the environmental permitting. This
is a topic demonstrated in the work described by France and Craul in Chapter I.7. As
Goldsmith notes, True interdisciplinary design is characterized by a creative, inclusive,
objective, and iterative process. Given the rapidly rising acumen and coincident expectations of the public for multiple functions, Goldsmith is correct in stating that no other
approach will be acceptable. The three-step process described in this chapter to achieve
this end should prove valuable to many undertaking water sensitive designs and planning.
The daunting challenge to accomplishing all this is correctly identified by Goldsmith:
educating clients to think in the long term and to understand that initial project development costs may be higher in the short term. Nicholas Pouder and Robert France experienced this challenge with their own project described in Chapter I.16, for example.
By moving from what Goldsmith calls defensive design toward sustainable designs,
in the spirit of Daniel Williams objectives for watershed management described in
Chapter II.11, an important lesson becomes clear: the need to consider factors beyond the
construction phase beyond in both space and time. The emphasis for sustainable
design in a watershed-friendly format is to move from running around and stomping
out individual fires to considering what is starting those fires in the first place. Or, to use
an expression popular in the environmental movement, to move from rearranging the
deck chairs on the Titanic for a better view of our collective demise, to grabbing the wheel
and steering us away from the danger all together.
Constraints, challenges, and

opportunities in implementing
innovative stormwater
management techniques
Participants: Richard Pinkham, Larry Coffman, Michael Clar, Thomas
Liptan, Herbert Dreiseitl, Wendi Goldsmith, Dennis Haag, Desheng Wang,
Catherine Berris, Bill Wenk, Alex Felson, and Robert France
Past
Desire for and love of tightly regulated, top-down control of an element water
with a physical nature that resists such an approach
Rapid drainage paradigm; i.e., sending water down the pipes and not taking responsibility
Imbalance between understanding of hydraulic engineering sciences (e.g., how to
design a trapezoidal channel) and hydrologic engineering sciences (e.g., what the
proper design storm is and how it may affect the entire watershed), leaving us
with fragmented, partial approaches that do not really solve the problem
Regulatory control restricted to a few professional groups with a narrow, institutionally enforced knowledge base
Flooding and water pollution are accepted standards of urbanization
Perspective of stormwater issues as a relocative problem that someone else at another
location will deal with later
Undervaluing waters cultural importance
Even if we knew today what ecological design was, we could not do it; i.e., often
the greatest limitations are due to institutional barriers and antiquated development codes that codify the status quo
Present
Acceptance of limits to urban growth, water resources, and human ingenuity with
respect to solving environmental problems
Early attempts at multiple use of water as it moves down the pipe
Recognition that alterations in stormwater management are multigenerational and
that solutions need to be self-sustaining through time
Understanding of the roles of natural, predevelopment processes, and how stormwater behaves there
Gradual evolution of institutional changes to circumvent previously codified and
unimaginative regulations
Narrowing the gaps between clients, regulators, and general society
1-56670-562-2/02/$0.00+$1.50
355
356

Intermingling of academic knowledge with real-world experience to facilitate and
speed up acceptance of innovative solutions
Beginning awareness of appropriate technologies with associated attempts at implementing innovative design solutions
Fusion of technologies and legislative actions for wetland protection, groundwater
recharge, and public recreation
Acknowledgment of the importance of cumulative effects
Future
Search for means to take new technologies and redesign them to deal with ecological issues
An approach for assessing performance that is more like medicine in terms of
requirements for quantifiable data, threshold criteria, and checks of implemented
solutions
Further education of the comprehensive role water plays in the functional infrastructure of cities
More knowledge needed about the hydrologic importance of groundwater recharge
in a watershed context
Implementation of a true hydrologically functional approach supported by a corpus of systematic and vigorous scientific investigation and engineering analysis
Desire for many more high-visibility demonstration projects ably demonstrating
creativity and revealing processes; i.e., need to try and prove techniques
Requirement for shared leadership of new, fresh voices along with a corresponding
movement toward embracing a plurality of visions in innovative stormwater management; i.e., adoption of the tradeoff leadership strategy exemplified by geese in
flight
Greater inclusion of ethics in professional practice; e.g., movement away from large
engineering, end-of-pipe solutions that generate large fees, toward small, dispersed, and less lucrative micromanaged solutions
Realization of the limitations of technological solutions alone
Moving from single-purpose

treatment wetlands toward
multifunction designed wetland parks
Participants: James Bays, Jake Cormier, Nicholas Pouder, Bryan Bear,
and Robert France
Past
Single-purpose functionality rules in wetland design
Shortage of awareness and absence of guidelines in how to bring people into
contact with wetlands, especially by those who fund projects
Fear and ignorance about how some treatment wetlands, especially those designed
for sanitary effluents, can be made multifunctional
Erroneous belief that this is just some new, untested, techno-fix temporary solution;
i.e., dont innovate with me!
Poor inter-/transdisciplinary communication among professionals designing and
constructing wetlands; e.g., between landscape architects and civil engineers
Shortage of empirical data on how designed wetland sites are used by the public
Absence of community involvement
Arrogance of some landscape architects in thinking that they can and should design
everything
Prevalence of the two extremes; i.e., barren and ugly, square-box wetlands and
overgardened, exotic-planted wetlands
Lack of long-term maintenance
Overwhelming majority of wetland creation and restoration projects are undertaken because they have to be; i.e., reactive instead of being proactive
Unrealistic belief in mitigation as a tried and proven strategy for replacement of
natural wetlands lost during development
Present
Recent perception of wetlands designed to achieve multiple objectives; i.e., treatment wetlands are now recognized to have an ancillary role of attracting wildlife,
and with them, humans
Development of wetlands for environmental education
Establishment of a good corpus of physical design guidelines for wetland creation;
e.g., importance and influence of size on biodiversity
Most are public projects located in the suburbs
Start of recognition of a wetland premium in terms of real estate values for proximal
property owners
1-56670-562-2/02/$0.00+$1.50
357
358

Involvement of landscape architects in the presentation of design plans to the
public as an accepted and integral step in raising public interest in wetland creation
projects
Gradual realization that nature is the final and best architect
Wetlands created in suburban locations are becoming appreciated for the feelings
of natural experience they engender; i.e., realization of Olmsteds grand vision of
extracting nature to make it convenient
Future
Multidisciplinary team right from the start; i.e., all created wetland parks that have
public-use applications must involve a landscape architect, a hydraulic engineer,
a wetland designer, a site civil engineer, and a watershed process engineer or
hydrologist, all sharing a strong belief in osmosis learning
Need to explore new and varied ways of allowing people to become enriched by
their wetland experiences while at the same time not disturbing wildlife
Opportunities for multiple uses need to be brought into planning and design
discussions much earlier in the process, and both the public and public works
officials need to be included
Turn multifunctionality on its head; i.e., ancillary functions such as enhancing
biodiversity need to be made premier with other uses, such as stormwater detention, pitched as by-the-way extra benefits
Need for much more documentation about the challenges, successes, and lessons
learned from established wetland creation projects
Cardinal requirement for more rigorous post-construction monitoring in order to
persuade others that these projects have value
More attention needs to be paid to safety concerns of wetlands as part of public
works projects; i.e., the liability issue in the U. S. compared with such places as
Canada and Europe may be the single greatest limitation to the design of aesthetically beautiful wetland parks that encourage unfettered visitor access
Avoidance of hubris; i.e., we should not lose sight that what we are really designing
(at least in the short term) is an interpretation or surrogate of nature and not
necessarily real nature (though well-designed projects may evolve toward such
over time)
Honest acknowledgment of the ethical and practical weaknesses and limitations
involved in wetland mitigation
Further education of clients toward moving from site-specific design issues to
larger regional planning objectives; i.e., water does end in the wetland
Embrace the concepts of, and find ways to further promote, diffusion innovation
in terms of education
Always remember that no matter who is paying the bills, the ultimate client is the
place the landscape
part II
Water sensitive planning

Overview: New interpretations in the management of watersheds
and riparian buffers and corridors
Part II comprises 17 chapters and responses that deal with two of the most important
issues associated with water sensitive planning: management of watersheds and riparian
buffer-corridors. Together, these chapters go far toward advancing the concept that, in
order to be sustainable and therefore effective, water sensitive planning has to be concerned just as much about managing human interactions as about simply managing water
resources alone. In other words, as the physical scale of projects increases, so too must
the sociological scale.
Four chapters focus on the management of riparian buffers and corridors. Frank
Mitchell provides an introduction to the complex issues and variable management options
for protecting water quality and preserving shoreline biodiversity. The chapter by landscape architect Charles Flink expands this vision by addressing the important role that
such regions play in supporting human recreation and societal infrastructure. Two of the
chapters those by James MacBroom and by Leslie Zucker, Anne Weekes, Mark Vian,
and Jay Dorsey integrate these objectives and discuss how rivers should be managed
and restored by viewing them as landscape systems rather than as independent aquatic
entities.
Five chapters focus on implementation of watershed management planning in urban
settings. Four chapters those by landscape architect Thomas Benjamin, by landscape
architect David Blau, by Dennis Haag, Stephen Hurst and Bryan Bear, and by Kelly Cave
demonstrate the various concepts that need to be addressed in order to protect urban
water resources in the face of existing degradation and future development, both often
taking place in an environment of complex sociopolitical interactions. The chapter by Gail
Krantzberg and Judi Barnes grapples with the intriguing and important issue of measuring
the recovery of impaired urban waters.
Eight chapters address water sensitive land-use planning in rural areas. Three of
these by landscape architect Margot Cantwell, by Robert France, and by Neil Hutchinson deal with protecting the integrity of boreal lakes from watershed development,
riparian forestry, and recreational use. In his chapter, landscape architect Daniel Williams
360
expands the physical sphere of water sensitive planning, arguing that in order for water
management to be truly sustainable in certain situations, it should be approached on the
scale of entire regions instead of specific watersheds. In a similar spirit, landscape architect
Amir Mueller, Robert France and landscape architect Carl Steinitz expand water sensitive
planning temporally in their presentation of a methodology based on alternate futures
modeling. Jeffrey Schlosss chapter introduces the benefits and pitfalls ensuing from use
of geographic information systems in water sensitive planning. The remaining two
chapters those by John Felkner and Michael Binford and by Robert France, John Felkner,
Michael Flaxman, and Robert Rempel provide examples of how GIS analysis can aid
decision making about both overt and subtle land-use management concerns.
chapter II.1
Shoreland buffers:
Protecting water quality and biological
diversity (New Hampshire)
Frank Mitchell
Abstract
Most interest in shoreland buffers has been related to water quality protection. Protecting
biological diversity, however, requires much larger buffers. This chapter presents several
concepts that apply to shoreland buffers whether water quality or wildlife habitat is the
primary concern:
Buffers are most effective when they are part of an overall protection strategy that
includes whole watersheds or habitat areas.
Connectivity is crucial. A continuous buffer along a stream, for example, will have
more integrity and effectiveness than a fragmented one.
Core protected areas will add a level of protection beyond what routine buffers
can provide.
Even a fragmented buffer provides some benefits regarding water quality and wildlife. Some gaps can be filled over time through restoration efforts.
Citizens appear to support the water resources and other features that shoreland
buffers can protect. They also appear to support actions needed to conserve ecological and social functions and values of buffers.
Shoreland buffers can be established through voluntary or regulatory means.
Buffers can be designed and established to adequately protect both water quality
and wildlife habitats.
Introduction
Buffer is used with different meanings in different contexts. For the purposes of this
discussion, a shoreland buffer is defined as a naturally vegetated upland area adjacent to
a surface water or wetland. This often means uncut or undisturbed forest, minimally
disturbed or managed forest, and abandoned pasture or fields reverting to forest
(Figure II.1.1). This is different from a setback, in which there is a distance requirement
between surface waters or wetlands and activity that might affect them but not a specification of the type of land cover in the setback zone. A filter strip usually refers to an area
1-56670-562-2/02/$0.00+$1.50
361
362
Figure II.1.1 Profile of stream buffer. The shoreland buffer in this profile consists of the forested
areas on either side of the stream, with farmland to the left and residential land to the right. The
natural roughness of the land supports retention and filtering of surface runoff and groundwater
infiltration. Tree roots help hold soil in place, and debris the trees produce nourishes stream ecosystems. The trees can also sequester nutrients flowing across the buffer aboveground and in
groundwater. (Drawing by Linda Isaacson.)
of herbaceous vegetation between agricultural or similar land uses and surface waters or
wetlands. The term riparian buffer is sometimes used to describe streamside, but not
necessarily wetland or lakeshore areas, and some who use the term apply it primarily to
wetlands adjacent to a stream or river, or to the immediate shoreline. Therefore, the term
shoreland buffer, as used here, differs from these other terms in that it implies:
The topography of the land adjacent to water or wetlands is natural

An expectation of natural vegetation
Inclusion of buffers for wetlands as well as surface waters
Inclusion of upland areas beyond a stream bank, lake shore, or wetland edge
Application to a variety of possible land uses beyond the buffer
A minimum but flexible width based on site-specific conditions
Some current views of buffers

Most historical interest in shoreland buffers has been related to water quality protection
and to some degree aesthetic concerns, but more recently the wildlife habitat functions of
shoreland areas have received increasing attention. If landscape design that incorporates
shoreland buffers is to be ecological, it must consider these habitat functions and values
as well as water quality and visual concerns. Chapter II.5 presents considerations for
meeting objectives of multiple-function stream corridor planning. Chapter II.3 also discusses how water-related greenways can serve to meet water quality, hydrologic, transportation, recreational, and other needs.
Research on buffer effectiveness has been the basis for recommendations about buffer
width and other aspects of design and maintenance.17 For example, in New Hampshire,
Chapter II.1:
Shoreland buffers: Protecting water quality and biological diversity
363
Table II.1.1 Wildlife within a 100-ft Shoreland Buffer: Habitat Needs of Representative Species
Species
Habitat Needs
Eastern (red-spotted) newt

Notophthalmus v. viridescens
Northern 2-lined salamander
Eurycea b. bislineata
Green frog
Rana clamitans melanota
Wood frog
Rana sylvatica
Spotted turtle
Clemmys guttata
Wood turtle
Clemmys insculpta
Woodland habitat for terrestrial juveniles (efts) for the 27 years

they spend on land
Foraging area adults may wander 330 ft on rainy nights;
dispersal of juveniles (only 25% return to natal streams)
Dispersal habitat
Northern water snake

Nerodia s. sipedon
Bats
Myotis and other spp.
Beaver
Castor canadensis
Mink
Mustela vison
Black bear
Ursus americanus
Bald eagle
Haliaeetus leucocephalus
Red-shouldered hawk
Buteo lineatus
Area-sensitive forest birds
Habitat for most terrestrial activity, often well away from water
Habitat for most terrestrial activity will travel up to 2600 ft from
water to find temporary food sources
Habitat for most activities; spend most of their time within 1,000 ft
of water, but will travel up to 1 mile to search for food; will nest
up to 330 ft away
Habitat for dispersal and hibernation
Roosting sites prefer to roost within 1,300 ft of water
Enough foraging habitat most foraging within 330 ft of water;
dispersal routes
Hunt up to 600 ft from water; den sites may be up to 330 ft from
water
Den sites; enough area for travel adult males require up to 19
square miles depending on habitat and food sources
Nest sites most eagle nests are within 1,300 ft of shorelines;
protection from human disturbance
Nest sites this species is found only where buffers are 330 ft or
more
Sufficient breeding habitat for species that need buffers wider than
330 ft
Source: Adapted from Chase, V.P., Deming, L.S., and Lataviec, F., Buffers for Wetlands and Surface Waters: A
Guidebook for New Hampshire Municipalities, 2nd ed., Audubon Society of NH, Concord, NH, 1997.
an interagency project that produced the publication Buffers for Wetlands and Surface
Waters, A Guidebook for New Hampshire Municipalities, used research results to recommend
a reasonable minimum 100-ft, naturally vegetated upland buffer around streams, rivers,
ponds, lakes, wetlands and estuaries. The basis for the recommendation was the recognition that a 100-ft-wide buffer can be expected to typically remove at least 60% of pollutants
such as eroded soil and phosphorus.1
Protecting biological diversity, however, requires much larger buffers. A 100-ft buffer
will provide food sources, cover, and other habitat components for many species, particularly smaller ones, but would only partially provide these needs for others. Buffers
recommended by wildlife biologists can extend to over 1,000 ft from a water body or
wetland.1,8 The rationale for buffers of this size is based on the travel and dispersal needs
of certain species, especially those with life cycles that utilize both aquatic and upland
habitats, and protection from nest predation and parasitism for animals such as interior
forest nesting birds.9 Table II.1.1 presents examples of the habitat needs of representative
species that use shoreland buffer areas.
Shoreland buffers alone will not fully protect water quality or wildlife. In both cases,
buffers will be most effective if they are part of a larger overall protection strategy that
includes whole watersheds or habitat areas.
364
Figure II.1.2 Woody debris in a stream. Leaves, branches, and other debris falling into streams from
the surrounding forest provide important habitat structure and food for stream fauna. In smaller
streams, such detritus can account for the majority of food energy in the system. (Photo by Frank
Mitchell.)
What buffers do (buffer functions)

Buffer functions and their benefits are ecological and social. They include:
Hydrologic effects such as promoting groundwater recharge, moderating flooding
by intercepting rain and snow melt, reducing sunlight penetration (thereby slowing
the rate of spring snow melt), and storing flood waters that overflow stream banks,
lakes, and wetlands.
Water quality maintenance through at least partial removal of sediment, phosphorous, and nitrogen from runoff passing through a naturally vegetated buffer. Forested buffers also contribute to neutralization of acid precipitation, due to ion
exchange that occurs as precipitation filters through the forest canopy.10 The percentage of pollutants removed depends on the pollutant load, the nature of the
material, the amount of runoff, and the character of the buffer area. The pollutant
removal rate is not generally a linear relationship with buffer width, but decreases
with increasing buffer width. More pollutant removal occurs in the first 100 ft than
the second hundred, for example.2
Wildlife habitat opportunities, such as providing foraging and nesting habitat as well
as cover for a mix of upland, aquatic, and wetland species. Buffer areas can also serve
as travel routes for migratory and nomadic, as well as resident, species. They also
support plant diversity in the ecotones represented. Buffers protect surface waters
and wetlands from temperature increases, which reduces waters capacity to hold
oxygen. Leaf litter and woody debris from buffers along smaller streams supply most
of the energy processed by the stream. The woody debris also traps the leaf litter,
making it available to organisms over a long period of time (Figure II.1.2). These
streamside buffers help stabilize banks as well, and naturally undercut areas beneath
tree roots offer cover for fish, turtles, and other creatures.1,11
Recreation and aesthetics benefit from the visual screen buffers provide along
surface waters, and they frame wetlands and surface waters in the landscape,
particularly in hilly or mountainous terrain.
Chapter II.1:
365
Factors affecting buffer functions

All buffers are not the same. Each, as part of a living system, functions differently. The
characteristics of a shoreland buffer affect its functional capacity. Several factors other than
buffer width influence the effectiveness of shoreland buffers. Some of these are described
here.
For water quality

Soil type Less erodible soils within a buffer pose less risk of buffer degradation
from runoff passing through it (channeling). More permeable soils allow greater
infiltration of rain and melted snow into the ground.
Vegetative type and density Woody plants at a natural density generally do the
best job of holding soil in place within buffers. They also contribute to buffer
stability by intercepting rain and snow (in the range of 20% in the northeastern
United States, depending on species) and reducing sunlight penetration, thereby
moderating snow melt that often precedes peak runoff times.
Surface roughness of the land Undisturbed terrain retains its capacity to trap
surface water through ponding and groundwater infiltration.
Season In northern areas, buffer effectiveness can be reduced substantially during times of the year when the ground is frozen and plants are dormant.
The nature of the land beyond the buffer Buffers have their limits as to how
much water or pollution they can absorb. The more intense the land use (more
likely to generate pollution such as eroded soil) above a buffer, the more the buffer
is in jeopardy of having its assimilative capacity exceeded.1
For wildlife habitat

What the buffer connects to (landscape position) along its length Buffers can
function both as resident (in-place) habitat and as travel routes for wildlife. As
resident habitat, a buffers value is supplemented by other habitats with which it
physically connects. This is important because larger habitat blocks are known to
support greater biodiversity than smaller ones. Two blocks of habitat, for example,
can function as a larger unit if connected by an area of adequate width. Animals
using a buffer as a travel route require the route to lead somewhere productive,
i.e., to a viable habitat block. In either case, then, resident habitat or travel route,
a buffer functions best as habitat when it is connected to other habitats, including
other buffers (Figure II.1.3).
Integrity of the connectivity of the buffer One of the greatest challenges in
establishing shoreland buffers is avoiding or minimizing the effects of fragmentation of the buffer. The network of roads in most parts of the country makes it
difficult to avoid fragmentation altogether, and the ecological effects of roads can
be substantial. These effects can include direct mortality (roadkill), modification
of animal behavior, alteration of the physical and chemical environments, and the
introduction of exotic species. These effects occur in terrestrial habitats along
roadsides as well as extending downstream in aquatic and wetland systems with
moving water (Figure II.1.4). This can fragment the aquatic and wetland habitats by
altering hydrology, increasing sedimentation, and introducing pollutants. These effects can, in turn, lead to discontinuous floral and faunal assemblages in such aquatic
systems.12,13
366
Streams and wetlands

Wetland edge
Stream
Ponds and Lakes
100 ft. Stream Buffer
Roads
Conservation Land
Areas for Conservation Focus
Figure II.1.3 Buffers as habitat connections. The most effective buffers connect to blocks of protected
land and connect with other wetland and surface-water buffers. This map shows how riparian
buffers can connect existing conservation areas and suggest areas for focusing on riparian buffer
protection. In the example, the circles depict wetlands, stream segments, and their surrounding
uplands that, if protected, would add significantly to the habitat value of the protected stream buffer
in the watershed. Such protection would have water quality, habitat, and aesthetic benefits.
Abruptness of the edge between the buffer and the land use beyond it When
buffers become fragmented strips between water and land, they may be subject to
negative edge effects of predation and parasitism as well as physical effects such
as wind, drying, temperature increase, and blowdown of trees. Edge habitats tend
to harbor disproportionate populations of predators such as blue jays (Cyanacitta
cristata), crows (Corvus brachyrhynchos), raccoons (Procyon lotor), skunks (Mephitis
mephitis), red foxes (Vulpes vulpes), and dogs and cats. Animals like these use the
edge habitats and penetrate from it into woodlands.1,8
Chapter II.1:
367
Legend
Roads
Road effect area
Affected stream section
Streams
Wetlands
Figure II.1.4 Road effects. The effects of roads follows flowing water downstream, sometimes changing physical, chemical, and biological conditions for a substantial distance downstream.
In the northeast, the brown-headed cowbird (Molothrus ater) is the nest parasite of
concern in relation to migrant songbirds. Cowbirds lay their eggs in nests of other species,
which then usually raise them as their own. Cowbird young hatch quickly and tend to
be larger than their host species, giving them a competitive advantage in feeding. They
may also physically remove their host species offspring from the nest. Cowbirds are native
to the midwest and have expanded their range throughout the northeast in response to
land clearing for agriculture and other uses.1
Some biologists feel that negative edge effects can be reduced by managing buffer
areas to have gradual or patchy (soft) edges on the upland side, rather than
abrupt, or hard ones. Another suggested technique is establishing or maintaining
dense vegetation along the upland buffer edge.8,9 In effect, this means a transitional
upland buffer to support the wetland/surface water buffer habitat functions more
fully (a buffer of sorts for the shoreland buffer).
Habitat structure (layering of vegetation, dead trees with cavities, etc.) The
structure provided by a riparian forest determines which species can use the
habitat. Habitat structure includes:
368

Horizontal diversity (the horizontal arrangement of plant communities in an
area)
Vertical diversity (the degree to which plants are layered within an area)
Soil qualities
Dead standing trees
Downed logs
Boulders, cliffs, etc.14
Vegetation type The species of plants in an area generally determine the animals
that will occupy the area. Dense stands of evergreen trees, for example, are well
know for their value as deer wintering areas in New England. Hard mast (nut
producing) trees, such as oaks and beech, provide food for a number of species
(such as bear, deer, squirrels, and jays).15
Establishing and maintaining buffers

Assuming the vegetation is already present, protected shoreland buffers can be established
through voluntary or regulatory means. The authors state of New Hampshire offers
examples of both of these approaches, as described in the following sections.
Voluntary measures
Publications
Education about buffers has become widespread in the United States in recent years,
resulting in a number of publications and other educational materials on the topic. The
northern New England state of New Hampshire is an example of such efforts. The publication Buffers for Wetlands and Surface Waters: A Guidebook for New Hampshire
Municipalities1 was produced cooperatively by four public and private organizations and
has been distributed widely in the state and used as the basis for dozens of educational
presentations. It was produced primarily to assist public officials in making decisions
about regulatory buffers. It has influenced public policy decisions from the local to the
state level. Buffers for Wetlands and Surface Waters recommends a reasonable minimum 100-ft
buffer for wetlands and surface waters to protect water quality and documents the rationale
for this recommendation. Though some habitat value is maintained with the 100-ft recommendation, the publication acknowledges that this is generally an insufficient size buffer to
protect biodiversity and provides guidance on using larger, voluntary buffers for this purpose.
A related publication, A Guide to Developing and Redeveloping Shoreland Property in
New Hampshire, A Blueprint to Help You Live by the Water,16 was also an interagency effort
involving 13 public and private organizations. The booklets purpose is to provide landowners with the understanding and tools necessary to develop your shoreland property
in a manner that meets [the landowners] goals while at the same time maintaining the
quality of the water body for everyone.
Good Forestry in the Granite State: Recommended Voluntary Forest Management Practices
for New Hampshire17 is yet another collaborative project, this time involving 15 groups.
Good Forestry in the Granite State presents recommendations for protecting water quality,
wetlands, and riparian areas in a forest management context. Its purpose is to provide
New Hampshire landowners, and the professionals work with them, practical recommendations on sustainable management practices for individual forest ownerships. Similar
to Buffers for Wetlands and Surface Waters, this publication has been widely distributed and
used in educational programs throughout the state. Because it emphasizes voluntary
Chapter II.1:
369
Table II.1.2 Legally Required and Recommended Voluntary Shoreland Buffers for Forestry in New
Hampshire
Feature
Intermittent streams
First- and second-order streams
Third-order streams
Fourth-order and larger streams
Pond (<10 acres)
Lake (>10 acres)
Nonforested wetland (<10 acres)
Nonforested wetland (>10 acres)
Legally Required
Management
No-Harvest
Zone
Zone
None
50 ft
50 ft
150 ft
50 ft
150 ft
None
None
None
None
None
None
None
None
None
None
Recommended
Management
No-Harvest
Zone
Zone
100 ft
100 ft
300 ft
600 ft
100 ft
300 ft
100 ft
300 ft
None
None
25 ft
25 ft
None
25 ft
None
25 ft
Note: Recommended Voluntary Best Management Practices (RVMPs) for shoreland buffers on forestry lands in
New Hampshire, presented in Good Forestry in the Granite State, advise substantially larger buffers than
regulations provide. The regulations are focused on water quality protection, whereas the RVMPs consider
habitat needs as well.
Source: Adapted from NH Forestry Sustainability Standards Work Team, Good Forestry in the Granite State:
Recommended Voluntary Practices for New Hampshire, Soc. Protection of NH Forests, Concord, NH, 1997.
adoption of buffer recommendations and a strong consideration of habitat requirements,

this document makes recommendations for substantial buffer widths. For example, Good
Forestry in the Granite State suggests buffers from 100 ft (from intermittent streams to
second-order streams, ponds, and open wetlands) to 600 ft for fourth-order and larger
streams (Table I.1.2)17
Landowner stewardship, or voluntarily maintaining sufficient shoreland buffers in forest,
residential, and other landscape settings is happening all the time.
Current use of property tax assessment can provide some incentive for landowners to
keep land in forestry, agriculture, or other open-space uses.
Landscape design for development projects can incorporate adequate or even optimal
protection for water resources by including shoreland buffers.
Land conservation can protect shorelands over the long term. Conservation easements
and land acquisition for conservation purposes are the most common tools employed.
Regulatory measures
State regulations
An example of state regulation of shoreland buffers can be found in New Hampshires
Comprehensive Shoreland Protection Act, which requires shoreland buffers (150 ft, except
a building may be built within 50 ft of the shore) along the states lakes (>10 acres), fourthorder streams, and tidal waters.
State forestry and agriculture best management practices can be legally enforceable
and focus mainly on reducing erosion and sedimentation from these land uses. They may
require buffers (filter strips) of varying width depending on slope and the type of water
resource being protected.
Municipal zoning
Many communities have municipal zoning ordinances that offer some degree of protection
for shorelands.
370
Public attitudes toward buffers

Results of public opinion surveys
Awareness of the functions and values of shoreland buffers among scientists, landscape
planners, and regulators is relatively recent, so one would not expect that the public would
be particularly well versed about the workings and importance of buffers in providing
critical habitats and maintaining water quality. However, there is evidence of public
support for the resources buffers can protect and for some types of buffer implementation
methods. Three examples from New Hampshire illustrate this.
In 1996 a Community Vision Committee surveyed town residents in Deerfield, NH
(pop. of 3124 in 1990), asking questions on a variety of topics related to the character
people wanted their town to have in the future. Questions included open-ended, multiplechoice, and rating-scale (agree/disagree, very important, important, not important, etc.)
types. The response rate was about one-third.18
In 1998 a private conservation advocacy group, the Concerned Citizens for a Safe
Environment, completed a survey of landowners along the Lamprey River in Raymond,
NH (estimated pop. of 9398 in 1999). The group got a 25% response rate to the survey,
which was done as part of a planned nomination of the river into the states Rivers
Management and Protection Program. Seventy-five percent of survey respondents reside
in the town and 72% use the property as a full-time residence.19
The Friends of the Cold River, a private conservation group, completed three landowner surveys in 1999, in cooperation with the University of New Hampshire Cooperative
Extension and the Upper Valley Regional Planning Commission. One survey questioned
landowners within the rivers watershed, the second targeted those within mile of the
river, and the third questioned those within mile of a lake that is the rivers headwaters.
The discussion below covers the first two surveys. Of the landowners in the river corridor
survey, one third owned land that was actually on the river. The Friends of the Cold River
received a response rate of 32% from the 1984 surveys it sent. Six rural communities are
in the Cold River Watershed.20,21
These surveys, though not scientifically conducted, give valuable insight into landowner attitudes and opinions, at least for the northeastern state in which they were done.
The consistency between the Lamprey River and Cold River surveys, in different parts of
the same state, supports an interpretation of credibility of the results despite the fact that
the surveys were conducted by laypeople.
Attitudes toward the value of surface waters, wetlands, and shoreland property
Deerfield community vision survey
The first question of the survey was open-ended, asking, What do you think is
(are) the most desirable features of Deerfield? Eighteen of 299 respondents (6%)
wrote lakes or rivers in a section on physical character of the town.
Ninety-seven percent of survey respondents answered that clean water was very
important or important to protecting the towns environment and natural resources. Ninety-five percent indicated that water resources were this important.18
Concerned citizens for a safe environment landowner survey

Landowners responding who indicated that the river played a part in their decision
to purchase their property: 60%
Chapter II.1:
371
Respondents who answered that they think the Lamprey River contributes to the
quality of life in the community: 95%
Top reasons for this view, in order of frequency, were
Fishing
Wildlife and waterfowl habitat
Scenic value
Open space
100%
95%
93%
75%
Landowners who cited shoreline development (the least cited category) as the
quality of life contribution:
15%19
Cold River area surveys

River
Watershed
corridor
landowners landowners
Live on the land that was the subject of the
survey (primary residence)
Plan to continue the current land use of their property
Plan to keep the property in the family
60%
84%
50%
69%
83%
41%
Rated the following quality of life features associated

with the Cold River as high on an importance scale
of lowmediumhighno opinion:
River
Watershed
corridor
Good water quality
90%
87%
Diverse wildlife habitat/wildlife and waterfowl habitat
76%
81%
Forest lands
77%
71%
Scenic qualities
75%
80%
Fisheries habitat
Not included
73%
Wetlands
Not included
64%
Agricultural lands
68%
55%
Cited as a top concern

(when asked to list top three concerns):
Water pollution
Building too close to the river
Clearing too close to the river
Not included
Not included
Not included
47%
28%
24%
Landowners observed:
Water pollution
Building too close to the river
Clearing too close to the river
Not included
Not included
Not included
19%
14%
22%20,21
Attitudes toward resource protection

In response to the question, Would you favor an increase, a decrease or no change in the
current levels of each of the following? (a list of land uses followed), 65% of respondents
chose town conservation land as a land use for which they wanted an increase. This
372
was the answer getting the greatest response for an increase. Private undeveloped land
and open space were the third and fourth most desired future land uses.18

Landowners who intend to keep their land in its present use
Those who plan to build on the land
Those who plan to subdivide
90%
2%
2%
In answer to the question, How important is it that each of the following characteristics associated with the Lamprey River and its corridor are conserved? responses of
very important or important were given as follows:
Open space
Scenic quality
Water quality
Wildlife and waterfowl habitat
Fishing access
Fisheries habitat
Wetland ecosystems
Residential development opportunity
Commercial or industrial development
93%
83%
82%
75%
66%
63%
63%
18%
2%
Respondents who reported noticing development too close to the river: 40%
In response to the question, Do you believe that any of the following measures
should be taken to protect the Lamprey River and the special opportunities it offers
to the area? landowners answered:
Limit shoreline commercial development
Protect water quality
Limit shoreline industrial development
Protect the scenic character of the river corridor
Protect wildlife and waterfowl habitat
Protect fisheries habitat
Limit shoreline residential development
Protect the free flowing nature of the river
No additional protection needed
83%
80%
75%
75%
73%
60%
50%
50%
8%19
Cold River area landowner survey

River
Watershed
corridor
Landowners who indicated intent to protect some
or all of their land from development
Landowners who plan to sell their land at some point
without subdividing
Those who plan to subdivide
Respondents who have a written management plan
for their farm or forest land
Those interested in developing one
37%
19%
18%
5%
13%
1%
19%
25%
19%
25%
Chapter II.1:
Landowners who were not aware of published best

management practices for farm and forest land
management (water quality BMPs)
For the question, Which of the following would be
of interest to you for your lands? responses were as
follows for high interest:
Maintain the lands natural beauty

Protect streams and wetlands
Protect from development
Identify and maintain wildlife habitat
Improve timber/firewood production
Improve production of forest products
Improve soils
Leave it alone (no management)
Grow horticultural products
Improve animal or crop production
Not sure/not applicable
58%
Watershed
landowners
84%
80%
73%
69%
44%
40%
34%
30%
24%
23%
13%
373
58%
River
corridor
landowners
Not included
Not included
Not included
Not included
Not included
Not included
Not included
Not included
Not included
Not included
Not included20,21
Opinions on appropriate protective methods

The Deerfield survey did not deal with issues of specific protective measures.18

In response to a question about which specific steps would be appropriate for river and
river corridor protection, landowners answered:
Minimum setbacks for new construction
Floodplain protection regulations
No additional protection needed
Purchase property in the river corridor (from willing seller)
Purchase of development rights in the river corridor
Voluntary easement donation program
65%
48%
48%
40%
35%
25%19
Cold River area surveys

The Cold River surveys included the similar questions, Which of the following options
are appropriate for conserving land and the natural resources associated with it? and
Which of the following do you feel are appropriate for protecting the Cold River and
the resources associated with it? Responses of a high level of appropriateness were as
follows:
River
Watershed
corridor
Incentives for landowners to manage land appropriately
68%
60%
Landowner education
60%
61%
Education of municipal officials
58%
53%
374

Coordinated planning among towns
Conservation planning in utility and road projects
Voluntary protection through conservation easements
Voluntary protection through public acquisition
Stricter enforcement of state regulations
Stricter enforcement of existing local regulations
Stronger local regulations
Stronger state regulations
58%
55%
53%
37%
34%
35%
20%
18%
46%
49%
40%
34%
34%
37%
20%
17%20,21
Conclusions from public opinion surveys

These survey results indicate that citizens and landowners in these geographic locations
generally support protection of the features and resources associated with water resources
in general and shorelands in particular. Many of the values respondents cite as important
to them are dependent to some degree on healthy shoreland ecosystems for their integrity.
Values intrinsic to aquatic and wetland systems are of primary importance to most survey
respondents. These include good water quality, wildlife habitats, scenic values, and openspace values. Survey respondents reported interest in nonconsumptive uses at a greater
level than interest in resource extraction activities.
Similarly, people are concerned about threats to these resources that shoreland buffers
can help ameliorate water pollution, land clearing and building near rivers, and loss
of wildlife habitat.
A high proportion of landowners report their intent to keep their land in its current
use, an attitude that bodes well for voluntary shoreland protection, at least until the land
changes ownership.
When asked what to do to protect resources related to rivers and shoreland areas,
people strongly approved of limiting shoreland development, protecting water quality,
wildlife habitats, and scenic and recreational values. They generally favored voluntary
over regulatory approaches to accomplish this, with the exception of Lamprey River
respondents, a majority of whom favored minimum setbacks for new construction (buffers
were not a choice in the question). Voluntary measures most preferred included landowner
incentives, education, and improved local planning, but permanent land protection
through conservation easements or public acquisition was favored by one-third to onehalf of the Lamprey and Cold River landowners.
An interesting variation from the consistency seen through much of the surveys
appeared in the Cold River survey. Thirty-seven percent of landowners in the watershed
said they intend to protect some or all of their land from development, but only 19% in
the river corridor said this. However, 5% of watershed landowners claimed to intend to
subdivide, as opposed to only 1% of river corridor landowners.
It is also interesting to note that the level of concern expressed by landowners in the
Cold River survey is apparently greater than their preparedness to act accordingly. A
majority of the Cold River landowners had no written management plan for their land
and were not aware of published best management practices (BMPs) to protect water
resources. Almost half (45%) of the Cold River Watershed landowners, however, stated
that they know where to find appropriate personal assistance in managing their lands.
Summary of buffer concepts

Several concepts apply to shoreland buffers for both water quality protection and wildlife
habitat conservation:
Chapter II.1:
375
Buffers are most effective as part of an overall watershed protection strategy.

Connectivity is crucial. A continuous buffer will have more integrity and effectiveness than a fragmented one.
Core areas can add a level of protection beyond what routine buffers can provide.
For water quality, this might mean a more substantial buffer around a wetland
that has a particularly high stormwater storage and pollution removal capacity.
For wildlife, the core area might be an ecological reserve of significant size to which
shoreland buffers (as travel routes for wildlife) connect.
Even a fragmented buffer provides some benefits for water quality and wildlife.
Some gaps can be filled through restoration efforts. (Chapter II.2 describes considerations for river restoration planning, including riparian greenways.)
The character of the land influences the functional quality of the buffer. Soil type,
vegetation type, and slope are the factors that most influence buffer function.
Conclusions
Buffers can protect both habitat and water quality functions and values, but protecting habitats requires wider, connected buffers.
Voluntary and regulatory methods may be used to protect buffers, preferably in
combination. Voluntary measures have the potential to provide optimal protection,
but rely on landowner initiative. Regulatory measures apply broadly but typically
provide less than ideal, and sometimes less than adequate, protection. Thus, a
strategy that encompasses both approaches can result in minimum standards being
met widely, with more desirable protection occurring in areas where landowners
have chosen to provide it. Such voluntary protection can be assured over long time
periods through conservation easements and land acquisition by public agencies
or private conservation groups.
Citizens appear to support the water resources and other features that shoreland
buffers can protect. They also appear to support actions needed to conserve ecological and social functions and values of buffers, particularly voluntary ones.
A landscape design and management approach that embraces water quality protection and conservation of biodiversity will give consideration to shoreland buffers as a means to support both functions. Though buffers alone cannot fully protect
water quality and biodiversity, they are a key piece in a larger landscape view that
would encompass these goals. A landscape design that follows this path will
complement regulatory efforts by others to protect the same features, resources,
functions, and values. Aesthetics and other social values will also be supported
by such an approach.
Literature cited
1. Chase, V.P., Deming, L.S., and Lataviec, F., Buffers for Wetlands and Surface Waters: A Guidebook
for New Hampshire Municipalities, 2nd ed., Audubon Society of NH, Concord, NH, 1997.
2. Desbonnet, A. et al., Vegetated buffers in the coastal zone, a summary review and bibliography. Coastal Resources Center Technical Report No. 2064, Univ. RI Grad. Sch. Oceanography, Naragansett, RI, 1994.
3. Castelle, A.J. et al., Wetland Buffers: Use and Effectiveness, Adofson Associates, Inc., Shorelands
and Coastal Zone Management Program, WA Dept. of Ecol., Olympia, WA, 1992.
4. Groffman, P.M. et al., An Investigation into Multiple Uses of Vegetated Buffer Strips, Univ. RI,
Dept. Nat. Resour., Kingston, RI, undated.
376
5. Palmstrom, N., Vegetated Buffer Strip Designation Method Guidance Manual, I.E.P., Inc., Consulting Environmental Scientists, Northborough, MA, 1991.
6. Rogers, Golden & Halpern, Wetland Buffer Delineation Method, NJ Dept. of Environ. Protection, Trenton, NJ, 1988.
7. Roman, C.T. and Good, R.E., Buffer Delineation Model for New Jersey Pinelands Wetlands,
Rutgers, State Univ. NJ, New Brunswick, NJ, 1985.
8. Smith, D.S. and Hellmund, P.C., Ed., Ecology of Greenways, Univ. MN Press, Minneapolis, 1993.
9. Noss, R.F. and Cooperrider, A.Y., Saving Natures Legacy, Protecting and Restoring Biodiversity,
Island Press, Washington, DC and Covelo, CO, 1994.
10. Hornbeck, J.W., Likens, G.E., and Eaton, J.S., Seasonal patterns in acidity of precipitation
and their implication for forest stream ecosystems, Water, Air and Soil Pollution, 7, 355, 1977.
11. Carroll, D.M., Deadfalls, turtles and trout, Wild Earth, 11, 57, 2001.
12. Trombulak, S.C. and Frissell, C.A., Review of ecological effects of roads on terrestrial and
aquatic communities, Conserv. Biol., 14, 18, 2000.
13. Forman, R.T. and Deblinger, R.D., The ecological road-effect zone of a Massachusetts (U.S.A.)
suburban highway, Conserv. Biol., 14, 2000.
14. DeGraff, R. et al., New England wildlife: Management of forested habitats, General Technical
Report NE-144, USDA Forest Service Northeast Forest Experiment Station, 1992.
15. DeGraff, R.M. and Yamasaki, M., Bird and mammal habitat in riparian areas, in Verry, E.S.,
Hornbeck, J.W., and Dolloff, C.A., Eds., Riparian Management of Forests of the Continental
Eastern United States, Lewis Publishers, Boca Raton, FL, 1999.
16. Lobdell, R., A Guide to Developing and Re-developing Shoreland Property in New Hampshire:
A Blueprint to Help You Live by the Water, 2nd ed., North Country Resource Conservation and
Development Area, Inc., Meredith, NH, 1995.
17. NH Forestry Sustainability Standards Work Team, Good Forestry in the Granite State: Recommended Voluntary Practices for New Hampshire, Soc. Protection of NH Forests, Concord, NH,
1997.
18. Deerfield Community Vision Committee, Results of the Deerfield vision survey, Town of
Deerfield, NH, 1996.
19. Concerned Citizens for a Safe Environment, Survey of landowners along the Lamprey River,
Raymond, NH, 1998.
20. Friends of the Cold River, Cold River area landowners survey, Acworth, NH, 1999a.
21. Friends of the Cold River. 1999b. Cold River corridor survey, Acworth, NH, 1999b.
Chapter II.1:
377
Response
Buffer strips: More than green eyelashes?
The region where land meets water the riparian ecotone is known to demand special
attention in environmental management. Shoreline buffers can function protectively as
the first line of defense for water bodies from land development, as, for example, discussed
by Robert France in Chapter II.16 with respect to soil erosion, or for human occupants of
floodplains, as, for example, described by Leslie Zucker, Anne Weekes, Mark Vian, and
Jay Dorsey in Chapter II.5 in terms of water management. Because shoreline buffers
provide not only these functions but also serve as human amenities in green recreational
corridors (Charles Flink in Chapter II.3), much attention has been devoted to measuring
their size and extent on a landscape scale (Jeffrey Schloss in Chapter II.12 and Robert
France, John Felkner, Michael Flaxman, and Robert Rempel in Chapter II.14). This chapter
by Frank Mitchell is important in emphasizing the need for bridging, conceptually as well
as physically, across spatial scales.
Mitchell repeatedly makes the point that to be truly effective, shoreline buffers must be
coupled to the larger landscape of whole watersheds and continuous habitat areas. The
importance of this simple precept for the profession of water sensitive planning cannot be
overstated. This is particularly the case for concerns about wildlife biodiversity. As Mitchell
mentions, though a buffer strip of only 100 ft or so may be sufficient to filter out sediments
and entrap contaminants (as, for example, shown by Frances work in Chapter II.16), such
widths are woefully inadequate for protecting riparian wildlife.
Mitchells chapter is most informative in its demonstration of how the concepts of
landscape ecology such as fragmentation, connectivity, edge effects, patches, nodes, etc.
should affect shoreline planning. As he correctly states, one of the greatest challenges in
establishing networks of shoreline buffers is in avoiding or minimizing gaps, which can
later be filled in with localized restoration efforts. Water sensitive planners must therefore
learn how to place their regions of interest and study back into the larger landscape in
order to effectively buffer deleterious land-use changes. For example, roads, which Mitchell touches upon briefly in this chapter, may have a negative influence upon riparian
wildlife (e.g., directly through roadkills or indirectly through breeding migration interference) even though they may be located many kilometers distant.
Finally, Mitchells chapter becomes even more significant in his encompassing of the
sociology implicit, yet frequently ignored, in shoreline protection. The role of education
in fostering initial interest and then in sustaining surveillance is of critical importance.
The results from the described public surveys demonstrate the support professed by
landowners for both human and environmental (wildlife) benefits accruing from shoreline
protection. Water sensitive planners desperately need this sort of information at an early
stage in order to justify to the skeptics (who may hold the purse strings) that their intents
are obviously in the best interests of both the wildlife and the human populations.
chapter II.2
River restoration planning (Connecticut)

James G. MacBroom
Abstract
River restoration refers to the modification of a previously disturbed rivers physical,
chemical, or biological properties to simulate those of a healthy natural river. This chapter
describes several types of river restoration projects with emphasis on multidisciplinary
collaboration and public participation.
Introduction
Restoration of degraded rivers is a complex process involving expertise in a broad range
of physical and social sciences. Historically, flood control, water supply, water quality,
urban waterfronts, hydroelectric power generation, and navigation improvements were
treated as independent programs to facilitate the nations growth and economic development; however, this has resulted in adverse environmental and social impacts. Today, we
seek to integrate the separate programs into unified efforts to balance ecological and
human needs and to create a sustainable environment. This creates a need for design
professionals and regulators to have cross training between the traditional scientific and
planning disciplines while maintaining the skills of a specialist. The classic academic
programs need to introduce students to the role of interdisciplinary teams that span a
broad range of expertise. Specific project types include dechannelization, removal of old
dams, creation of linear greenways and trails, instream habitats, daylighting enclosed
channels, and the cleanup of abandoned industrial properties.
River restoration projects require a thorough knowledge of watershed processes and
their interactions, including hydrology, hydraulics, water quality, fluvial morphology, plus
the biological sciences. In addition, developed areas must focus on community-sensitive
designs that address land use, demographics, aesthetics, recreational opportunities, public
access, historic sites, and linkages to cultural features.
Several trends have reinforced the interest in river restoration in recent years. The first
is the water quality improvements since passage of the federal Water Pollution Control
Act in 1972. Few people are interested in recreation in polluted streams. As water quality
improves, so does public demand for water access, recreation, and aesthetics. A second
major factor is the demise of many older 19th-century, water-dependent industries. Many
dams built for waterpower or water intakes are no longer actively used and can be
removed, and urban riverfronts with crowded industrial facilities that were formerly
water-dependent are now available for adaptive reuse.
1-56670-562-2/02/$0.00+$1.50
379
380
Human impacts on rivers

Many early communities developed along rivers that provided transportation, water
supply, fisheries, waterpower, and irrigation. The clearing of forestland, filling of wetlands,
and drainage of agricultural lands decreased groundwater recharge and increased surface
runoff even in colonial periods. More recently, urbanization has had a tremendous impact
on both surface-water quality and quantity. Impervious areas increase the percentage of
precipitation that becomes surface runoff, while paved gutters, channels, and storm drains
accelerate the rate at which water moves downstream, causing larger flood flow rates.
The higher peak runoff rates found in heavily disturbed watersheds and urban areas
increase water velocities in rivers, leading to increased erosion of the channel bed and
banks with a general enlargement of the waterway area. The erosion process disturbs the
channel substrate and aquatic habitat, reducing biological diversity and productivity.
Downstream areas may receive excessive sediment loads, leading to poor water quality
and habitat impacts.
Other human activities have impacted riverine systems by directly altering the natural
rivers. Many communities dredged or realigned rivers to channelize their flow for navigation, flood control, or land reclamation purposes. Channels that have been straightened
will have steeper slopes with higher velocities that encourage erosion, and the shorter
length reduces the size of aquatic habitats. Widening channels for flood control reduce
flow velocities, often leading to very slow velocities and shallow base flow depths with
subsequent sediment deposition. Channelization may also increase downstream peak
flood flows by reducing water storage on floodplains and in wetlands.
Filling channels or floodplains and stabilizing riverbanks with stone, concrete, and
retaining walls reduce riparian vegetation and habitat diversity. In addition, artificial
linings limit public access and affect aesthetics. Many smaller tributary streams have been
enclosed underground, resulting in a loss of all ecological functions (Figure II.2.1).
Watershed management
A rivers flow rate and water quality are dependent on the land-use practices within their
entire watershed. Consequently, watershed management is an essential part of maintaining healthy productive rivers.
Comprehensive watershed planning involves three separate but interrelated planning
levels that must be coordinated (MacBroom, 1998). The watershed includes all the land
that naturally drains to the specific point of interest. Its perimeter, called the watershed
boundary, marks the divide between one basin and another. Watersheds may range in
size from a few acres to thousands of square miles. Small watersheds are generally more
sensitive to land-use conditions and changes while large watersheds typically have broad
floodplains and channels that attenuate much of the land-use influence. The hydrologic
cycle and pollution sources need to be addressed for the entire drainage basin, which
includes all lands that contribute runoff to the project site. The influences of topography,
soils, vegetation, land uses, and runoff need to be considered.
The intermediate planning level consists of the stream corridors, including natural
resources such as floodplains, wetlands, most aquifers, and riparian buffer zones. These
areas are essential transitions between upland areas and waterbodies. Riparian corridors
carry flood flows, recharge aquifers, filter pollutants from overland flows, reduce water
velocities and peak flow rates, and provide wildlife habitat areas. Many riparian corridors
have been disturbed by agricultural and urban land uses, plus linear features such as
highways, railroads, and dikes that tend to follow rivers and thus isolate riverfront areas.
Figure II.2.1
Retaining Wall
Source of Organic
Material and Shade
Filters Runoff
and Sediment
Culvert
Channelized
Channel
Piped
Drainage
Original
Stream
Channel
Bridge
Culverted
Tributaries
Dam w/
Diversion
Culvert
Concrete
Liner
Direct
Discharges
Fill
Wildlife
Corridor
Floodplain
Chapter II.2:
Progression of Urbanization
PROGRESSIVE URBANIZATION OF A STREAM
381
382
Important elements of urban river corridor planning include their relationship with adjacent
neighborhoods, active recreation needs, environmental education, and habitat linkages.
The final planning level focuses on the actual streams and river channels, plus ponds
and lakes. It addresses aquatic habitat, channel capacity and stability, fish passage, and water
quality. In urban areas, water use, public access, and recreation are also important factors.
Intense land-use activities in urban, industrial, and agricultural areas can produce
significant point and nonpoint sources of pollution. The need to address the water quality
impacts of combined point and nonpoint sources of pollution is forcing the U.S. EPA and
individual states to develop total maximum daily (pollutant) loads for impaired rivers.
This places more emphasis on the overall watercourse quality rather than setting effluent
discharge limits or specifying the best management practices. Nonpoint sources originate
from a broad range of human activities, therefore, a watershed management approach is
necessary to evaluate and manage cumulative water quality impacts.
Watershed management is also adaptable for controlling peak flow rates and runoff
volumes. This involves land-use and site design criteria to minimize impervious cover,
increase infiltration to groundwater, preserve wetlands, and avoid accelerating runoff
movement. Low-impact site developments include use of porous paving, grass swale
drainage systems, curbless roads to encourage sheet flow, and drywells for roof runoff.
The emphasis is on minimizing stormwater runoff at its source rather than trying to collect
it and rapidly pass it downstream.
The concept of river restoration has evolved in response to an interest in improving
degraded resources. River restoration ranges from urban areas that seek to revive community values, to rural areas that seek to improve aquatic habitats. Even small restoration
projects require careful attention to the basic water resource-related sciences, as noted in
the examples that follow.
Dam removal
Many of the nations rivers have dams that were constructed for public, or industrial water
supply, waterpower, navigation, and recreation. The United States has an estimated 75,000
significant dams (Maslin and Sicchio, 1999). There is nationwide interest in the removal
of old, obsolete dams that are no longer functional. Some have been removed due to
physical deterioration of aging structures, others due to the abandonment of 19th-century,
water-power-operated mills.
Whereas many dams continue to provide essential public benefits such as storing
water supplies or creating hydroelectric power, other dams that are no longer functional
can actually have negative impacts. Specific negative impacts due to dams include altering
the rivers sediment transport, trapping contaminated sediment, reducing flow velocities,
decreasing dissolved oxygen levels, inundating riparian wetlands or floodplains, obstructing canoes or kayaks, and, most important, blocking fish movement, which creates fragmented habitats. Dams also affect downstream areas because their discharges have
reduced sediment loads that encourage downstream erosion as the river seeks to balance
sediment loads versus transport capacity (ASCE, 1997). In addition, older dams that are
not properly maintained can create hazardous conditions in the event of a failure.
The impact of dams on fish movements and populations is a major issue in coastal
areas with anadromous species, as well as in inland waters. Anadromous fish such as
salmon, shad, alewife, and herring spend their adult lives in the ocean but return to swift
freshwater rivers to reproduce. Passage facilities, including fish ladders, elevators, and
trap and haul operations have had limited success.
Chapter II.2:
383
At first glance, some may think of dam removal as a simple demolition project, but
the actual removal of the structure is often the easy part. The key issue is to create a
subsequent channel suitable for fish passage and to control excess sediments. Each species
of fish has unique needs in terms of allowable and optimum flow depths and velocities
required for passage. The post-dam channel design requires a careful hydrology analysis
of seasonal flow rates to identify criteria for mean flows during the migratory runs in
addition to flood flows. A hydraulic analysis is required to determine the final channel
dimensions and slope for fish passage and flood conveyance, integrated with a geomorphic
assessment of river patterns and stable conditions (Figure II.2.2).
One of the critical issues related to dam removal projects is the management of
sediments often found in the impoundments. The impoundments trap most of the rivers
bedload sediments and a portion of the suspended sediments, depending on the ratio of
the watershed size to the pond volume and area. Sediments within the impoundments
must be explored and decisions made with regard to leaving the sediments in place or
removing it. In some cases, sediments may be contaminated and could harm downstream
water quality. Sediment management options include removal to off-site disposal areas,
stabilization in place, on-site relocation, or allowing gradual erosion to occur.
After dams are breached or removed, the unvegetated soils in their former impoundments may need to be reshaped or landscaped to minimize erosion. The reclaimed land
may be used for wetland habitat or parks.
The Naugatuck River restoration project in western Connecticut is an example of a
complex, multifaceted program to rehabilitate a formerly polluted urban river. For 150 years,
the Naugatuck River Valley supported copper and brass industries dependent on the river
for mechanical power, cooling water, rinse water, steam boiler water, and waste assimilation. Most of the industries have either closed or become less water-dependent, enabling
communities to reclaim former industrial properties and restore the river. Key elements
include an advanced wastewater treatment plant in Waterbury to improve water quality,
new sanitary sewer interceptors to prevent overflows, and fish passage at eight abandoned
industrial dams for species that can take advantage of improved water quality. The fish
passage component of the project includes removal of the Union City, Freight Street, Platts
Mill, Chase Brass Dams, and Anaconda Dams, a fish ladder at Brays Buckle Dam, a ramp
at the Plume and Atwood Dam, and a bypass channel at the Tingue Dam. Additional
project features include improved public access, fish stocking, riverbank clean upstream,
tree planting, and modification of sediment bars at tributary stream confluences. The
program stakeowners include federal and state agencies, local communities, private businesses and industries, and nonprofit conservation groups (Wildman and MacBroom, 2000).
The latter includes Trout Unlimited, the Naugatuck River Watershed Association, and The
National Fish and Wildlife Foundation.
Dechannelization
Dechannelization refers to efforts that reverse previous channelization processes. It may
include removing deteriorated retaining walls or linings, replanting the river banks,
reshaping the channel, providing low-flow channels, etc. All these activities require a
thorough evaluation of the rivers hydrology and hydraulics. In addition, the channel
shape and alignment should conform with geomorphic principles. The inner channel
should be proportional to annual peak flow rates, with a supplemental vegetated floodplain to convey excess flows (Brookes, 1988). Figure II.2.3 depicts some of the available
design concepts.
Figure II.2.2
MEANDERING LOW
FLOW CHANNEL
EXISTING
GR
AD
PROPOSED FLOODWAY
SHELF AND TRAIL SYSTEM
NOT TO SCALE
56
DEPTH
E XISTING GRADE
PROPOSED FLOODWAY
SHELF AND WETLAND SYSTEM
TYPICAL FLOODWAY CROSS SECTION
23
DEPTH
MAINTAIN EXISTING WIDTH

OF THE BROOK CHANNEL
30' AVERAGE WIDTH
PROPOSED WETLAND
POND AND ISLAND(S)
EXISTING
VEGETATION
3:1 GRASS
SLOPE
PROPOSED GRADE
3:1 SIDE SLOPES TO

MATCH EXISTING GRADE
EXISTING GRADE
384
PROPOSED PAVED
PEDESTRIAN TRAIL
STEPPED GRASS
FLOODWAY SHELF
EXISTING
VEGETATION
3:1 GRASS
SLOPE
HARBOR BROOK CHANNEL RESTORATION
Remove Culvert, Add

Floodplain and Vegetation
Encourage Overland
Flow
Replace Culvert
w/Wider Open
Bottomed Culvert
De-Channelize
Stream
Encourage
Overland Flow
Grass
Swale
Widen
Bridge
Re-open
Tributaries
Figure II.2.3
Relocate
Stream
Channel
Relocate or
Remove
Structures
Culvert
Remove Liner,
Add Floodplain
and Vegetation
Relocate or
Remove
Structure
Add
Vegetation
Chapter II.2:
Concrete Liner
Remove Dam
STREAM ENHANCEMENT
385
386
Traditionally, manmade channels have relatively straight alignments with compact,

steep-sided cross-sections of a trapezoidal or rectangular shape designed for large flood
flows. They often have high flow velocities that require concrete or stone riprap linings
for erosion protection. As a result, they have low environmental value due to limited
vegetation, poor diversity, lack of in-stream shelter, wide depth and velocity fluctuations,
warm-water temperatures, and minimal aeration. Large flood control channels tend to be
too wide to concentrate dry weather base flows and provide poor sediment transport.
A strong interaction exists between restored rivers and their communities. This
demands that the project team include urban planners and landscape designers with
community involvement in the planning process.
Channel daylighting
In extreme cases, communities have totally enclosed watercourses and placed them underground, typically to reclaim floodplain land for development or roads and to control
flooding. Enclosed watercourses vary from small streams placed in storm drains to larger
rivers that are in conduits or beneath buildings and parking lots. In Hartford, CT, two
miles of the Park River is totally underground as part of flood control efforts. An added
incentive to bury rivers occurred along polluted areas with obnoxious odors or unsanitary
conditions. Enclosing rivers obviously prevents recreation, destroys aquatic habitat, and
disconnects people from water resources (Pinkham, 2000).
Among the benefits to opening up previously enclosed rivers are enlarged habitat,
water quality, hydraulic capacity, recreation, fish passage, boat passage, and reduction of
expensive maintenance of underground structures. Efforts to daylight rivers are often
related to urban redevelopment and the creation of open spaces and linear parks that
emphasize water resources and public access. Daylighting projects requires extensive
community planning participation and a flexible, multidiscipline project team.
The new channel should be designed to provide a viable aquatic habitat and to
replicate natural channels. Opportunities are created for urban parks and linear trail
systems. Although most documented daylighting projects have been small, several larger
urban projects are currently in the planning and design stage.
Riverfront access
Riverfront access in urban areas is often restricted by water-dependent industrial uses or
highways and railroads built along river corridors. Other riverfronts have been isolated
by flood control dikes or walls. As a result, communities are left with little or no access
to rivers that stimulated the communitys early growth. Poor access limits aesthetic opportunities, recreation, and reduces a communitys interest in environmental conditions.
The Cherry Creek project in Denver, San Antonio River Walk, Baltimore Harbor, and
Hartfords Riverfront Recapture are fine examples of adaptive reuse of previously isolated
or underused waterfronts in urban areas. Successful projects combine urban planning and
landscape design to create public space and areas for special events. Urban waterfronts
stimulate economic and recreational activity and provide public access.
The Riverfront Recapture Inc. program was initiated in 1981 to reunite the Connecticut
River with its adjacent communities. Flood control dikes, railroad tracks, and an interstate
highway had isolated downtown Hartford from its waterfront, blocking physical access
and the view of the river. A riverfront master plan was created in 1982, including an
ambitious vision of the future that included greenways, parks, trails, boat launches, excursion boats, an amphitheater, a tall-ships wharf, and a plaza. The riverfront project is
Chapter II.2:
387
credited for stimulating other programs, including a new University of Connecticut football stadium (under construction), and the $500 million multiuse Adrians Landing project
with a hotel in addition to residential and retail space. Recent riverfront activities include
bass tournaments, triathlon races, festivals, fireworks, concerts, and rowing clinics.
Habitat improvements
Rivers that have been channelized or simply subject to increased flooding due to watershed
changes often have degraded aquatic habitat. Specific adverse conditions include reduced
diversity, wide, shallow channels, loss of vegetation, limited shelter, poor water quality,
few benthic species, rapid flow variations, high sediment loads, and excessive velocities.
The U.S. EPA has recognized that water quality improvements alone cannot restore
all streams to the desired swimmable and fishable level without additional habitat and
ecosystem help (EPA, 1996).
Many of these problems can be addressed by physical alteration of the channels shape
and form. Among the available techniques and features available to improve aquatic
habitat are erosion controls, stormwater quality management, variable channel depths,
pools and riffles, flow deflectors, boulder clusters, undercut banks, coarse substrate, and
use of an irregular alignment. Aquatic habitats also benefit from replanted riverbanks that
shade the water, provide cover, and help to attenuate runoff pollutants.
Flow management
Human activities have often resulted in altering the natural river flow rates due to both
land-use activities and the diversion of water for consumptive or irrigation uses. It is not
uncommon for urban watersheds to have higher flood flows and lower base flows than
natural watersheds. Consumptive water uses that reduce river flow include public water
supply, irrigation, and evaporative cooling water. Interbasin water diversions also reduce
the river flows necessary to protect aquatic systems, provide recreation, and assimilate
wastewater. Short-term river flows can be altered by nonconsumptive uses such as hydroelectric plants and industry.
Flow management is an important part of restoring rivers with regulated discharge
rates. There must be sufficient flow to preserve the ecological integrity of riverine systems
and maintain water quality. Flow rates must be sufficient to transport fine sediment, avoid
excessive deposition, and periodically scour coarse substrates. The area and volume of
aquatic habitat are directly proportional to critical flow rates. For most of the country, low
flows occur during the summer and often limit aquatic habitats. In-stream flow needs
vary on a seasonal basis to correspond with the normal life stages of aquatic species. The
management of water quantity and water quality is an integral part in achieving the
federal Clean Water Act goal for all rivers to be both fishable and swimmable.
In some watersheds, human activities may intentionally or inadvertently increase dry
weather streamflow, helping aquatic habitat, recreation, and aesthetics. For example, in
Connecticuts Farmington River watershed, large flood control and water supply reservoirs retain excess wet-weather runoff in the winter and during storms, then slowly release
water in the summer to augment natural runoff. The flow release pattern was developed
via a detailed instream flow study involving the interested parties, led by the National
Park Service, Connecticut Department of Environmental Protection, Metropolitan District
Commission, and Farmington River Watershed Association. The augmented flows
enhance recreational activities including white-water boating, fishing, and aesthetics.
388
Geomorphic considerations
The naturalistic approach being applied in many river restoration projects would not be
possible without key scientific advances over the past 20 years. The evolution of river
restoration sciences is shown on Figure II.2.4. It is recognized that the alignment, slope,
and cross-sectional shape of natural rivers have evolved in response to their flow rates
and sediment loads and that aquatic ecosystems of plants and animals that live in or
adjacent to rivers have adjusted to these conditions. Our improved ability to restore rivers
in recent years is due to the gradual merger of hydraulic engineering and fluvial morphology. These disciplines historically evolved independent of each other along two separate tracks: hydraulic engineering focused on numeric analysis of hydrodynamic systems
and structures, while fluvial morphology developed from geologists observing natural
systems. This allows for increased quantification of river mechanics, which can be applied
to forecast fluvial behavior and responses.
The geomorphic design process is similar for dam removal, daylighting, dechannelization, and habitat improvement projects. It begins with a thorough assessment of existing
hydrologic and geologic conditions, including the characteristics in nearby stable channels.
One then has to identify the preferred natural channel slope, alignment, and cross-section
parameters for the project site. If the preferred natural conditions cannot be provided, then
other variables must be altered (slope, alignment, sediment, etc.) to create alternate conditions.
Conventional hydraulic analysis of open channels defines the flow capacity for a given
size, or conversely determines a channel size for a stipulated flow rate. For any given flow
rate, however, many different combinations of channel width, depth, slope, roughness,
and shape are possible. Consequently, supplemental techniques are necessary to find the
optimal geometric proportions to simulate natural channels.
Several approaches are used to assess stable conditions for unlined channels. The
empirical regime relationships provide general guidance but have been criticized for their
incomplete consideration of physical processes. The hydraulic geometry equations of
stable channel width, depth, and slope are based on plots of empirical field data that are
best used on a regional basis. The theoretical tractive stress approach and sediment transport analysis are available for technical evaluation of stream stability and equilibrium
conditions. They provide guidance on channel slopes versus cross-sectional shapes, but
do not address channel shape or patterns. The author has applied a combination of the
empirical regional regime data and theoretical analysis in the design of channels that
simulate alluvial rivers, in order to minimize the use of rigid channel linings and provide
habitat opportunities.
The first technique consists of locating a stable channel, upstream or downstream of
the subject area, that can be used as a model to determine channel widths, depths, and
slopes for the restored channel segment. It requires that the duplicate channel have similar
discharge rates and geology and that the surrogate channel be in equilibrium. For these
reasons, it is difficult to apply where rapid watershed changes are occurring.
The second technique is based on the empirical field data on channel geometry measurements collected by the U.S. Geological Survey and others beginning in the 1950s
(Leopold, 1994). The studies found that within a specific climatic zone, the width, depth,
and slope of stable natural channels were proportional to their bankfull flow capacity and
that the bankfull capacity was equal to flood flow with an average recurrence frequency
of approximately 1.5 years. Consequently, by determining a channels flow, one could then
determine optimum channel proportions to simulate natural conditions. This is a powerful
tool for use with alluvial channels in sedimentary floodplains but is less valid for degrading or aggrading channels. It is also unsuitable for watersheds undergoing rapid hydrologic changes resulting from urbanization.
EMPIRICAL
SYSTEMS
CLASSIFICATION
FLUVIAL
MORPHOLOGY
GEOMETRY REGIME
RELATIONS METHODS
INTEGRATED
MOBILE BOUNDARIES
FLUVIAL HYDRAULICS
RIPARIAN BUFFERS,
HABITAT MODELS,
IFIM
FISH PASSAGE,
WETLANDS
CONSERVATION,
INSTREAM FLOWS
HABITAT
CONTINUITY,
DISRUPTION
AQUATIC
ECOLOGY
RIVER MANAGEMENT
COMPUTER MODELS
UNSTEADY FLOW,
NON-UNIFORM,
UNIFORM FLOW,
RIGID BOUNDARY
OPEN CHANNELS,
PHYSICAL MODELS
SOIL EROSION,
INCIPIENT MOTION,
STABILITY,
SHEAR STRESS
SEDIMENT
TRANSPORT,
DEPOSITION
HYDRAULICS
SEDIMENT
TMDL,
DE-NITRIFICATION,
DECHLORINATION,
BMP S
T ,
IN ES ES
T
P O RC S T
IN
U A
P O ES
SO . W
C
N R
D
N O OU
IN
S
PHYSICAL
PARAMETERS,
PATHOGEMS
WATER
QUALITY
RIVER RESTORATION
DAM REMOVAL
INSTREAM HABITAT
RIVER DAYLIGHTING
WETLAND MGMT.
1978
SPECIES ACT
ENDANGERED
RIVERS
WILD and SCENIC
CWA 1987
SDWA 1986
WPCA 1972
NEPA 1969
NFIP 1968
WQA 1965
BIG DAMS
IRRIGATION
FLOOD CHANNELS
NAVIGATION
WETLAND FILLING
DREDGING
NATIONAL
ACTIVITIES
Figure II.2.4
WARMING
METEOROLOGY,
GLOBAL
CONTINUOUS
SIMULATION,
WATERSHED
MANAGEMENT,
BMPS
DETENTION
URBAN RUNOFF,
MODELS,
HYDROGRAPH
RELATIONS
STRATIGRAPHY
GEOMORPHOLOGY,
STATIONS,
GAUGING
GEOLOGY
HYDROLOGY
1900
TIME
Chapter II.2:
2000
RIVER MANAGEMENT
Scientific Evolution
389
390
Whereas the first two techniques are empirical, the final approach to channel restoration is based on engineering theories for substrate stability and sediment transport. The
basic concept for providing a channel in equilibrium is that a rivers ability to transport
sediment will be equal to the sediment inflow rate to a given segment. Mathematical
techniques vary from simple allowable velocity and shear stress analysis to one or two
dimensional computer models (Yang, 1996).
Using the previously described techniques, we are learning how to recreate naturalistic
channels and to forecast the impact of future watershed conditions.
Public participation
River restoration has broad public support and is often initiated by local community
groups or nonprofit conservation organizations working in conjunction with government
agencies. Public participation is a key element in creating successful projects that form a
linkage between community needs and our environment.
Public awareness of environmental and community needs plus local knowledge of
restoration opportunities help to generate political support for funding, land acquisition,
and regulatory permitting. In some cases, private funds are raised to prepare master plans
for grant applications and cost estimates, and private funds can be used as a matching
contribution to generate more funds.
Other areas of public interest that can be part of river restoration programs in urban
areas include preservation of historic sites and buildings, environmental education, job
creation, and youth programs.
A typical project flow chart (Figure II.2.5) would include public participation at several
levels. Early participation helps to scope the project and establish local goals and objectives, while later participation helps to define project alternatives and priorities.
Harbor Brook restoration project A river restoration example

Meriden, CT has embarked upon an ambitious program to restore 4 miles of Harbor Brook.
Harbor Brook is a post-industrial river with an urban watershed of 12.3 square miles. It
has been dammed, diked, diverted, polluted, partially filled, and partially enclosed over
the past 200 years. It is prone to frequent damaging floods, chronic bank erosion, and
variable water quality and provides no recreational uses.
The proposed project was stimulated by severe downtown flooding in 1982, 1992, and
1996, causing millions of dollars of damages, plus a strong desire for removal of derelict
riverfront factory buildings and reviving the central city. Although initiated due to flood
control concerns, the project has evolved into a multifaceted river restoration program.
Key components of the $30 million project include:
Acquire open space.

Detain and treat stormwater at four sites.
Replace nine undersized, deteriorated bridges over the river.
Remove seven bridges without replacement.
Open up and daylight two enclosed conduits, creating a downtown park.
Restore open channel, with floodway and instream habitat.
Create a linear park and trail system along the river.
Improve instream habitat with pools, riffles, coarse substrate.
Remove old channel retaining walls, reduce bank slopes.
Reduce bank slopes.
Replant groundcover and overhead canopy trees.
Chapter II.2:
PROJECT FLOW
CHART
391
STAKE OWNERS
INITIATE PROJECT
PROJECT SCOPING
GOALS and OBJECTIVES
PUBLIC
PARTICIPATION
ENVIRONMENTAL
ASSESSMENT
EVALUATE
ALTERNATIVES
PRELIMINARY
DESIGN
FUNDING
REGULATORY
PERMITS
LAND RIGHTS
FINAL DESIGN
PUBLIC
PARTICIPATION
CONSTRUCTION
MONITOR AND
ADJUST
OPERATE AND
MAINTAIN
Figure II.2.5
As part of the project, a partially vacant shopping plaza and parking lot that were
built over the river in the 1960s will be demolished to allow daylighting over half a
mile of enclosed river. The site will become a city park in the downtown area, creating a
village green (Figure II.2.6).
Figure II.2.6
PROPOSED PARK AT MERIDEN HUB SITE
392
Chapter II.2:
393
The upstream stormwater detention sites are being proposed along tributary streams
to reduce peak flows and redistribute runoff, preventing the project from having impact
upon downstream areas. In addition, land-use policies will reemphasize stormwater infiltration and runoff quality controls plus stormwater treatment to improve water quality.
Wetlands will be created in previously filled floodplain areas.
The project planning was guided by an appointed Flood Control Implementation
Agency that included the City Engineer, City Council officials, and members of the public.
The agency coordinated with city planners, the City Planning Commission, Wetlands
Agency, Conservation Commission, nonprofit watershed association, downstream city of
Wallingford, plus state and federal agencies.
Literature cited
ASCE Task Committee, Guidelines for the Retirement of Dams and Hydroelectric Facilities, American
Society of Civil Engineers, New York, 1997.
Brookes, A., Channelized Rivers, John Wiley & Sons, New York, 1988.
Environmental Protection Agency (EPA), Ecological Restoration: A Tool to Manage Stream Quality, Office
of Water, Washington, DC, 1996.
EPA, Stream Channel Restoration: Principals, Processes, and Practices, Federal Interagency Stream Restoration Working Group, Washington, DC, 1998.
Leopold, L.B., A View of the River, Harvard Univ. Press, Cambridge, MA, 1994.
MacBroom, J.G., The River Book, CT Dept. Environ. Protection Bulletin #28, Hartford, CT, 1998.
Maslin, E. and Sicchio, M., Ed., Dam Removal Success Stories, Friends of the Earth American Rivers
and Trout Unlimited, Washington, DC, 1999.
Pinkham, R., Daylighting, New Life for Buried Streams, Rocky Mountain Institute, Old Snowmass, CO,
2000.
Wildman, L. and MacBroom, J.G., Dam removal, a tool for river restoration on the Naugatuck River,
Proc. ASCE Water Resourc. Eng. Conf., Minneapolis, 2000.
Yang, C.T., Sediment Transport, Theory and Practice, McGraw-Hill, New York, 1996.
394
Response
Water quality improvements are not enough
River restoration, James MacBroom considers, is a complex process involving physical
and social sciences directed toward seeking balanced approaches to ecological and human
needs in a sustainable environment. Water quality improvements alone are not enough
to restore rivers due to the host of physical degradations that have been wrought upon
them. Dechannelization, dam removal, daylighting, riparian brownfield cleanup and redevelopment, in-stream habitat creation, and greenway establishment are all factors that
should be addressed in engaging in water sensitive planning along river corridors.
MacBroom believes that the planning and implementation of river restoration projects
will not succeed unless four tenets are met:
1. All aquatic elements in the watershed are considered such as floodplains, wetlands,
and aquifers, in addition to rivers and streams.
2. Public participation is sought at an early stage to form a conceptual linkage between the needs of riparian communities and those of the environment, such that,
through time and the planning process, this will develop into physical linkages
including riverfront access and greenway trail networks, as discussed in Chapter II.3
by Charles Flink.
3. River management professionals require cross training between the traditional
scientific and the planning disciplines while maintaining the skills of the specialist.
4. All participants in the process, from the public to the planning professionals, need
to have, to varying degrees, education and understanding about fluvial processes
and morphology, as discussed in Chapter II.5 by Leslie Zucker, Anne Weekes, Mark
Vian, and Jay Dorsey.
chapter II.3
Greenways as green infrastructure

in the new millennium
Charles A. Flink
Abstract
Greenways are not a new land-use concept in the United States; they have, in fact, been
in existence since the mid-1800s (Little, 1990). The modern American greenway landscape
has taken on a new and important function. Recognized for years as a recreational amenity,
greenways have emerged in the 1990s as a multiobjective land use that can satisfy a variety
of concerns facing urbanizing America. Greenways are now viewed as an important
element of green infrastructure. The term infrastructure is fairly well known throughout the United States and has historically been used to describe things that humans have
created to improve the qualify of life, such as sewers, roads, highways, and utilities.
Green infrastructure identifies the natural elements of the land that also serve to define
the quality of life. The streams, lakes, native vegetation, landforms, animal life, and geologic
formations of an area define a sense of place, community character, and livability of
American towns and cities. Greenways have successfully been used as a method for
cataloging these unique community resources and setting in motion the strategies necessary to preserve, enhance, and manage these resources.
This chapter profiles a variety of greenway projects and programs from throughout
the nation and define how greenways serve as resources for floodplain management, water
quality protection, alternative transportation, recreational resource, and open-space stewardship.
Introduction
As we begin a new millennium, the United States finds itself at a critical juncture point.
More than 100 years of sustained growth and development across a vast, hospitable
continent has resulted in a nation that is the envy of most modern societies. This growth
has been made possible by virtue of a comprehensive and integrated system of infrastructure. When judged by its size, level of service, delivery to even the most remote locations,
and accessibility, infrastructure within the United States has served to define our standard
of living as one of the highest in the world; however, there is a flipside to our manmade
infrastructure superiority. The resulting damage that has occurred to both our natural and
social infrastructure has severely impaired numerous ecological systems and destroyed
the character of our communities.
1-56670-562-2/02/$0.00+$1.50
395
396
Figure II.3.1 Raleigh greenway. This natural trail winds along Crabtree Creek in Raleigh, NC, and
is an example of a greenway corridor that serves multiple benefits as green infrastructure. (Photo
by Charles Flink.)
One of the greatest challenges facing the United States in the next 25 years will be to
conserve, manage, and modernize our nations infrastructure to meet the future needs of
a growing society. In order for this to be accomplished in a realistic and cost-effective
manner, fundamental changes will need to occur in the way that we view the purpose,
function, and operation of infrastructure. As a nation, we need to redefine infrastructure
more broadly to include the critical interrelationships that exist between man and nature.
As part of this redefinition, we need to promote land uses and systems that enable
us to best utilize our limited resources. Our society is in need of exemplary models of
land use that extol the virtues of preservation and conservation, promote energy efficiency,
encourage community action, and provide multiple benefits. Greenways offer us such a
model and can become an integral part of our natural, manmade, and social infrastructure
(Figure II.3.1).
Infrastructure, by definition
According to The American Heritage Dictionary of the English Language (1981), infrastructure
is defined as the basic installations and facilities on which the continued growth of a
community, state, etc. depend, (such) as roads, schools, power plants, transportation and
communication systems. In order to more clearly define the full impact of infrastructure,
and the essential interrelationship between man and nature that impacts infrastructure, I
suggest that three important types or divisions of infrastructure deserve consideration:
manmade, natural, and social.
Manmade infrastructure is the type that most people are familiar with and includes
utility systems, such as electrical generation, manufacturing systems, potable water, graywater treatment, solid waste disposal, stormwater discharge, and telecommunications transmission; transportation systems, such as roadways, highways,
bridges, mass transit service, and all other land- and water-based travel ways.
Natural infrastructure is the naturally occurring physical foundation of the world,
which includes major components of our planets ecological systems, such as soil,
water, air, and vegetation.
Chapter II.3:
Greenways as green infrastructure in the new millennium
397
Social infrastructure consists of the political systems, religious beliefs, moral values,
laws, conduct, and livelihood that we abide by as individuals and that collectively
forms the spiritual foundation of our communities and societies.
If you accept the definition for these three types of infrastructure, then you might also
agree that there has been conflict among these divisions for thousands of years. It is the
relationship between these divisions that ultimately defines mans quality of life and
shapes our common future. It has been proven by Lynch (1962) and McHarg (1969) that
achieving balance and harmony between these divisions offers man an opportunity to
enjoy a higher standard of living. But balance is difficult to achieve even under controlled
conditions.
Gilbert Grosvenor, past president of the National Geographic Society, concluded (1990)
that the forces of man are now of such immense power, (that) they can threaten the very
forces of nature, the very physical forces that make up the Earth itself. Perhaps for the
first time in the history of humankind, it is fully understood that collectively, our species
of 6 billion is capable of tipping the balance between the three divisions of infrastructure.
As worldwide population continues to grow at an exponential rate, the conflict between
these divisions becomes more evident. It is very possible that the worlds entire standard
of living could be dramatically altered and degraded if we fail to change the methods by
which we develop and manage the interrelationships of infrastructure.
Infrastructure in the United States

It is worthwhile to briefly consider the plight of the three types of infrastructure within
the context of the United States, partly because many societies of the world measure their
standard of living by the quality of life Americans have enjoyed for more than 100 years.
To a large degree, the success of this nation has been built around the worlds most
comprehensive, accessible, and integrated system of infrastructure.
Our human-made infrastructure provides us with a network of roadways and highways that span the continent, linking urban and rural areas from coast to coast; fiber optic
lines that transmit voice and data from the most remote areas of New Mexico to the
burgeoning metropolis of New York City in a matter of seconds; water management
systems that reroute the natural drainage of the Rocky Mountains to desert areas of California
and Arizona, where abundant supplies of fruits and vegetables are grown to feed our
population. Our green infrastructure has supported this nations steady growth and development, providing ample timber to build homes, minerals and raw materials to supply
factories and manufacturing centers, abundant fertile soil to grow crops, and a seemingly
endless supply of fresh water to quench our thirst. Our social infrastructure, built on
democratic principles, has supported the land of the free for more than 200 years. We offer
safe haven to the oppressed, religious freedom to the disenfranchised, equal opportunity
to the minority, and the ability for an individual to participate in the governing of our
communities.
Despite these heralded successes, the United States faces daunting infrastructure
problems that threaten our standard of living and economic future. For example, Americans lose more than 1.5 billion hours of productive time each year stuck in traffic jams
because our urban mobility has been severely impaired (Surface Transportation Policy
Project, 1991). Drinking water systems in major metropolitan areas, such as Milwaukee
and New York City, are aging and over capacity, such that the delivery of potable water
to millions of consumers is no longer taken for granted. Thousands of miles of rivers are
classified as degraded and severely threatened by urbanization, poor agricultural practices, and illegal dumping of waste products. The below-cost deforestation of our National
398
Forests not only destroys a significant national resource, but also threatens thousands of
species of plants and animals. Our nations health care system is spiraling out of control,
suffering as much from an unhealthy populous as from the greed of the medical industry.
Civil and ethnic dissatisfaction has degraded the quality of life for inner-city neighborhoods in Los Angeles, Detroit, and Miami. Finally, the American family, pillar of social
structure, has been weakened during the past 40 years by a variety of competing interests,
not the least of which has been a lack of quality-oriented time to enjoy with each other
the pursuits of life.
How did the United States arrive at this juncture?

For the most part, we have developed our human-made infrastructure with little respect
or concern for the natural infrastructure. We have changed the function of entire ecological
systems to lessen the impact of potential natural hazards. For example, expensive flood
control projects have dramatically altered vast, complex upstream and downstream ecosystems, while benefiting only a few local residents. We have used a manufacturing-based
philosophy to exploit natural resources without first understanding the full function of
these resources. As one example, the stormwater management systems of experimental
superfarms in eastern North Carolina significantly altered the salinity of the second-largest
estuary on the continent. Combined with poorly managed fishing practices, this indiscriminate change to the infrastructure of a land-based ecological system has severely
crippled an entire water-based food chain and abundant fishery.
Our human-made infrastructure systems have historically been developed as individual units, with little regard for multiple use of land and water resources. Because we have
always viewed this continent as bountiful and capable of supporting individual pursuits,
little effort has been made to coordinate the development of transcontinental, regional, or
local infrastructure systems. These systems are now so vast, uncoordinated, and old that
it takes supercomputers to manage their operation and delivery. Today, minor problems
in these systems now cause major interruptions of service to increasing numbers of
consumers.
Furthermore, we have historically destroyed the social fabric of local communities in
order to satisfy straight-line, high-speed, human-made infrastructure systems. This has
been accomplished under the heading of progress. In other cases, to improve the efficiency
of an infrastructure system, we have rerouted its path of travel around areas of congestion,
eradicating the unique charm and indigenous qualities of local communities; thus, we
have created a monotonous landscape whose character can be found repeatedly within
every town and city in the nation.
As the American landscape continues to change from rural to urban, we realize that
land uses have been encouraged during the past 50 years that are not compatible with an
optimal and efficiently managed infrastructure system. Looking to the future, we are
forced to encourage the redevelopment of what is now low-intensity land-use areas into
higher densities, in order to make more efficient use of existing infrastructure.
Additionally, in a capitalistic society, the following question must always be asked:
How much will it cost to modernize and maintain our existing human-made infrastructure,
as well as develop new systems to meet the demands of urbanizing America? Some
infrastructure specialists place the cost somewhere in the hundreds of billions of dollars,
a price tag that may be out of reach, irrespective of a healthy world economy.
Have we reached the point of no return? Are there opportunities to change course? I
suggest that one emerging land-use concept offers Americans hope for our future. Greenways, often regarded as recreation based corridors or linear parks, is a land use concept
that has been evolving during the past 20 years and is presently being implemented to
Chapter II.3:
399
resolve infrastructure-related problems in several communities throughout the United

States.
Greenways as infrastructure
Greenways are multiple-use corridors that respect the inherent qualities of natural systems
and accommodate manmade systems in a way that is compatible with nature. Greenways
conserve open space, protecting riverine systems, vegetation, native soil, and geomorphology that is essential to the health and vitality of the local landscape. They soften patterns
of urban growth, offering green vegetated buffers between land uses that can effectively
cool the air mass, deflect and absorb objectionable noises, and block unsightly views (Flink
and Searns, 1993).
Well-planned greenway systems have been proven to offer efficient means of alternative transportation for community residents. An integrated trail system provides choice
in the mode of travel between popular points of origin and destination. Greenways offer
a way to protect our nations natural and cultural heritage, promoting eco-tourism, protecting historic settlements, buildings and landscapes, and offering opportunity for selfinterpretation and education. They have also been shown to promote economic development by increasing the value of adjacent property, offering improved access to community
resources, and improving the quality of life for local residents.
By protecting the basic foundation of natural infrastructure, vegetation within greenways can cleanse the air of pollution, and absorb overland runoff to minimize pollution
to our creeks, streams, and lakes. They can mitigate the effects of urban flooding by
protecting flood-prone areas from encroachment and development. Greenways serve as
important corridors for wildlife, protecting habitat, offering sanctuary during migration,
and promoting biological diversity.
Successful greenways promote goodwill throughout a community. They offer access
to landscape resources close to where people live and work. They provide an outlet for
community activism and have become the new main street for many towns and cities
throughout the United States.
Greenways offer communities with an opportunity to establish infrastructure that is
based on the local abilities of ecological systems. This linear system, when planned and
developed properly, provides community residents with safe, efficient access and service
to a variety of natural resources, much like a roadway network, water supply lines, or
electrical system. Under ideal circumstances, communities would plan for greenways in
the same manner.
Greenways have been shown to have a positive impact on the different types of
infrastructure. A few specific examples are described next.
Transportation
Traditionally, our transportation system, a component of human-made infrastructure, has
been measured by mode of travel, number of highway miles, automobile carrying capacity,
and speed of travel. If transportation systems were instead measured on the basis of their
impact on nonrenewable resources, air quality, energy efficiency, and urban mobility, then
the resulting end products the travel way, route, and location would have a different
look and function.
Greenways offer an opportunity to transform traditional transportation design, development, and management. They do so first by accepting the fact that mobility is of prime
importance in defining urban transportation systems and by offering a choice in the mode
of travel. Second, energy efficiency is absolutely critical to our future survival. George
400
Figure II.3.2 Greenway cyclists. A grandfather and grandson enjoy the Trolley Line Greenway in
downtown Kansas City, MO. Thousands of miles of greenway trails serve the transportation needs
of Americans daily. (Photo by Charles Flink.)
Woodwell, director of the Woods Hole Research Center in Woods Hole, MA, observed
that from the perspective of a biologist, it appears urgent to recognize that the era of
fossil fuel has passed (Hay, 1990). Yet, as we evaluate our nations current transportation
system, we realize that nearly 50% of all oil consumed is for automobile travel (Surface
Transportation Policy Project, 1991).
In fact, greenways are a reasonable choice of travel for trips under 5 miles and might
be the preferred choice for many trips under 2 miles if the infrastructure were in place
today (Flink and Musser, 1992) (Figure II.3.2). Several communities are enjoying the benefit
of having comprehensive and interconnected greenways as one component of the local
transportation network. Seattle, WA, Portland, OR, Davis, CA, Denver, CO, Madison, WI,
and Raleigh, NC, have come to rely on their greenway systems as one method for providing a choice in travel. Offering a choice in travel may be the most effective method for
reducing traffic congestion, improving air quality, and promoting a more healthy lifestyle.
The Transportation Efficiency Act for the Twenty First Century (TEA-21) emphasizes
multimodalism as the future goal for our nations transportation infrastructure. Because
greenways are built to be fully accessible to all persons regardless of ability, they satisfy
the concern regarding mobility. Greenways offer nonpolluting, nonoil-consumptive, clean,
and green travel ways throughout our communities.
Water quality management

Protecting our potable water supply is dependent on mitigating the effects of nonpoint
source pollution which in urban areas is discharged through our stormwater management systems. For the past 75 years, stormwater management has largely been associated
with street system drainage and the detention of post-development exit flows from intensive land-use development. The solutions to these drainage concerns have been based on
calculating how much water is capable of flowing through a circular pipe otherwise
defined as hard engineering. As part of revisions to the Clean Water Act in 1990, the
Chapter II.3:
401
Figure II.3.3 Denver flood channel/greenway, Denver, CO, has transformed several of its urban
streams into multiobjective stormwater management, transportation, and recreational corridors.
(Photo by Charles Flink.)
EPAs new national water quality program is aimed at rethinking the hard engineering
approach to solving stormwater management. Promoting a soft engineering approach,
EPAs Nonpoint Discharge Elimination System guidelines radically change the way exit
flows are discharged and accommodated within downstream sites (Flink and Musser,
1993). In the future, stormwater will have to be absorbed within floodplain lands, or
through Best Management Practices, rather than being discharged directly into flowing
streams (Figure II.3.3).
Greenways as a component of stormwater management infrastructure provide a mechanism for protecting floodplain lands and establish areas where floodwaters can be
absorbed naturally. Denvers South Platte River Greenway system and its tributary
streams are an early example of the soft engineering approach to stormwater management.
The greenway is an inexpensive alternative to a proposed $600 million concrete-lined
channel flood control project originally offered by the Army Corps of Engineers. The
greenway project cost $18 million to construct, resulting in a savings of $582 million from
the original Corps proposal. It not only provides flood protection for lands along the South
Platte River and its tributaries, but also an interconnected 40-mile-long system of multiuse
trails that serve as access points for maintenance of the riverine system. The greenway
provides residents with access to the flowing waters of the Platte and safeguards the citys
urban core from flooding. In addition, Denver earns bonus points from the Federal Emergency Management Agency, National Flood Insurance Program, under its Community
Rating System, which awards communities with lower insurance rates for protection of
floodplain lands.
Other communities are following Denvers lead and beginning to develop comprehensive greenway systems where the primary function is flood control and water quality
management. The LouisvilleJefferson County Metropolitan Sewer District is one such
community where a comprehensive effort is underway to revise the traditional approach
to channelization of natural streams. The community expects to implement a multiobjective approach to stream corridor management, which will serve to protect natural habitat,
improve the capacity of native streams, and offer an interconnected system of off-road
trails for recreation and alternative transportation.
402
Figure II.3.4 The Swift Creek Recycled Greenway, Cary, NC, is a publicprivate partnership project
that demonstrates the creative use of waste materials and trash in the construction of public-use
facilities. (Photo by Charles Flink.)
Recycling of solid waste

The town of Cary, NC, became one of the first communities in the nation to develop a
greenway facility using the principles and products of recycling. Faced with a diminishing capacity in the local landfill, and a successful community-wide recycling program,
the town accepted an invitation from Greenways Incorporated, a local environmental
and landscape architecture consulting firm, to develop the Swift Creek Greenway using
recycled materials. A unique publicprivate partnership was established between the
town and the firm, which included more than 22 other public and private-sector organizations. A one-mile pilot project was envisioned that would demonstrate how typical
household solid waste could be reused to create essential facilities for the greenway,
including bridges, benches, trash receptacles, signage, and safety railing. The hardsurfaced trail was developed using bottom ash from a local coal-fired electrical generation plant. Recycled asphalt and shredded rubber tires were used to develop the surface
of the trail tread (Figure II.3.4).
The greenway demonstrates that trash and debris can no longer be thought of as
disposable items. Instead, they are new resources for communities all across the United
States who face similar problems regarding waste management. Further, a slogan developed for the project, Buy Recycled. Its Second Nature, offers a needed boost to the
recycling industry, which has suffered in recent years due to a lack of appropriate product
marketing. The project clearly demonstrates that recycled products can achieve equal or
greater performance than similar products made from natural resources. In addition to
the stated project goals, the greenway has an environmental education component and
serves to link local neighborhoods with a large office park complex, providing an efficient
alternative route for travel from home to work. The project received a Take Pride in
America Award for its accomplishments and is serving to enhance Carys reputation as a
progressive and livable community.
Chapter II.3:
403
Figure II.3.5 The Stones River Greenway, Murfreesboro, TN, is an example of a greenway that offers
educational opportunities. The greenway trail links together Civil War battlefield sites and documents one of the most important battles in American history. (Photo by Charles Flink.)
Environmental education
The city of Murfreesboro, TN, has a storied if not dubious place in American history. Site
of the one of the bloodiest battles in the Civil War, the town has struggled with its identity
and future as a bedroom community for metropolitan Nashville. With assistance from
Congressman Bart Gordon and the National Park Service, the community decided to link
together the resources associated with the Stones River National Battlefield and develop
a multipurpose greenway that will highlight the communitys unique history. One of the
primary motivations behind the project is to increase tourism and economic development
for the community through the development of a riverside greenway. Education is the
other principal pursuit of the project (Figure II.3.5).
Community leaders developed a series of wayside exhibits that provide greenway
users with opportunities to learn more about the Battle for Stones River. Additionally, the
local landscape has been interpreted so that residents and visitors can better understand
the unique natural features of Stones River and middle Tennessee. The project links the
downtown area to suburban neighborhoods and has become the first off-road bicycle
facility within the community, offering an alternative route of transportation. Local greenway advocates firmly believe that the greenway has improved the quality of life and
standard of living for residents of the community.
Improving social conditions

In the classic work Design with Nature, Ian McHarg states (1969), We need not only a
better view of man and nature, but a working method by which the least of us can ensure
404
Figure II.3.6 Greenways can be a deterrent to crime. In downtown Nashville, TN, the Shelby Safewalk has helped transform an urban neighborhood. Similar projects can now be found in Chattanooga, TN, and Miami, FL. (Photo by Charles Flink.)
that the product of his works is not more despoliation. Several greenway projects
throughout the United States have served to reverse poor land-use practices and enhance
local community pride. The BrooklynQueens Greenway is a remarkable effort to remove
concrete and asphalt in downtown New York City so that trees can be installed, and urban
residents will be encouraged to walk and bike from neighborhood to neighborhood. In
Nashville, TN, and Greensboro, NC, two economically depressed neighborhoods have
embraced the greenway concept as a way of restoring social order and quality of life to
degraded urban conditions. Devoid of significant natural features, these communities have
rallied around the concept of multiple-use greenways as a method for instilling pride and
ownership in the neighborhood landscape (Figure II.3.6).
The typical urban trappings concrete sidewalks, telephone poles, overhead utility
wires, vacant lots, and abused vegetation have been reorganized and replaced with
decorative paving, human-scale street lights, trees, pocket parks, and flowering plants.
Project implementation involves local participation in the demolition and construction of
the greenway. From this involvement, local residents have become stakeholders in the
future success of the greenway and will hopefully pass along their pride and accomplishment to future residents of the community. In addition, restoration of the designated
greenway corridor offers local residents with a viable choice in travel and improves the
efficiency of utility systems throughout the community.
Conclusion
The future development of our infrastructure will determine to the greatest extent the
impact that civilization has on our planets natural environment. We have clear and
relatively easy choices to make. We can choose to modify the way in which we define,
plan, develop, and manage our infrastructure, or we can continue to promote infrastructure development that will eventually degrade our social and economic future and destroy
the ecological systems necessary to sustain life on the planet.
Chapter II.3:
405
Since the release of the Presidents Commission on Americans Outdoors Report in

1987, our nation has narrowly defined greenways principally as a recreational amenity
for our communities (Americans Outdoors, 1987). I suggest that the most significant
attribute of greenways is not their recreational value, but their potential as a land use that
successfully accommodates multiple environmental and social functions. As a component
of infrastructure, greenways can provide our communities with a land resource that
successfully balances the needs of man with the abilities of the nature.
As we face the next 25, 50, and even 100 years, it is very appropriate that we take
time to rethink traditional forms of land use, especially with regard to infrastructure
development. I believe that multiple use of linear land corridors will be one key to
accommodating our future physical growth. At the heart of this multiple-use concept, I
feel that greenways will serve as the foundation within which other infrastructure components can and should be developed.
Literature cited
Acting in the national interest: The transportation agenda, Surface Transportation Policy Project,
Americans Outdoors: The Legacy, The Challenge, The Report of the Presidents Commission, Island Press,
Flink, C. and Searns, R., Greenways, a Practical Guide to Planning, Design and Management, Island
Press, Washington, DC, 1993.
Flink, C. and Musser, T., Current planning guidelines and design standards being used by state and
local agencies for bicycle and pedestrian facilities, working paper for the U.S. Dept. Trans.,
Fed. Hwy. Admin., Washington, DC, 1992.
Flink, C. and Musser, T., A nationwide inventory of stream corridor planning programs, working
paper for the LouisvilleJefferson County metropolitan sewer district, Louisville, KY, 1993.
Grove, N., Greenways paths to the future, National Geographic, 177, 76, 1990.
Hay, K.G., Using natural gas rights-of-way as greenways: A feasibility study, The Conservation
Fund, Washington, DC, 1990.
Little, C.E., Greenways for America, The Johns Hopkins Univ. Press, Baltimore, MD, 1990.
Lynch, K., Site Planning, MIT, Cambridge, MA, 1962.
McHarg, I.L., Design with Nature, Natural History Press, Garden City, NY, 1969.
406
Response
Corridors that integrate natural, societal, and social elements
As Charles Flink discusses in this chapter, water sensitive planning should transcend
consideration of mere physical spaces. In particular, greenways can very much be utilized
to help define and instill a quality of life at the same time as shaping a sustainable future.
This is similar to the broad aspirations characterizing comprehensive, regional planning
as Daniel Williams describes in Chapter II.11.
Greenways, once regarded as only recreational amenities, are now recognized to
provide many more benefits to communities, functioning as sort of green main streets.
Such systems, Flink argues, represent models for integrating human and environmental
objectives in land-use planning that can come together in concepts of open-space stewardship. Not only does modern greenway planning address the capability of these linear
parks to ably serve their original resource purpose for humans in terms of tourism,
protection of historic settlements, alternative transportation areas, and increased nearby
property values, such planning also considers these river borders with respect to their
roles in floodplain management, water quality protection, and wildlife habitat preservation, as discussed elsewhere in this book (Frank Mitchell in Chapter II.1; James MacBroom
in Chapter II.2; and Leslie Zucker, Anne Weekes, Mark Vian, and Jay Dorsey in Chapter II.5).
chapter II.4
Natural resource stewardship planning

and design: Fresh Pond Reservation
(Massachusetts)
Thomas S. Benjamin
Abstract
Fresh Pond Reservation represents one of metropolitan Bostons premier public open spaces.
The reservation surrounds Fresh Pond Reservoir, the terminus of the city of Cambridges
water supply. As such, Cambridge places a high value on protecting the reservoirs water
quality. Originally designed by Olmsted and Eliot in the 1890s, the reservation is now aging
under heavy recreational use, with erosion from hillsides and pathways threatening the
reservoirs water quality. Further, ecological values, public safety, and aesthetics are now also
threatened as diverse vegetation types and zones become overrun by a few invasive vine
and shrub species, their proliferation related to ongoing soil disturbance.
The Fresh Pond Reservation master plan (adopted January 2001) provides long-term
guidance for resource management and enhancement as the reservation moves into its
second century. The plans goals are now being realized through a number of projects,
the most significant of which is a comprehensive Landscape Management Plan to control
soil erosion and invasive species. Two other major infrastructure projects currently under
construction at the also reflect the goals of the plan: (1) Cambridges new water purification
facility; and (2) a sewer separation/bikeway project. Additional projects to manifest the
plans vision include a master plan for landscape and drainage improvements at the
reservations golf course, landscape improvements surrounding a new assisted-living
facility on the premises, and possible naturalization of a former soccer field currently in
use as a construction staging area. The last two projects have been incorporated into one
Northeast Sector Landscape Improvements Project covering approximately 25 acres at
Fresh Pond. Finally, two pilot projects have emerged from the plan as well the Perimeter
Road Alternative Paving Pilot Study and the Reservoir Fencing Alternatives Pilot Study.
Introduction
Many of our nations parklands and prized urban open spaces face a common challenge
how to balance water quality and natural resource protection with the diverse needs of
the public. This challenge is being faced internationally, as well, as population pressures
increase on ever more scarce open space and water resources. Fresh Pond Reservation in
1-56670-562-2/02/$0.00+$1.50
407
408
Cambridge, MA, is no exception. Here, the connection between water quality protection
and wise stewardship of the landscape in which water resources reside, or the natural
interaction between land and water resources, has been broadly realized in a major forward-looking planning effort that is now being implemented. With a little luck, the anticipated success of this effort will provide an international model for natural resource
stewardship planning and design.
After a century of increasingly heavy recreational use, the 162-acre Fresh Pond Reservation is showing its age, especially with soil compaction and erosion from hillsides.
Invasive plant species are overrunning natural areas. Moreover, shoreline erosion is contributing to the degradation of numerous water bodies on the reservation, including Fresh
Pond Reservoir the terminus of the city of Cambridges water supply. Mitigating these
environmental stressors is critical if Fresh Pond is to remain a natural haven in the city
for a second century (Figure II.4.1).
Thankfully, the city of Cambridge and its residents have recognized the need for a
long-term natural resource management program for Fresh Pond. They are moving forward with an ambitious master plan (Fresh Pond Reservation master plan) to preserve
the water quality of Fresh Pond Reservoir while implementing much-needed landscape
enhancements. This master plan is a model for water sensitive landscape design and
restoration. Its lessons are poignant and applicable to the preservation of our countrys
urban parks for generations to come (Figure II.4.2).
History
Fresh Pond Reservation is one of metropolitan Bostons premier public open spaces.
Originally designed by the Olmsted firm and Charles Eliot in the 1890s, the 162-acre
reservation has long served as a natural haven in the city. It is an interesting and complex
ecosystem, loved no less fervently by its diverse users than Central Park by New Yorkers.
The reservations natural features are a challenge to manage: wetlands and upland
woods; stands of large evergreen and deciduous trees; shrub borders; steep hillsides;
ravines and small ponds; seasonal marshes, meadows, grassy areas, dense thickets, irregular vegetated shorelines; and various soil types and hydrologic conditions. Adding to
this complexity is the fact that the reservation serves as a protective buffer for the 163-acre
Fresh Pond Reservoir, the terminus of the city of Cambridges water supply (Figure II.4.3).
Throughout much of the reservations history, municipal management and maintenance have been minimal. City of Cambridge residents, long characterized by their strong
tendency toward activism, have taken the initiative to preserve the reservations natural
resources for more than four decades. In 1997, however, the city of Cambridge and the
Fresh Pond Master Plan Advisory Committee (FPMPAC) joined forces to develop a natural
resources stewardship plan (Fresh Pond Reservation Natural Resource Stewardship Plan)
one focused primarily on watershed management (Benjamin, 1999).
Natural resource inventory and stewardship plan

The resulting stewardship plan process for Fresh Pond consisted of two major components:
the natural resource inventory (Fresh Pond Reservation Natural Resource Inventory) and
the plan itself. The inventory phase comprehensively investigated and recorded natural
resource conditions at the reservation and around the reservoir. This systematic, ecological
study of the reservations open spaces focused on water quality issues and the identification of surface conditions such as erosion and compaction, vegetation cover types, and
shoreline and stream conditions. It also identified many areas of notable natural or cultural
character, including healthy stands of old trees and historic water treatment structures.
Glacken
Field
Weir
Meadow
Kingsley
Park
Water
Treatment
Plant
Lusitania
Field
Con
cord
Ave
nue
Neville
Manor
Pa
r
kw
ay
W
S
Scale: 1" = 2000'

Scale Is Approximate
Watershed Boundary
Site Features
Stewardship Plan
Fresh Pond Reservation
Source: City of Cambridge GIS Aerial

Photograph: Dated; March 29, 1995
Figure II.4.1 Fresh Pond Reservation site features, March 29, 1995. (Source: city of Cambridge GIS Aerial).
Hur
on
Ave
nue
Golf
Course
Little
Fresh
Pond
Golf
Course
Blacks
Nook
Fr
es
h
Chapter II.4:
Po
nd
North
Pond
Natural resource stewardship planning and design: Fresh Pond Reservation 409
410
FRESH POND RESERVATION
MASTER PLAN
CAMBRIDGE, MASSACHUSETTS
Figure II.4.2 Fresh Pond Reservation master plan, prepared by the Fresh Pond Master Plan Advisory
Committee, adopted 2001.
Stewardship plan recommendations

With strong community support, the inventory findings and the stewardship plan recommendations for Fresh Pond were accepted in early 1999. The plan provided recommendations
for general topics reservation-wide, such as slope stabilization, wetland management, and
trail management. It also provided extensive recommendations for 11 subareas of the reservation including several water bodies and a municipal golf course. As a reflection of the
reservations diverse needs, the recommendations themselves ranged widely in character:
Erosion control and shoreline/upland stabilization

Reduction of invasive plant species
Enhancement of meadow areas
Control of trail access
Improvements to trail surface and drainage
Reuse of on-site materials, such as downed trees, for trail resurfacing
Composting of leaves for widespread reuse in soil improvement efforts
In January 2001, the Cambridge City Council took these recommendations to heart. It
gave the green light for work to proceed on seven water sensitive design projects at the
reservation, as part of the comprehensive Fresh Pond Reservation master plan into which
the stewardship plans recommendations were incorporated (Figures II.4.4, II.4.5, and II.4.6).
A key player in taking the stewardship plan and, later, the master plan from conception
to reality has been the city of Cambridges Water Department. In particular, the citys
Figure II.4.3 Original plan for Fresh Pond by the Olmsted Firm and Charles Eliot, (1897).
Chapter II.4:
Figure II.4.4 Fresh Pond vegetation types map prepared in geographic information systems (GIS) format for the Fresh Pond Reservation Natural Resource
Inventory. (Source: Rizzo Associates, Inc., 1998.)
Source: City of Cambridge GIS Aerial

Photograph: Dated; March 29, 1995
Scale: 1" = 470'

Scale Is Approximate
412
Softwood Forest
Hardwood Forest
Mixed Forest
Scrub/Shrub Upland
Meadow
Forested Wetland
Scrub/Shrub Wetland
Emergent Wetland
Landscaped/Maintained Land
Developed Land
Legend
Vegetative
Cover Types
Fresh Pond Reservation

Natural Resourse Inventory
Chapter II.4:
Table 1. Developed Land, Open Water and

Vegetated Land in Fresh Pond Reservation
Table 2. Vegetative Cover Types

at Fresh Pond Reservation
Major Land Cover Type
Vegetative Cover Type
Developed Land
Open Water
Fresh Pond
Little Fresh Pond
Blacks Nook
North Pond
Vegetated Land (see Table 2)
Total Reservation Area
Area (acres)
10.3
162.6
156.6
3.5
2.0
0.6
152.0
324.9
Upland Forest
Softwood forest
Hardwood forest
Mixed forest
Scrub/Shrub Upland
Meadow/Open Field
Wetland
Forested wetland
Scrub/shrub wetland
Emergent wetland
Landscaped/Maintained
Golf course
Other
Total Vegetated Land
Area (acres)
52.1
8.1
38.5
5.5
2.1
0.5
21.4
17.4
3.3
0.7
75.9
50.4
25.5
152.0
Source: Fresh Pond Natural Resource Inventory,

Table 1
Figure II.4.5 Tables from the Inventory.
watershed manager, Chip Norton, has been a vital link in coordinating the multiple
interests represented by the committee and others with the citys ongoing work, especially
with regards to large, preexisting infrastructure improvement projects. With both a strong
understanding of the larger context in which the reservation and reservoir sits and the
citys fiscal constraints and day-to-day operational needs, Mr. Norton has helped mold
the master plans grand vision into tangible and feasible capital projects. It has been crucial
to the process to have a proactive public official located on-site who possesses the broad
view of how all the pieces fit together and who also has the ability, and level of commitment, to get things done in the city.
Master plan implementation

Of the seven projects currently under way or about to begin that incorporate Fresh Pond
master plan goals, the Landscape Management and Maintenance Plan represents the first
step toward manifesting the goals of the natural resource stewardship plan: to preserve
water quality through improvement of soil conditions; and to increase biodiversity and
overall aesthetic interest through vegetation management. The Landscape Management
Plans scope exemplifies and mirrors the many recent and current efforts elsewhere in the
emerging, multidisciplinary field of restoration ecology. The key components of this project
are:
Vegetation management through control of invasive tree, shrub, vine, and herbaceous species to enhance biodiversity and native species growth
Erosion control and stabilization of the most severely eroded and highly visible
upland slopes and nonreservoir shoreline section(s) to protect the reservoir from
silt-laden surface runoff
Priority areas have been selected based on the severity and nature of invasive plants
and/or erosion conditions, the visibility to the public, the proximity to other preexisting
construction projects and their impacts (e.g., recently disturbed soils), potential ecological
value (e.g., edge zones considered highly valuable), and the size and manageability of the
.
N ST
GROSVEN OR
RD.
BETTS RD.
and Fence
Perimeter Road
IN
.
T
UN
LW
IN
TH
TA
OR
Pond
Fresh
FO
HO
Little
TE
RR
S T.
R
O
N
A
P
A
V
V
A
N
C
O
R
D
.
L
P
I
L
L
E
IN
P
IS
R
G
D
L
O
T
.
S
Reservoir
Fresh Pond
North Pond
Nook
Blacks
ES
PO
T
L
.
T
PA
Weir Meadow
FR
IT
and Fence
Perimeter Road
AVE.
.
T
Facility
Purification
New Water
Y
A
W
A
F
Field
Lusitania
Manor/
Neville
.
T
W
A
P
R
Y
A
W
E
K
IF
Berm
RESERVATION
FRESH POND
PROJECTS AT
FUTURE
CURRENT AND
KEY MAP TO
Title:
New Bikeway/
MANAGEMENT AREAS
LANDSCAPE
Key:
Figure II.4.6 Key map to current and anticipated capital projects at Fresh Pond Reservation. (Source: Rizzo Associates, 2001.)
Golf Course
Municipal
ING TO
LIVERMORE RD.
WA SH
T.
S
E
V
O
R
G
.
E
A
V
E
B
A
N
E
E
D
.
DR.
E
LARCHWOOD
V
A
RD
N
E
O
V
A
RC
LA
LEXIN GTON
LN
AR
SS
C
O
VA
T
.
IG
414
A
V
Chapter II.4:
Figure II.4.7 Oriental bittersweet (Celastrus orbiculatus) vine, an invasive species, choking a hardwood forest and edge zone at Fresh Pond.
unit. An area in excess of 25 acres has been proposed for vegetation management, slope
stabilization, or both. This represents approximately half of the reservations natural
area, typically woodlands or overgrowing old fields (Figure II.4.7).
Landscape Management Plan vegetation and soils management

In terms of the level of effort, complexity, and cost, the vegetation management component
is the larger piece of the Landscape Management and Maintenance Plan. Its primary
objective is to control invasive species in high-priority areas so that biodiversity with
native species may be increased. These areas include forested wetland and upland areas,
as well as meadows and edge zones.
Methods under consideration for invasive species control include both mechanical and
chemical techniques. The mechanical techniques being considered are cutting, hand-pulling,
mowing, smothering, and prescribed burning or spot burning. Due to their reliance on
abundant labor, mechanical techniques may only being considered for limited areas, such as
highly environmentally sensitive areas near the reservoir and open-water bodies; however,
mechanical or manual removal techniques will be emphasized to the greatest possible extent.
For larger, less sensitive areas, environmentally sound chemical techniques, specifically
the use of herbicides, may represent the most feasible and cost-effective method for controlling
invasives. This approach is endorsed by leading experts in invasives control. In fact, a strong
argument in favor of chemical treatment over mechanical methods is that mechanical removal
of established plants is more disturbing to soils. This is due to the trampling by workers
engaged in the act of pulling plants out of the ground, especially where invasive stands are
dense (e.g., buckthorn in Lusitania Woods) (Figures II.4.8, II.4.9, and II.4.10). Further, state
and federal level environmental regulations strongly discourage soil disturbance of any kind
within wetlands. Approximately half of the area currently proposed for landscape management work at Fresh Pond falls within regulated wetland resource areas.
Environmentally sound chemical techniques emphasize direct herbicide application
into the individual plant stems or trunks to be treated, rather than area or blanket spraying.
These techniques include injection of herbicide into invasive tree species (e.g., tree-of-heaven,
416
Figure II.4.8 Buckthorn (Rhamnus spp.), an invasive shrub species that thrives on disturbed soils,
dominates as much as 95% of the forest understory in parts of Fresh Pond Reservation.
Figure II.4.9 Mechanical removal of invasive species along Fresh Pond Reservoirs shoreline. The
Fresh Pond Reservation Landscape Management Contract currently being prepared will address
invasive species control.
Glossy Buckthorn
Gray Dogwood
Arrowwood
Sheeps Laurel
Swamp Azalea
Highbush Blueberry
OPEN
Rhaninus frangula
Cornus racentosa
Viburnum denatum
Kalmia angustifolia
Azalea viscosum
Vaccinium corymbosum
Vincetoxicum nigrum
Celastrus scandens
Athyrium felix-femina
Arctostaphylos uva-ursi
Polygonatum biflorum
Shrub
Shrub
Shrub
Shrub
Shrub
Shrub
Shrub
Groundcover
Groundcover
Groundcover
Groundcover
Groundcover
Groundcover
5
0
10
10
15
40
10
50
10
10
5
10
5
35
20
15
20
10
25
5
0
0
0
70
90
10
0
0
0
0
0
30
20
10
10
30
20
5
+10
+10
+15
30
80
+40
+10
+10
+5
+10
+15
+5
0
+5
+10
20
x
x
P
V
P
V
V
P
P
Figure II.4.10 Vegetation management monitoring form. Monitoring will be an integral component to landscape management strategies being planned for
the reservation.
F
P
E
E
G
E
E
G
G
G
G
F
G
G
Percent
Percent (%)
Percent (%) Change
Vegetation Health
(%)
Cover At
() Since
Planted (P)/
(Poor, Fair,
Cover Last Monitoring
Last Monitoring
Invasive Volunteer (V) Good, Excellent)
Comments: Fairly dense native groundcover developing throughout plot area; remaining buckthorn appears stressed; tree canopy almost fully covering area
Black Swallow-wort
Oriental Bittersweet
Lady Fern
Bearberry
Solomons Seal
OPEN/BARE
Silver Maple
American Elm
Green Ash
River Birch
OPEN
Common Name
Acer saccharinum
Ulmus americanus
Fraxinus pennsylvanica
Betula nigra
Botanical Name
Tree
Tree
Tree
Tree
Tree
Vegetation
Layer
Chapter II.4:
Monitoring Data
Monitoring Location: Water Purification Facility

Monitoring Station Number: 1
Monitors Name: Jane Doe
Monitoring Date: 1 June 2003
Date of Last Monitoring: 27 October 1999
Fresh Pond Vegetation Management Project

[Sample] Monitoring Form
418
Norway maple) and cutting and dabbing herbicide onto the stumps of invasive shrubs
and vines (e.g., buckthorn, oriental bittersweet) called the cut-stump method. The only
exception to these direct and highly controlled applications would be to herbaceous
invasives (e.g., garlic mustard), which can only be chemically controlled through foliar
spraying. Generally speaking, herbaceous invasives have not yet become a major problem
at Fresh Pond.
Two widely available general-use herbicides that have been endorsed by numerous
experts in natural areas management include Round-Up (registered trademark of Monsanto Company, St. Louis, MO) (active ingredient: glyphosate) and Garlon (registered
trademark of The Dow AgroSciences Company, Indianapolis, IN) (active ingredient: triclopyr). Both products have been used extensively around the country for invasives
control. Round-Up has been particularly endorsed for its rapid chemical breakdown after
application, although Round-Up is a nonspecific herbicide that, if improperly used or
accidentally overthrown, can kill both woody and herbaceous species. By contrast, Garlon
can have a longer chemical breakdown time but is specific to woody species only and will
not harm herbaceous species if spilled or overthrown. These chemicals have been successfully used to control invasives in Massachusetts by the New England Wildflower Society,
Trustees for Reservations, and the Middlesex County (MA) Mosquito Control Unit, among
others. Nationally, controlled herbicide use has proven effective in combatting invasives to
restore native species diversity by the Nature Conservancy, Minnesota Department of Natural
Resources in state parks, the McHenry County (IL) Conservation District, and the University
of Wisconsins Curtis Arboretum, among others (http://www.newfs.org/ publication.html,
http://tncweeds.ucdavis.edu/handbook.html, http://www.nps.gov/plants/alien/).
Following chemical or mechanical treatment, selected areas will receive soil amendments
(mechanically treated areas only) and/or seeding with a noninvasive or less invasive cover
species such as annual ryegrass, sheep fescue, white clover, buckwheat, and/or switchgrass. The cover species will serve a number of purposes, including:
Immediate soil cover providing erosion control and protection from trampling
impacts in newly opened areas
Immediate competition for invasive species germination
Initial decompacting of compacted soils
Organic mulch and nutrient source for impoverished soils
Improved stormwater runoff uptake
Improved aesthetics for areas with extensive vegetation removal
The Management Plan will provide the first phase of natural resource improvements
by focusing on invasives removal and selective thinning of the tree canopy (in forested
sections) to increase light penetration to lower forest strata. Restoration plans will be
prepared and executed under the closely associated Northeast Sector Landscape Improvement Plan. The restoration plans will focus upon rebuilding the dense, diverse forest
understory strata through intensive replanting with native species. The restoration strategy will also diversify edge zones/hedgerows with native berrying shrub plantings. The
old field and meadow areas restoration will focus upon periodic mowing to maintain the
ecologically productive open character, overseeding/planting with native wild grasses
and wildflowers, and establishing new meadows in previously disturbed areas. All vegetation improvements are intended to increase stormwater filtration and uptake while
better protecting soils from excessive erosion. Improvements will also increase wildlife
habitat opportunities and year-round aesthetic interest for reservation users.
Chapter II.4:
Figure II.4.11 Low-tech bioengineered shoreline stabilization concept for small ponds in the reservation. (Source: Fresh Pond Reservation Natural Resource Stewardship Plan, Rizzo Associates, Inc., 1999.)
Vegetation management monitoring

An important component to the Landscape Management and Maintenance Plan will be
to explore different management techniques in different areas and to closely monitor the
effectiveness of the various techniques in general and relative to one another. The monitoring program will provide the basis to assess future management needs for the selected
areas. It is important to note that the initially treated areas will likely require additional
follow-up management. Management activities, including reapplication of herbicides and
follow-up cutting or hand-pulling, may be necessary for a number of years subsequent
to the initial treatment to achieve an acceptable level of invasives control and to begin to
restore vegetative diversity (Figure II.4.11).
Erosion control and slope stabilization

The second component of the Landscape Management and Maintenance Plan is erosion
control and slope stabilization. These efforts will be directed toward the steepest, most eroded,
and most visible slopes and shorelines, as well as those that directly threaten the water quality
of Fresh Pond Reservoir. Low-tech, low-cost bioengineering approaches will be used to
achieve initial stabilization. Techniques may include some or all of the following:
Securing base of slope or bank with biodegradable coconut roll (coir fascine) and
porous gravel base
Intercepting gully-bound runoff in gravel based dry wells or french drains
420

Spreading composted leaf litter over the selected slope areas
Creating organic pockets with vertically driven wood stakes and compost
Harvesting acorns and nuts at the reservation and spreading them over the organic
layer (composted leaves)
Spreading topsoil over the organic layer
Securing the organic layer with biodegradable erosion control fabrics
Seeding the areas with a noninvasive cover crop (e.g., white clover, switchgrass)
and/or meadow wildflowers and wild grasses in sunny exposures
Planting native sapling trees, shrubs, and herbaceous plants
Placing wind-fallen branches, branch bundles, and other coarse-textured woody
debris over the treated slopes to discourage trampling
In addition, the placement of cut logs and live brush bundles (fascines) or the like
buried and staked in place across slope faces may also be considered where necessary
and feasible. Slope stabilization will also involve the removal of invasive trees (e.g.,
Norway maples) to improve light penetration to the ground level. The goals of the upland
slope work are soil stabilization in the short term and full forest structure restoration of
all vegetation strata in the long term. Final design methodology will be informed by
extensive soil sampling of the areas in question (Figures II.4.12, II.4.13, and II.4.14).
Water purification facility/Weir Meadow landscape improvements

The Landscape Management and Maintenance Plan is an independent capital project.
However, other notable projects in the Fresh Pond Master Plan involve bringing existing
landscape plans into compliance with the goals of the natural resource stewardship plan,
now incorporated into the master plan. indeed, it was the publics call for a well-coordinated approach to landscape planning at the reservation that helped to galvanize the
community-based master planning process in 1997. One such project is the improvement
plan for the citys new state-of-the-art water purification facility (which went online in
summer 2001) in the reservation and an associated meadow area (http://www.ci.cambridge.
ma.us/~Water/startup.html).
The Weir Meadow is a popular, 5-acre sloping lawn area that has historically been
home to the reservoirs weir structure. The structure regulates piped flow into Fresh Pond
from Cambridges upcountry reservoirs located approximately 8 miles to the west.
Landscape improvements to the water purification facility and the Weir Meadow area
have been accomplished through restoration-oriented (naturalized) planting strategies.
Such strategies have emphasized the use of native species to protect water quality and
improve aesthetics and ecological values. The strategies have included:
Replacement of lawn areas with lower maintenance groundcovers and meadows
Use of low-maintenance lawn species where lawn remains
Restoration of forest strata with addition of groundcovers, shrubs, understory
trees, and canopy trees
In addition, heavy revegetation of the site is intended to improve overall soil health by
protecting the surface from erosion, breaking up the soil through root growth, increasing
infiltration opportunities, and adding organic matter as plant materials are annually cycled
back into the earth. Improved soils equate to reduced surface runoff into the reservoir.
Another important design strategy is the inclusion of substantial wetland-oriented
bioswales, at both the water purification facility and Weir Meadow sites, to capture and
Figure II.4.12 Upland slope stabilization and forest restoration concept. (From the stewardship plan.)
Chapter II.4:
422
Figure II.4.13 Deep shade of invasive Norway maple (Acer platanoides), at left, prevents lower forest
strata, growth leaving soil surface on steep slopes open to erosion, which, in turn, increases silt load
of surface runoff into Fresh Pond Reservoir (directly behind fence).
Figure II.4.14 Plans for the reservations Weir Meadow include heavily planted bioswales, designed
for the primary purpose of surface runoff filtration, to be located approximately where pools form
on the existing lawn and overflow into the reservoir. (Right) Newly installed bioswale.
filter runoff prior to its entry, primarily as groundwater, into the reservoir. The bioswales
will be heavily seeded and planted with wetland herbs and encircled by dense native
wetland shrub borders. One of the most valuable contributions of the overall design will
be a substantial increase of high-quality edge habitat and hedgerows a great benefit to
wildlife (Figures II.4.15 and II.4.16).
A public art piece has also been included in the facilitys design, near a major entrance,
in the hopes of raising public awareness about the citys watershed and water delivery
system (http://www.harriesheder.com/www.camb.htm).
Reservoir fence and Perimeter Road resurfacing alternatives

Two exciting studies are emerging from the Fresh Pond master plan both of which
examine alternatives to better aesthetically frame the reservoir while protecting it from
Chapter II.4:
Figure II.4.15 Cambridges new state-of-the-art water purification facility, to go online in 2002, is
complemented by a water-friendly landscape around it. (Left) View of new facility from Kingsley
Park. (Right) View of newly landscaped facility with new bikeway in foreground and Fresh Pond
Parkway (vehicular) at left.
Figure II.4.16 Persistent flooding and drainage problems at Fresh Ponds municipal golf course are
being addressed by a master plan for the course now under way.
surface runoff contamination. The Reservoir Fence pilot study will determine the most
aesthetically complimentary and economically feasible fence type or types to replace a
deteriorating existing chain-link fence. An existing fence encircles the entire 2.5-mile
perimeter of Fresh Pond Reservoir, with the notable exception of one short unfenced
section of shoreline immediately adjacent to the new water purification facility. Due to
the highly urban context in which the reservation and reservoir sit, the city has determined
it necessary to entirely fence its drinking water supply. In the citys view, unlimited water
access by people and dogs would represent an uncontrollable threat to drinking water
quality. Further, the reservation has a number of other significant water bodies, most
notably Little Fresh Pond and Blacks Nook, which are easily accessible to reservation
users. However, due to their accessibility these ponds tend to be highly eroded along their
unprotected shorelines. Current and future projects will address reinforcement of these
shorelines.
424
Figure II.4.17 Pilot section of stabilized aggregate Perimeter Road resurfacing. Stabilized aggregate
surface is porous, natural in appearance, and easy on the feet. Stabilized aggregate acts as a sponge
when wet preventing runoff into the reservoir.
In the meantime, the fence pilot project implementation is under way with a new 4ft-high, black polyvinyl-clad chain-link fence installed along the shoreline in front of the
new purification facility. Related to the fence pilot study are ongoing efforts to improve
views to the reservoir from the reservation, primarily through the selective thinning of
vegetation along the reservoirs shoreline. Some consideration is also being given to the
creation of additional controlled access openings in the reservoirs fence to get reservation
users closer to or over the water, though this is only in the conceptual stage.
More directly related to surface runoff management, the Perimeter Road resurfacing
pilot study is investigating the viability of the Stabilizer (brand) product (http://www.
stabilizersolutions.com), a porous, nonchemical aggregate binder material. The goal of the
study is to repave the entire 2.5-miles of deteriorating asphalt road that encircles the
Fresh Pond Reservoir with an aggregate bound with Stabilizer.
Stabilizers use here is expected to allow for much increased infiltration of surface
runoff prior to its reaching the reservoir. This durable, low-maintenance material has
proven effective at Battle Road National Historic Park in Lexington, MA. An approximately 1500-ft-long pilot section of stabilized aggregate was installed in 2001 and is
under study. (Figure II.4.17).
Northeast sector landscape improvements

A recently begun project at Fresh Pond Reservation is the design of landscape and site
improvements to Neville Manor, which is being converted into an assisted-living center,
and nearby Lusitania Field, a former soccer field being used as the construction staging
area for the new water purification facility. The project area also encompasses a large
upland and wetland forest and old field edge zones.
Although in its early planning stages, current components of this project are envisioned to include construction of a new youth soccer field and a meadow to maximize
songbird habitat with an environmental art concept closely integrated into the overall
design. As discussed earlier, the 25-acre project area will significantly overlap with highpriority areas for vegetation management, as specified in the Landscape Management and
Chapter II.4:
Maintenance Plan. As such, the project will take the management strategies specified
therein to the next level by preparing full restoration plans for the areas to be managed
for invasive species. Design work is expected to be completed by the end of 2002 with
site work expected to begin in 2003.
Golf course master plan

In addition to providing open space around Fresh Pond Reservoir, the reservation is home
to the citys only municipal golf course. The Thomas P. ONeill Municipal Golf Course,
which covers approximately one third of the reservations total area (51 acres maintained), has been in continuous play since the 1920s. Although it has been upgraded at
various times since its inception, the course has been plagued with significant drainage
problems. Moreover, its playability has not been examined comprehensively for years.
Further, the complex stormwater drainage system which traverses the course, and involves
the neighboring community of Belmont, has never before been comprehensively studied,
to the citys knowledge. However, the city has recently hired a consultant to do exactly that.
Master planning for the golf course is being approached from the dual perspectives
of playability and environmental soundness. The planning is underway and conceptual
recommendations for playability improvements will be reviewed in the near future. With
the city conducting a comprehensive stormwater and drainage study of the golf course,
the master planning for its use will soon be informed by critical drainage data. The
drainage study will identify and quantify stormwater infrastructure problems contributing
to flooding at the course and recommend solutions for related improvements. Such a study
will be critical to improving drainage problems at the course and for protecting water
quality in the adjacent reservoir (Figure II.4.18).
Bikeway corridor landscape improvements

Yet another project currently under way is the Fresh Pond Bikeway Corridor project. This
project focuses on landscape improvements to a 1,000-ft-long section of new bikeway
corridor associated with a much larger sewer separation project occurring within and
along existing Fresh Pond Parkway, a major arterial route in Cambridge. The landscape
improvement area covers about two acres. About half of the bikeway, within the project
area, runs adjacent to the reservoirs highly used Perimeter Road within the narrowest
section of the reservation. The reservations width here is only about 60 ft from the
reservoirs shoreline to the edge of curb at Fresh Pond Parkway. The project may be broken
into two distinct sections. The first covers the narrow section between the reservoir and
the Parkway, the second extends the bike path along the Parkway edge of a lowland forest
subject to seasonal flooding.
Due to the projects immediate proximity and groundwater connections to the reservoir, water quality protection was a central goal of the landscape design work, which was
completed in 1999 and 2000. Other important design goals included visual and noise
buffering of Perimeter Road from adjacent Parkway traffic (and ongoing roadway reconstruction), strong separation between bike path and Perimeter Road to avoid user conflicts
between bicyclists and pedestrians, and enhancement of views toward the immediately
adjacent reservoir. Further, in keeping with the master plan goals, the design work also
aimed to maximize habitat values and forest restoration opportunities.
In response to these goals, a number of design strategies were employed here and a
variety of water-conscious features included. Perhaps, the most visually notable feature
was the creation of a 3- to 5-ft high, 30-ft-wide linear berm running between the Perimeter
Road (reservoir) and the bike path itself. The berms north side, closest to the Parkway
426
Figure II.4.18 Landscape plan for Fresh Ponds new bikeway, now under construction, calls for berms and heavy evergreen plantings
to buffer the reservation and reservoir from adjacent parkway traffic impacts. (From Fresh Pond Bikeway Corridor landscape design,
phase one, Rizzo Associates, Inc. and Carol R. Johnson Associates, Inc.)
Chapter II.4:
Figure II.4.19 Berm construction along new bikeway, April 2001. Perimeter Road on right along
reservoir fence, bikeway on left (behind berm), and Fresh Pond Parkway on far left (behind trees
and Jersey barriers).
and adjacent to the bike path, will be heavily planted with native evergreen trees and
shrubs, while the sunny south side will be planted with a low-maintenance lawn mixture
to provide an informal grassy sitting area with opened views to the reservoir. The berm
will serve multiple purposes. The primary purpose will be to buffer noise and visual
impacts from the Parkway. The berm will also help keep stormwater runoff and airborne
dust/garbage (particulates) out of the reservoir. To complement the view to the reservoir
from the berm, selective tree thinning along the shoreline will occur. The thinning will
also remove invasive species and increase the presence of favored desirable natives species
(e.g., river birch), which will further enhance aesthetic and habitat values.
Another important strategy used in the woodland section adjacent to the Parkway
was the creation of densely vegetated bioswales. The bioswales will capture and filter
runoff from areas adjacent to the Parkway and slowly filter it into the groundwater table.
This will serve to protect water quality in both the adjacent lowland woods and in the
nearby reservoir. Other planting strategies included dense plantings of a diverse variety
of native trees, shrubs, perennial grasses and flowers, and groundcovers to increase edge
habitat values and restore complex forest strata.
As with all landscape improvement projects at the reservation guided by the master
plans environmentally conscious vision, the Bikeway Corridor design work was the
product of extensive public involvement at all phases, but particularly during the critical
design development phase (Figure II.4.19).
Conclusion
The extensive and complex landscape improvements occurring at Fresh Pond Reservation
today, and in the pipeline for the near future, are very much shaped by the singular goal
of water quality protection. In the design and execution of these projects one is constantly
reminded of the critical connection of the landscape to water bodies via surface and
groundwater drainage and the need for water-minded stewardship. As the Reservation
enters its second century, armed with the strong vision outlined in the master plan, the
future for water quality preservation and the restoration of highly functioning natural
428
Figure II.4.20 The master plans goals seek to protect the reservoirs water quality through wise
stewardship of landscape that surrounds it.
Figure II.4.21 Soil stabilization and improvement through intensive vegetation management, emphasizing restoration of biodiversity, is a central concept in the plans approach.
systems looks bright, indeed. Hopefully, the forward-looking strategies used at Fresh Pond
will succeed in preserving water quality long term and serve as an excellent example for
stewardship of aging, yet much-loved public open spaces everywhere (Figures II.4.20 and
II.4.21).
Acknowledgments
Special thanks to Karen Maki and Sandra Fischer of Rizzo Associates for editorial review
and graphic support, respectively. Further thanks to Amy Green, Jane Taylor, Audrey Sall,
Richard Grant, Cynthia Baumann, Dave George, and Ken Deshais, all of Rizzo Associates,
and to Joseph Ingoldsby of J.E. Ingoldsby & Associates for their respective roles, large and
Chapter II.4:
small, in the preparation of the Fresh Pond natural resource stewardship plan and related
planning and design efforts. From the city of Cambridge, the significant efforts of the
following individuals and committees must be recognized: Chip Norton; Fresh Pond
Master Plan Advisory Committee (FPMPAC 19992001); Fresh Pond Stewardship Plan
Oversight Committee members (19981999), including James Barton, Deborah Howe,
Patricia Pratt, Janice Snow, and Louise Weed; and Julia Bowdoin, Richard Rossi, Lisa
Peterson, Paul Ryder, Robert Carey, and J.C. Girouard. Additional thanks are extended to
others who have significantly participated in the implementation phases of the projects
at Fresh Pond Reservation, including John Amodeo, Jennifer Jones, Randy Sorenson, Ben
Dieterle, and Howard Snyder, Carol R. Johnson Associates, Inc.; John Kissida, John Olcott,
and Nate Sanford, Camp, Dresser & McKee, Inc.; George Sargent, Tom Devane, and Debby
Sargent, Wogan & Sargent Golf Course Architects, Inc.; Julie Mair Messervy.
Literature cited
Benjamin, T.S., Building consensus for natural-resource and water-quality preservation at the Fresh
Pond Reservation, Erosion Control, Sept.Oct. 1999.
City of Cambridge (Fresh Pond Master Plan Advisory Committee), Fresh Pond Reservation master
plan, prepared May 2000, adopted January 2001.
City of Cambridge Water Department Web site:
http://www.ci.cambridge.ma.us/~Water/startup.html.
Harries, Mags, Web site: http://www.harriesheder.com/www.camb.htm.
New England Wildflower Society, Conservation Notes, Invaders, 2, 1998,
http://www.newfs.org/publication.html.
Rizzo Associates, Inc. and J.E. Ingoldsby Associates, Fresh Pond Reservation natural resource stewardship plan, February 1999.
Rizzo Associates, Inc. and J.E. Ingoldsby Associates, Fresh Pond Reservation natural resource inventory, October 1998.
Rizzo Associates Web site: http://www.rizzo.com/portfolio_cambridge_fresh_pond.htm.
Stabilizer Solutions, Inc. Web site: http://www.stabilizersolutions.com/.
Tu, M., Hurd, C, and Randall, J.M., Weed Control Methods Handbook: Tools and Techniques for Use in
Natural Areas, The Nature Conservancy, April 2001,
http://tncweeds.ucdavis.edu/handbook.html.
U.S. National Park Service/Plant Conservation Alliance, Weeds gone wild project, Washington, DC,
[email protected] and/or http://www.nps.gov/plants/alien/.
430
Response
Protecting and restoring treasured landscapes:
Complexity and integration
Landscapes are not only situated in physical space, they can also be layered within the
same space as determined by variable uses. The case study described by Thomas Benjamin
is particularly interesting due to the high visibility of the site. Not only does the Fresh
Pond Reservation protect the city of Cambridges drinking water supply, it is also the
largest green space in the city and is used by golfers, pedestrians, cyclists, dog walkers,
joggers, etc. On top of that, the landscape has historic significance, as, for example, being
designed and planted by the master landscape architects, Olmsted and Eliot. Dealing with
the concerns raised by each of these interest groups, some mutually opposed, is an onerous
task indeed.
Certainly, the most straightforward message from Benjamins study is one of recognizing of the false dichotomy often existing between what is a water sensitive design
project and what is a water sensitive planning project. This chapter, more than most in
this book, is very much a straddler between the two thematic sections because Benjamin
describes a series of site-specific designs whose conceptual origin and strategy of implementation are very much rooted in approaches of traditional planning. The approach taken
of moving from a natural resource inventory as part of a stewardship plan used to
identify problem areas and provide recommendations to a water sensitive management
plan is a strategy to be admired. The watershed assessment planning process described
by Dennis Haag, Stephen Hurst, and Bryan Bear in Chapter II.8 is patterned in a similar
vein.
This chapter also addresses an array of topics considered in more detail elsewhere in
this book, such as shoreline bioengineering covered by Wendi Goldsmith in Chapter I.17
and Nicholas Pouder and Robert France in Chapter I.16, recreational planning of riparian
trails introduced by Charles Flink in Chapter II.3, and biofiltration swales described by
Robert France and Philip Craul in Chapter I.7, and Thomas Liptan and Robert Murase in
Chapter I.6. Benjamins chapter assumes real significance in his presentation of a case
study showing how all these elements, in addition to landscape aesthetic enhancements,
can be effectively integrated into a single, comprehensive solution.
Several other elements of this study offer important lessons to water sensitive planning
professionals that are worth reiterating:
1. The extensive public participation process at all phases of planning and acknowledgment that such bottom-up concerns were instrumental to motivating the entire
project from the start
2. The piggy-backing of the resource plan with ongoing public utility improvements
at the water purification facility, including art installations
3. The use of pilot projects as an incremental approach to both educate the public
and affect the watershed alterations required
4. Understanding that nonstructural planning solutions involving tending and management may be just as important as structural designs
5. Recognition of the need to experiment with different management techniques for
different locations, with the consequent requirement for monitoring or surveillance
through time
chapter II.5
Treating rivers as systems to meet

multiple objectives
Leslie Zucker, Anne Weekes, Mark Vian, and Jay D. Dorsey
Abstract
Planners, local decision makers, and landowners are increasingly asked to meet multiple
objectives for stream resources. The ability of the stream system to meet multiple objectives
depends on complex relationships between system components (e.g., the channel, riparian
area, and floodplain). Similar to other systems with multiple variables, rivers appear to
exhibit an emergent order that transcends the properties of component parts; however,
management approaches have tended to ignore the interconnectedness of river processes.
This chapter examines aspects of river system management that could be improved
through the application of complex systems theory. The application of systems theory to
river planning and management can help to meet multifunctional goals.
After three years
I no longer saw this mass.
I saw the distinctions.
But now, I see nothing
With the eye. My whole being
Apprehends.
My senses are idle. The spirit
Free to work without plan
Follows its own instinct
Guided by natural line,
By the secret opening, the hidden space,
My cleaver finds its own way.
I cut through no joint, chop no bone.
from Cutting Up an Ox, Chuang-Tzu (approx. 300 B.C.E.)
Introduction
This enigmatic poem presents a simple meat cutter as artful master of his trade. The meatcutter is able to move his blade to the place where the least effort is required to meet his
1-56670-562-2/02/$0.00+$1.50
431
432

riparian area
channel
floodplain
channel meander belt
Figure II.5.1 A river includes a channel, riparian area, and floodplain among other natural features.
The river uses the riparian area and the floodplain to dissipate an immense quantity of hydraulic
energy. The channel meander belt might shift over the entire floodplain in a period of several
hundred years; however, within management time frames, the meander belt might be confined to
a relatively fixed position. (Adapted from Palmer, L., River management criteria for Oregon and
Washington, in Geomorphology and Engineering, D.R. Coates, Ed. Dowden, Hutchinson & Ross, Inc.,
Stroudsburg, PA., 1976.)
goal. Rivers are the artful masters of their own existence. They balance the forces of moving
water and sediment against the landscape, assuming a form that minimizes the energy
required to do the necessary work.
In this chapter, we examine rivers from the perspective of systems science. The elements that make up a river (such as water flow, nutrients, plant and animal communities,
and bedload) have a tendency to organize into an interconnected and complex system.
One assumption of systems theory is that systems self-organize into emergent states that
cannot be inferred from their parts (Waldrop, 1992).
This assumption suggests the potential to expand river science beyond traditional
approaches. We explore the management implications of ignoring the tendency of rivers
to behave as systems, and suggest that application of systems theory to river planning
and management might improve our ability to meet multifunctional goals. The word
river is used in reference to systems of all sizes: river, stream, creek, or ditch.
In 1976, the geologist Leonard Palmer wrote that given reasonable freedom to continue its normal functions, the river will provide for the discharge of water and sediment,
and will sustain an environment of high quality for wildlife habitat and for humans
(Palmer, 1976). He goes on to say, Management must provide a sufficient corridor to
allow continuation of meander belt progression [downstream]. Meander belt progression
refers to the downstream migration of meander bends as the stream erodes the outside
of banks and deposits sediment on the inside of bends and on adjacent floodplains
(Figure II.5.1).
Palmers management philosophy recognizes that such dynamic processes can cause
a river to adjust its course and perform a number of other actions, or functions. This
philosophy recognizes a certain kind of stability that a river given reasonable freedom
to continue its normal function will discharge water and sediment while sustaining a
high-quality environment. Although stream management goals may include stability, we
have generally failed to recognize that the river is often able to provide the outcomes we
Chapter II.5:
Treating rivers as systems to meet multiple objectives
433
desire without human manipulation. Some engineers and scientists have begun to describe
the river as the carpenter of its own edifice (Leopold, 1994). One well-known practitioner
of applied river morphology has said that river form and fluvial processes evolve simultaneously and operate through mutual adjustments toward self-stabilization (Rosgen, 1996).
Generations of hydraulic engineers have used mathematical equations and rigid structural controls to define and refine the operations of a rivers component parts. Although
reductionist science, statistical variability, and empirical field techniques are powerful tools
with which to examine both the universal and landscape-specific processes of rivers, these
tools necessarily ignore the collective behaviors that arise from the properties of parts.
Traditional stream management has mirrored the tendency of science to break systems
into parts. Separate resource agencies emerged to address management goals related to
individual stream functions. Examples of stream function include runoff and stormwater
conveyance, decreased storm flow peaks, groundwater recharge, sediment and debris
sorting, nutrient cycling, pollutant retention and transformation, water cooling and oxygenation, channel maintenance, fish and wildlife habitat, and recreation and aesthetics.
Planners, engineers, and designers within separate organizations are increasingly asked
to sustain one or several stream functions simultaneously and, to that effect, choose
designs that manage the relevant properties of component parts. Often, however, these
designs fall short of meeting multifunctional goals (NRC, 1992). Greater success might be
achieved if multifunctionality is viewed as a desirable property of dynamic river systems.
New management approaches will emphasize the complexity of river systems,
addressing characteristics such as channel morphology and the integrity of the biological
community, instead of isolated attributes such as channel stabilization or simplification
of the stream fringe. This might require a significant change of attitude for the designer,
the resource manager, and the public at large. Historically, the activities of managers
(including resource agencies, planners, and individual landowners) have not been well
coordinated. As competing interests in individual functions increasingly become apparent,
stream managers and planners must find ways to coordinate their management goals and
practices. The required changes can be viewed as properties of a social system that is as
dynamic and complex as the river itself.
The following sections explore application of systems science to aspects of river
management and planning.
Interactions
According to systems theory, systems are controlled by interactions among component
parts. Like other systems, a rivers ability to maintain multiple functions is dependent
upon the magnitude and dynamics of such interactions. Loss of connectivity between
component parts can alter the flow regime, degrade water quality, and affect the health
and distribution of aquatic communities.
Connectivity in a stream is expressed in several dimensions: longitudinally
(upstream/downstream), laterally (channel/floodplain), between scales (microhabitat/macrohabitat), and temporally (timing of flows). One of the most significant examples
of connectivity is the interaction between channel flows and the floodplain. Filling of
floodplains, channelization, berming, and riparian vegetation removal are common practices that limit interactions between stream system components. A number of processes
are potentially affected, including vegetative growth, groundwater recharge, stream flow
regime, erosion and deposition, and nutrient cycling.
For example, some of these management practices result in lowering of the water table
below the riparian rooting zone, which alters nitrogen cycling within riparian areas.
434
Figure II.5.2 A schematic representation of a food web showing few species with many links and
many with only one or two links. The food web is presented as a graph. Highly connected species
can exist at different trophic levels depending on the particular food web. (Adapted from Sole, R.V.
and Montoya, J.M., Complexity and fragility in ecological networks, Santa Fe Institute Working
Paper 00-11-060, 2000.)
Nitrogen removal within the riparian area is biologically mediated through plant uptake
and bacterial processes. The biophysical structure necessary for maximum nitrogen
removal includes a dense and deep plant root structure, and organic matter such as dead
roots and fallen leaves. By lowering the water table, management practices can prevent
nitrate-laden shallow groundwater from interacting with the rooting zone where these
conditions exist.
Nitrogen removal is also an example of how organisms help to mediate local and
regional flows of energy and materials (Naeem et al. 1999). Many stream ecosystem
processes are sensitive to declines in the diversity of aquatic organisms. Biological diversity
stems not only from the numbers of species involved, but also from the richness and
variety of interactions organisms have with their environment and with other organisms.
It appears that successful management of aquatic communities requires the maintenance of characteristic types and levels of connectivity to which biotic communities are
adapted. Some aquatic food webs are characterized by numerous weak connections
between species and very few strong interactions (Montoya and Sole 2000) (Figure II.5.2).
Although food webs show resilience to random removal of species, disturbances that
result in selective removal of highly connected species may be enough to promote ecosystem collapse (Sole and Montoya 2000).
Thus, the type and rate of disturbance, and in addition, the degree to which its spatial
effect is either amplified or muted by connectivity, are important factors in maintaining
biodiversity and multiple ecosystem functions. Within stream systems that exhibit moderate levels of connectivity, aquatic communities can be locally unstable, yet globally
stable. In contrast, landscapes that are highly connected can be globally unstable, because
disturbances are able to spread rapidly through a homogeneous system (Green 1994).
Periodic ecosystem disturbances, such as flooding or fragmentation, provide critical
input to the structure and dynamics of stream communities. Some evidence suggests
aquatic communities require a moderate regime of disturbance fluctuation over time and
in space in order to remain healthy (Reice et al. 1990). In contrast, as a part of watershed
development and water management, we tend to drastically alter disturbance regimes by
preventing flooding, clearing floodplain vegetation, or by directly modifying the channel.
These changes are particularly harmful to gravel-spawning fish and benthic invertebrates
when alteration of natural channel morphology and hydrology changes sediment deposition dynamics (Kondolf et al. 1987). In rivers with natural channel morphology (e.g.,
pool/riffle development), channel-forming flows at or near bankfull tend to remove fine
sediments deposited on stream beds while leaving beneficial coarse gravels (Kondolf and
Wilcock 1996; Wilcock et al. 1996).
Chapter II.5:
435
Premodification
Channel
Constructed
Channel
Downcutting
Undercutting
Sloughing
Aggradation and Stabilization
Figure II.5.3 A representation of channel evolution following channelization observed in West

Tennessee. The modified cross-section is shown superimposed over the original natural channel.
The dashed line represents the original modified cross-section. After modification, the channel begins
to downcut and then undercut banks, which eventually slump into the channel. A new, low-flow
channel cuts a path through sediment deposits. (Adapted from Simon, A. and Hupp, C.R., The
recovery of alluvial systems in response to imposed channel modifications, West Tennessee, U.S.A.,
in Vegetation and Erosion, Thornes, J.B., Ed., John Wiley & Sons, New York, 1990.)
In order to maintain high levels of biodiversity, planners, designers and managers

should strive to maintain natural ranges of the type, rate, and extent of disturbance, and
to avoid creating overly homogeneous or fragmented landscapes that limit interactions.
System adjustments
The functionality of a system depends on the nature and arrangement of its component
parts. From this perspective, channel form at the scale of the stream reach might be
understood as an emergent property of dynamic interactions between variables. Within
the field of fluvial geomorphology (the field that studies how running water and the
landscape shape each other), stream form is considered a function of the following interrelated variables: stream discharge, sediment load, sediment size, slope, velocity, width,
and depth as these adjust to resistance created by vegetation and geology (Leopold
et al., 1992). A change in any one of the variables causes the other variables to adjust.
Systems theory proposes that a preferred arrangement for the system exists and that
systems evolve toward the preferred state. The flux of energy and matter through dissipative systems is a driving force generating order. Of the many possible channel forms
that could result from the driving variables, the channel tends toward a most probable
form related to the dissipation and distribution of energy within the system (Leopold,
1994). Researchers in fluvial geomorphology have noted that alluvial channels appear to
travel through different stages of channel evolution. One widely cited example of this
theory is shown in Figure II.5.3, which documents evolution of channel morphology
toward a new state following disturbance (Simon, 1989).
The culturally accepted stream-planning and management approaches in many areas
of the country do not allow river systems to self-organize. This is particularly true in
urbanized or settled areas, where infrastructure has been built within the floodplain, and
in areas that require drainage of wet soils for development. In these situations, stream
channels are typically maintained in a modified form.
The history of the midwestern United States is closely associated with the conversion of
wetlands and wet soils to productive agricultural lands. As settlement of wetlands proceeded,
436
Floodplain
Bankfull
Baseflow
(a)
Floodplain
Bankfull
Baseflow
(b)
Figure II.5.4 Cross-sectional views of (a) a trapezoidal-shaped drainage ditch, and (b) a natural
channel cross-section. The drainage ditch has been deepened to increase water-carrying capacity
and to lower the water table in a timely fashion following storm events. Notice the dimensions of
the channel have been changed so that the bankfull flow is now spread over a wider area, slowed
and made shallower. This promotes deposition. In addition, flows that once accessed the floodplain
are now contained within the channel. This increases erosive force on the ditch banks. (Adapted
from Rosgen, D., Applied River Morphology, Wildland Hydrology, Pagosa Springs, CO, 1996.)
the first drainage efforts were ditching programs, introduced primarily to remove standing
surface water. Today, a mixture of surface and subsurface drainage measures are used to keep
basements and yards dry; remove water from parking lots and roads; lower the water table
for agricultural production; and treat septic system effluent in leach fields.
The traditional drainage ditch was historically designed for, and highly effective at,
achieving one function: rapid downstream conveyance of unwanted rainwater. Drainage
improvement has been seen as a way to help rivers and promote stewardship of the land.
A straightened, trapezoid-shaped channel was typically accepted as the ideal design, as
evidenced by the comments of one Wisconsin engineer who wrote that the streams
running through [swamps and bottoms] are not swift enough, meandering about and
taking double, or even more, the necessary time to cross a certain distance. This meandering, crooked or irregular course of a stream is also the course [sic] of its unequal width
and depth, and of the wear and tear of its banks (Kessenger, 1890).
A schematic of a typical drainage channel cross-section compared to a natural stream
cross-section is shown in Figure II.5.4. In an attempt to achieve channel stability and
control flooding, natural channels are often widened and straightened. This design is
intended to expand channel capacity by increasing the width of the channel. Channel
straightening increases the channel slope, which accelerates mean flow velocity during
peak runoff events and facilitates the timely removal of water.
Over time, however, adjustments within the channel can threaten these functions,
necessitating frequent repair of the modified channel. Flow distributed across the uniform
Chapter II.5:
437
Figure II.5.5 Some straightened channels begin to develop deposition and erosion patterns that force
meander redevelopment shortly after modification. Water rarely flows in a straight line.
bottom of the ditch promotes sedimentation and channel aggradation. Flows begin to
accelerate around depositional areas as the channel regains meander geometry and slope
(Figure II.5.5). Meander formation acts against ditch walls, which erode and slump, adding
additional sediment to the channel (this process is also illustrated in Figure II.5.3).
In some systems, straightening and piping of headwater streams to facilitate development result in faster delivery of larger flows downstream. Increased discharge leads to
downstream channel instability and adjustment to new and often larger channel configurations. If adequate channel capacity is not available downstream, flooding occurs.
A typical response to flooding is construction of levees or other containment structures. This strategy can fail to prevent flooding if aggradation continues to cause flood
elevations to rise. For example, for hundreds of years upstream land management practices
added large amounts of silt to the Yellow River of China. Levees were continuously built
higher to contain flood flows. As a result of this centuries-long cycle, the river bed is now
at a higher elevation than the surrounding land in some areas.
Urbanization exacerbates the effects of channelization by changing the water cycle
within watersheds (Ferguson, 2001). Water drained from suburban housing developments
and urban roofs, roads, and parking lots is typically directed to the existing stream
network. Water that would have seeped slowly through soils and back to the stream
channel now reaches the channel as runoff. The same storm event produces higher flood
peaks and larger, faster flows. As a result, erosive forces on stream banks and beds increase.
A typical management response to bank erosion is hardening of the banks with
concrete slabs, tires, or other means in order to hold soil in place. This prevents the channel
from adjusting to a new size that accommodates increased stormflows. The river system
responds by carving a new pattern in reaches that have not been hardened, and the
resulting erosion can lead to bed aggradation in another part of the system.
Both urban and agricultural drainage systems must be maintained in the desired state
at considerable expense. In 47 western-Ohio counties alone, almost $3 million was spent
on drainage ditch maintenance in 1996 (Atherton, 1998). These costs and others must be
balanced against the benefits of traditional approaches that maintain a single objective.
Planning and management approaches that recognize typical river system responses
should be investigated for their ability to lower long-term costs and maintain a full range
438
of natural function. Todays river scientists increasingly view meandering channels and
connected floodplains as mechanisms by which potentially destructive energy is dissipated from the system.
Planning approaches
Although a great deal is known about river system function (i.e., the importance of floodplains, nutrient cycling, riparian corridors, channel processes, etc.), society still continues to
make behavioral choices decisions regarding stream management that have severe
unwanted consequences, such as ecological degradation and risk to life and property. Why
do we continue to make these choices? Similar to river ecosystems, the human community
is a self-organizing system whose dynamics may be understood through the lens of
complexity theory.
In any particular watershed, the number of players with a stake in managing the river
is startlingly large (Figure II.5.6). Interactions between stakeholders are fluid and complex.
Each of the stakeholders holds different meanings and values for the river. In the terms
of Luhman (Luhman, 1982), each stakeholder group approaches the dialogue around
stream management from the perspective of its own meaning world. These meaning
worlds are related but functionally differentiated for example, the worlds of property
rights, of economics, of governmental bureaucratic culture, of environmental law, etc.
Within particular meaning worlds, individuals engage in conversations that manifest
the social system, articulate the individuals perceived interests, and define the boundaries
of stakeholder behaviors. In the current socioeconomic climate, the discourse between
meaning worlds tends to result in one or several meaning worlds trumping all others as
the ultimate determinant of resource management acts. As a result, the dialogue produces
actions that can privilege one function of the river system at the expense of many others.
The fact that stakeholders, including watershed residents, policy shapers, and educators, generally share a limited vision of the stream system may contribute to this outcome.
Stakeholders who view their streamside property or function of interest as separate from
the entire system may not recognize the full consequences of individual actions. As a
result, the most politically expedient or profitable solution is often chosen, although the
consequences include costly system adjustments and loss of overall stream function.
Systemic changes are often difficult to predict within a complex system. To continue
basing actions on ignorance of system responses, however, puts land managers in the
position of constantly reacting to the negative consequences of past decisions. Approaching the planning process with a vision of the river as a complex system could change the
ultimate valuation of design and management approaches.
One strategy for transforming the shared vision of stakeholders is to introduce a new
meaning world in the guise of a new institution or organization that holds a vision of the
stream as a complex system and explicitly makes this vision a subject of its discourse.
Two organizational formats that possess this capacity are (1) the stream, river, or watershed
association, and (2) the stream, river, or watershed management advisory group, both
containing members with diverse interests in stream function. These groups tend to have
development of multiobjective management plans or related planning processes as a
primary objective.
Participation in a stream, river, or watershed association ostensibly expands the members frame of reference to a larger scale of system function (i.e., from the reach traversing
the individuals property, to the entire system). Participants realize that the function of a
particular reach (for example, stability of a stream bank near ones house) is contingent
on the way other actors manage their respective properties. Where individual interest
transcends the limits of a particular meaning world, a new state of stream corridor
SHANDAKEN
BR
W
NYS
DPRHP
ACOE
DEP
Emergency
Services
Utilities
Through
Traffic
Trout
Unlimited
FEMA
TROUT
Phoenicia
Recreational
Users
Onteora
School
District
NYS
DOT
olid
ns
ate
Phoenicia
Fish and Game
o
nc
e
ish
dF
an
rm
NRCS
Developers
Greene
County
New York City

Water Users
Tourists
Ulster
County
d
te
da s
oli ner
s
n w
co do
Un Lan
Hunter
Govt
Hunter
Ulster County
Soil & Water
Ulster County
Hwy and Bridges
Shandaken
Town Highway Dept.
Hunter
Hwy Dept
E
CRE
VE ED
O
L
H
C
NY ERS
STO WAT
Lanesville
Chichester
HTC
School
Hudson Basin
River Watch
Stony Clove
Rod & Gun
Club
OTHER WILDLIFE
NYS
DEC
USGS
Loggers
Green County
Highway
Figure II.5.6 Map of stakeholders in two adjacent subwatersheds in the Catskill Region of the New York City Water Supply Watershed, produced by
representatives of landowners, municipal officials, sport fishing groups, and area resource agencies during a participatory stream management planning
session initiated by the NYC Dept. of Environ. Protection, Stream Management Program. Each triangle, circle, or rectangle represents a distinct management
interest.
Local Business
Association
ESOPUS CREEK
Catskill
Watershed Corp
l
Environmenta
Groups
Watershed
Ag Council
Town of
Shandaken
Govt
OA
DS
W TR
A
TE EET
RS HO
HE LL
D
O
Shandaken
Town Highway Dept
US
F&W
Ulster County
Hwy and Bridges
SEMO
Greene Lexington Greene

County Highway Dept County
Highway
Soil & Water
Chapter II.5:
NYS
DOT
Town of Lexington
Govt
Green County
439
440
management may emerge that emphasizes long-sighted values such as ecology, community accord, and quality of life.
In the case of a stream or watershed management advisory group, previously uncoordinated, disparate management goals and practices are considered in a unified context,
and contradictions in management approaches can be made explicit. The group is a forum
for a new discourse that addresses the consequences of practices aimed at preserving a
single stream function. Process linkages in the physical system can become the subject of
conversation within and between management organizations.
Although these organizational forms have the capacity to generate new meaning
worlds that reflect the complexity of rivers, the emergence of unifying dialogues is not a
foregone conclusion. Because stakeholders may have historically antagonistic discourses
(and practices), it can be useful to employ professional facilitators of small-group process,
whose primary role is to mitigate the tendencies of participants to regress into purely selfreferencing discourses. Such processes might reveal coupled interests that were not previously apparent, allowing management responses to emerge that support multiple
human interests and natural system functions.
In addition to facilitation, simple frameworks for understanding both river and human
behavior might be key to shifting the dialogue between diverse interest groups. One
example is use of the Rosgen stream classification system (Rosgen, 1994) (Figure II.5.7) to
impart some understanding of channel processes to a lay audience. While fluvial geomorphology has been the subject of academic study for decades, Rosgens typology has
enabled discussion of the relationship between stream form and function, and the effects
of various management practices on system function, with local officials, planning boards,
and interest groups throughout the country (Benoit and Wilcox, 1997; Azary, 1999; Lovegreen,
2001).
Summary
Rivers of any size, whether a tiny rivulet or the mighty Mississippi, are systems that follow
physical laws and exhibit similar patterns of behavior that can be described roughly with
mathematical functions. This allows us to predict their behaviors reasonably well. For
instance, we know that streams will flood, and we can predict the frequency with which
certain sized floods will return.
What we cannot predict is the exact date or year that floods will occur. We know that
rivers within certain valley types and within certain climates tend to form patterns that
can be classified. We cannot predict the exact course a river will take, or the exact way an
aquatic ecosystem will respond to a flood event. The reason we cannot make these exact
predictions is that rivers, at all scales, are multivariate, complex systems. Rivers exhibit
properties that cannot be predicted from their component parts. Furthermore, rivers are
subsystems nested within and connected to larger global systems.
Although we can predict that rivers will flood (albeit, without knowing exactly when),
we continue to build structures within floodplains. As a result, floods produce predictable
outcomes such as loss of property that we unreasonably continue to see as unmanageable
tragedies. If our conversations begin to reflect an awareness and appreciation of the stream
as a system, it is likely we will develop organizational structures that consider our collective interests in all of the functions that rivers manifest.
As river managers and planners, perhaps we too can become as artful as the meat
cutter in the poem at the beginning of this chapter. We suggest that ecologically sensitive
water planning, design, and management can meet multiple goals for the river by applying
the strategies in the following section.
DOMINANT
SLOPE
RANGE
CROSS
SECTION
VIEW
Aa
+
Aa +
10
A
4-1
0%
B
2 - 4%
< 2%
< 4%
D
DA
DA
< 0.5%
< 2%
< 2%
G
2 - 4%
Figure II.5.7 Level I of Rosgens stream delineation and classification process provides a general characterization of valley type
and landforms and identifies the major stream types, A through G. (From Rosgen, D., Applied River Morphology, Wildland
Hydrology, Pagosa Springs, CO, 1996. With permission.)
STREAM
TYPES
>
Chapter II.5:
PLAN VIEW
LONGITUDINAL, CROSS-SECTONAL and PLAN VIEWS

of MAJOR STREAM TYPES
441
442
Planning and management

Allowing the stream to utilize its floodplain this requires limiting floodplain
uses to low-impact activities without altering or narrowing the floodplain
Recognizing rivers and streams as systems with emergent properties at all scales
of organization
Creating organizational structures to deal with management of streams as systems
Site design and engineering

Allowing the channel to manage itself, except in extreme cases (such as logjam
removals that threaten existing infrastructure) or for initial restoration of stream
processes
Designing culverts and bridge crossings to allow maximum meander belt width
Maintaining important connections between and within physical and biological
components (subsystems) of the river system
Biological
Maintaining and restoring biologically healthy stream corridors
Maintaining important connections between and within physical and biological
components (subsystems) of the river system
How does the meat cutter become a master? By transcending, without losing, the
components of the system:
But now, I see nothing
With the eye. My whole being
Apprehends.
Literature cited
Atherton, B., Drainage improvement assessment methods and subsurface drainage practices in Ohio,
Ph.D. thesis, Ohio State Univ., Columbus, 1998.
Azary, I., Using the Rosgen stream classification system as a pedagogical tool, presented at the
California Geographical Society annual meeting, San Diego, CA, May 1, 1999.
Benoit, T. and Wilcox, J., Applying a fluvial geomorphic classification system to watershed restoration, Stream Notes, Fort Collins, CO, July, 1997.
Ferguson, B., Stormwater management and stormwater restoration, in Handbook of Water Sensitive
Planning and Design, France, R.L., Ed., CRC/Lewis Publishers, Boca Raton, FL, 2002.
Green, D.G., Connectivity and complexity in landscapes and ecosystems, Pacific Conserv. Biol., 1,
194, 1994.
Kessenger, H., Draining swampy districts, The Drainage J., 12, Sept., 1890.
Kondolf, G.M., Cada, G.F., and Sale, M.J., 1987. Assessment of flushing flow requirements for brown
trout spawning gravels in steep streams. Water Resour. Bull., 23:927935.
Kondolf, G.M. and Wilcock, P.R., 1996. The flushing flow problem: defining and evaluating objectives. Water Resour. Res., 32(8):25892599.
Leopold, L.B., A View of the River, Harvard Univ. Press, Cambridge, MA, 1994.
Leopold, L.B., Wolman, M.G., and Miller, J.P., Fluvial Processes in Geomorphology, Dover Publications,
New York, 1992.
Chapter II.5:
443
Lovegreen, M., Bradford County, Pennsylvania Conservation District Manager, personal communication, 2001.
Luhman, N., The Differentiation of Society, John Wiley & Sons, New York, 1982.
Montoya, J.M. and Sole, R.V., Small world patterns in food webs, Santa Fe Institute, working paper
00-10-059, 2000.
Naeem, S. et al., Biodiversity and ecosystem functioning: Maintaining natural life support processes,
Issues in Ecology, 4, 1999.
National Resource Council (NRC), Restoration of Aquatic Ecosystems: Science, Technology, and Public
Policy, National Academy Press, Washington, DC, 1992.
Palmer, L., River management criteria for Oregon and Washington, in Geomorphology and Engineering,
Coates, D.R., Ed., Dowden, Hutchinson & Ross, Inc., Stroudsburg, PA, 1976.
Reice, S.R., Wissmar, R.C., and Naiman, R.J., Disturbance regimes, resilience, and recovery of animal
communities and habitats in lotic ecosystems, Environ. Manage., 14, 647, 1990.
Rosgen, D., A classification of natural rivers, Catena, 22, 169, 1994.
Rosgen, D., Applied River Morphology, Wildland Hydrology, Pagosa Springs, CO, 1996.
Simon, A., The discharge of sediment in channelized alluvial streams, Water Resourc. Bull., 25, 1177,
1989.
Simon, A. and Hupp, C.R., The recovery of alluvial systems in response to imposed channel modifications, West Tennessee, in Vegetation and Erosion, Thornes, J.B., Ed., John Wiley & Sons,
New York, 1990.
Sole, R.V. and Montoya, J.M., Complexity and fragility in ecological networks, Santa Fe Institute,
working paper 00-11-060, 2000.
Waldrop, M.M., Complexity, Simon & Schuster, New York, 1992.
Wilcock, P.R. et al. 1996. Specification of sediment maintenance flows for a large gravel-bed river.
Water Resour. Res., 32(9):29112921.
444
Response
Beyond the banks:
Holistic planning of rivers as more than the sum of their parts
If there is a single and strong, take-away message from this chapter by Leslie Zucker,
Anne Weekes, Mark Vian, and Jay Dorsey, it is that rivers are not isolated entities in the
landscape, but behave as complex systems whose management requires consideration of
the interconnectedness of both elements and processes. Traditional approaches to river
management have been limited due to tendencies to break up and study the individual
influencing variables such as stormwater runoff, groundwater recharge, sedimentation,
nutrient cycling, etc. as if they were independent, ignoring the simple fact that water, more
than anything else, is a landscape integrator. This belief is echoed in Chapter II.2 by James
MacBroom.
For water sensitive planning of river systems to occur in an effective way, we need
to dramatically shift our focus in planning for rivers to planning for river systems. Systems, Zucker and her co-authors argue, are not static systems but are constantly adjusting
(in the short term) and even evolving (in the long term) in a dynamic equilibrium of selforganization.
As Zucker et al. state, for this process to be truly successful, water sensitive planners
must not forget that the complex river systems also support and are supported by complex
social systems. Such planners must search for more effective means of technology transfer
to help educate the lay public and riparian stakeholders. Adjustments to current institutional frameworks might also have to be reexamined to help implement planning decisions. Finally, homeowners must be constantly and reiteratively educated about the importance in regarding river systems from a collective watershed perspective that extends
beyond the bounds of their individual property lines.
chapter II.6
What progress has been made in the

Remedial Action Plan program after ten
years of effort? (Ontario, Canada)
Gail Krantzberg and Judi Barnes
Abstract
This chapter provides a brief history and the current status of the Canada/Ontario Great
Lakes Remedial Action Plan (RAP) program. It highlights the progress that has been
achieved in developing and implementing actions to rehabilitate the Great Lakes since
the inception of the RAP Program in 1987, and argues for new methods of documenting
and communicating accomplishments and impediments. Strategies that address the
impairments related to micropollutants are progressing. Governments have been focusing
on reducing loadings of chemicals from industrial sources and extensive abatement activities have advanced point source controls. This has resulted in declining levels of contamination in water, sediment, and biota, including the elimination of several fish consumption
restrictions in a number of areas of concern (AOCs). Habitat rehabilitation is also proceeding well, due to the ability of the RAP process to engage volunteers and partners in local
action that is both tangible and visible. Conversely, the diffuse nature of nonpoint source
inputs, including stormwater and combined sewer overflows, continues to require considerable financing. The current climate of constraints in government spending means that
limited resources are available to implement storm and wastewater controls at a rate anticipated when RAPs began. Nevertheless, noteworthy accomplishments include new, costeffective technologies that are effective at controlling nutrients and bacterial loadings. To date,
more than half the actions predicted to be necessary to restore beneficial uses in Ontario have
been implemented. Our environmental quality index features the number of beneficial uses
impaired and the extent of impairment. The higher the value of the index, the greater the
degree of impairment. In many locations, the index values have declined appreciably.
The Remedial Action Plan program: Why and how?

Industrial, municipal, and recreational uses of the Great Lakes have imposed great stress
on the basin ecosystem. Insults include toxic substances in air, water, and sediment,
bioaccumulation and food web magnification, bacterial contamination, and eutrophication. Native fish, bird, mammal, and plant species have been lost due to inputs of pollutants, the introduction of exotic species, and the profound destruction of habitat.
1-56670-562-2/02/$0.00+$1.50
445
446
The CanadaU.S. Great Lakes Water Quality Agreement (GLWQA), first signed in
1972 and amended in 1978 and 1987, is the primary mechanism for ensuring a coordinated,
binational approach to management of environmental quality in the Great Lakes basin.
Its purpose is to restore and maintain the chemical, physical, and biological integrity of
the Great Lakes basin ecosystem (IJC, 1988). In 1985, the Great Lakes Water Quality Board
of the International Joint Commission identified 42 degraded areas of concern around the
Great Lakes. Areas were characterized by some or all of the following conditions:
Restrictions on fish and wildlife consumption

Tainting of fish and wildlife flavor
Degradation of fish and wildlife populations
Fish tumors or other deformities
Bird or animal deformities or reproduction problems
Degradation of benthos
Restrictions on dredging activities
Eutrophication or undesirable algae
Restrictions on drinking water consumption, or taste and odor problems
Beach closings
Degradation of aesthetics
Added costs to agriculture or industry
Degradation of phytoplankton and zooplankton populations
Loss of fish and wildlife habitat
According to the language of the Great Lakes Water Quality Agreement, these are
referred to as impaired beneficial uses. With the signing of the 1987 Great Lakes Water
Quality Agreement, Canada and the United States agreed to restore these locations by
developing and implementing remedial action plans (RAPs). Figure II.6.1 depicts the
locations of the Canadian areas of concern.
Restoration of beneficial uses within the Areas of Concern is the primary goal of RAPs,
which are characterized as proceeding in three stages. Stage 1 involves identifying
impaired beneficial uses and sources of environmental degradation. In stage 2, the restoration goals are determined by the local communities with help from agency experts.
Actions required to restore the impaired uses are identified along with a timetable for
implementation. Stage 3 records that the restoration goals and targets have been met and
impaired beneficial uses have been restored. Once beneficial uses have been restored to
the satisfaction of the public and the governments, the site is stated to no longer have the
attributes of an Area of Concern and is removed from the list. This is termed delisting.
In Canada, one area has been delisted (Collingwood Harbour) and by the end of 1998,
the Areas of Concern were beyond the planning phase of stage 2. All are implementing
actions to restore the areas of concern. RAP success depends on accomplishing visible elements. This is partly due to the tenet that informed individuals, groups, and clubs are
interested in getting involved and making a difference. Public involvement in RAPs has been
a major breakthrough for ecosystem management and recovery (Hartig and Zarull, 1992).
Methods for tracking progress

Canada and Ontario reaffirmed their commitments to joint development and implementation of Remedial Action Plans for the Canadian Areas of Concern through the 1994
CanadaOntario Agreement Respecting the Great Lakes Basin Ecosystem (COA, 1994).
The 1994 COA committed Canada and Ontario to restore 60% of impaired beneficial uses
across all 17 areas of concern, leading to the delisting of 9 areas by the year 2000. All AOCs
Chapter II.6:
What progress has been made in the Remedial Action Plan program?
447
Figure II.6.1 Location and status of the Canadian Great Lakes areas of concern. (From Krantzberg, G.,
Aquatic Restoration in Canada, Backhuys Publications, 1999. With permission.)
and impairments were considered to be equally important, and efforts were directed at
all locations. It is becoming increasingly apparent, however, that while this is was challenging
objective, it does not account well for cumulative and step-wise progress in environmental
recovery. It was also conceived that if all implementation was complete by 2000, then an area
of concern would be delisted. This, in fact, is not the case, as we point out in the following
paragraphs.
Progress on RAP development and implementation has been described by the parties,
jurisdictions, and the International Joint Commission in a fairly bureaucratic context, that
is, the completion and submission of stage 1, 2, and 3 reports to the IJC. Progress has been
criticized as slow when measured against the finding that only one of all Canadian and
U.S. Areas of Concern has been delisted. This is a very limited evaluation of the current
status and achievements in the areas of concern.
What are needed are measurements of incremental progress toward fully achieving
the goals and restoration targets at the Areas of Concern. RAPs require a long-term
commitment in order to restore beneficial uses (Water Quality Board, 1991). Along the
way, intermediate indicators of progress could be reductions in the number and extent of
stressors such as declining chemical discharges or additive initiatives to rehabilitate
degraded habitat (Environment Canada and U.S. EPA, 1994) By recognizing that progress
is achieved, and should therefore be documented and celebrated in a step-wise fashion,
RAP participants will be better able to maintain and broaden partnerships and momentum
for Great Lakes rehabilitation and protection for the long term (Water Quality Board, 1996).
To address the shortcoming to date in reporting progress, Canada and Ontario have
been making use of a status pie diagram that depicts the extent to which planning,
448

Stage 3
restoration
Stage 1
planning
implementation
Stage 2
Figure II.6.2 Status pie diagram showing RAP development, implementation, and restoration of
beneficial uses. (From Krantzberg, G., Aquatic Restoration in Canada, Backhuys Publications, 1999.
With permission.)
implementation, and restoration are complete (Figure II.6.2). In each Remedial Action Plan
status pie diagram, the first three quadrants represent problem identification (stage 1),
plan preparation (stage 2), and plan implementation (component of stage 3). The fourth
quadrant represents the degree to which the beneficial uses have been restored, that is,
the progress made toward meeting the Area of Concerns delisting targets. The shaded
portion of each quadrant shows the degree to which the RAP has accomplished its task.
The numbers around the edge of the pie diagram are points at which reports are submitted
to the International Joint Commission. Of note in the pie diagrams, progress in stage 2 is
depicted as occurring in two quadrants in planning and implementation. This demonstrates that implementation is ongoing and not contingent on completion of the overall
plan.
The pie diagram as shown in Figure II.6.2 portrays that problem definition has been
completed (stage 1), a plan for restoring beneficial uses, along with an implementation
framework, has been developed (stage 2), and all actions proposed for restoring beneficial
uses have been implemented (third quadrant). The fourth quadrant is not entirely shaded,
however, reflecting that environmental recovery is not complete.
This scenario could exist for a number of reasons. For example, in the case of eutrophication, several RAPs have set phosphorus loading reduction targets from point and
nonpoint sources in order to achieve receiving water quality of specific characteristics.
When phosphorus load reductions have been met, residual phosphorus in the system may
slow ecosystem response. The attainment of the delisting target in the receiving waters
could take time. Similarly, RAPs are setting targets for rehabilitation of riparian zones and
coastal wetlands. It is feasible that the necessary actions to physically rehabilitate habitat
to a condition anticipated to support healthy wildlife communities have been completed.
Colonization and recruitment may require time to respond, and a sufficient number of
years of monitoring will be required to clearly document change. Another increasingly
familiar condition is in the context of sediment management. In some Areas of Concern,
the preferred option is to institute source control and allow for natural recovery. Once
sources have been controlled, impairments will remain until cleaner sediment covers the
more polluted deposits, or biodegradation proceeds; recovery could take several to many
years.
In the instances depicted in Figure II.6.2, it is important to report to the Great Lakes
community that all reasonable actions have been taken and no further active intervention
Chapter II.6:
449
is needed or planned at the present time. This in itself is a statement of victory. Of course,
there should be sufficiently rigorous criteria applied to the phrase all reasonable action,
to ensure that the decision to no longer actively intervene is not misused as an excuse for
inaction. Also paramount is that a formal monitoring program is in place to track natural
recovery and attainment of the RAP delisting targets. We suggest that until the beneficial
uses are restored to the satisfaction of the community and government agencies, delisting
is inappropriate.
We propose the following criteria be used to verify that all reasonable intervention
has been taken, and natural recovery is required prior to delisting. We also recommend
this be used as a starting point for discussion among RAP participants, the parties, the
jurisdictions, and the IJC:
All reasonable and practical implementation has been completed, with the tools
available in the present. What is reasonable and practical will vary in time, and
future opportunities may exist to further intervene and speed recovery.
The severity of the impairments will influence the rate of recovery. The time scale
for natural recovery, however, should be mutually agreed upon by all RAP participants.
The local public are to be satisfied that current conditions will respond to actions
implemented and that natural recovery will be monitored.
A pollution prevention or other maintenance plan is in place to reduce the risk of
future degradation and to ensure that natural recovery can proceed.
A process is in place to respond to future development pressures such that environmental recovery is sustainable and to implement new actions as technology
advances. As a corollary, contingency plans have been developed to further intervene if monitoring results warrant a more aggressive strategy.
Commitments to a monitoring plan and program are in place to measure progress
toward environmental restoration and a mechanism is established to report periodically to the public.
Delisting would occur at such time as experts and the public concur that the delisting
targets detailed by the AOC have been met. This places significant pressure on the parties
to design and commit to assessing ecological recovery and intervening further if predicted
recovery is not responding in the manner projected.
Status of the Canadian Areas of Concern

Across the Basin, federal, state, and provincial agencies are discussing policy and process
around delisting the areas of concern. No RAP has been completed yet in the United
States, and several RAPs are coming to completion in Canada. Completing RAP implementation, however, is not the same as restoring environmental quality. The focus on
delisting has, in some instances, detracted from the objective of incrementally measuring
gains in environmental recovery. Most recently, we have been examining methods to more
precisely quantify and illustrate the last two quadrants of the status pie diagram, that is,
estimating the extent to which recommended remedial actions have been implemented
(third quadrant) and the degree to which the restoration targets for the beneficial uses
have been attained (fourth quadrant). The point is to highlight obstacles and celebrate
progress at many levels in order to sustain long-term rehabilitation and recovery (WQB,
1996). To continue to restore the Great Lakes will require a clear articulation of the
accomplishments to date and the challenges remaining.
450

St.Lawrence River
Bay of Quinte
Port Hope
Metro Toronto
Hamilton Harbour
Niagara River
Wheatley Harbour
Detroit River
Detroit River
St. Clair River
Collingwood Harbour
Severn Sound
Spanish Harbour
St. Marys River
Peninsula Harbour
Jackfish Bay
Nipigon Bay
Thunder Bay
following a different process
Canadian side
Binational
20
40
60
80
100
Percent Implementation Complete/Percent

Implementation
Restoration
*RAP delisting targets are determined locally, and may or may

not be equivalent to complete restoration of beneficial uses
Figure II.6.3 Progress on RAP implementation and achievement of restoration targets. (From
Krantzberg, G., Aquatic Restoration in Canada, Backhuys Publications, 1999. With permission.)
Figure II.6.3 presents a synopsis of progress for each Area of Concern. The mean degree
of implementation across all AOCs by 1997 was 57% 35%, with the mean degree of
attainment of the restoration targets being 51% 33%. The values portrayed in Figure II.6.3
are based on an evaluation of progress on individual impairments at each AOC. This
evaluation is reported in the Remedial Action Plan Update (COA RAP Steering Committee,
1997) and by Krantzberg et al. (1997).
Progress varies with the nature of the impairment

It is instructive to examine the variability associated with progress in restoring the beneficial uses. Table II.6.1 presents the number of Canadian AOCs that have particular beneficial uses designated as either impaired or requiring further assessment. Those impairments associated with contaminated sediment and habitat destruction are pervasive across
the AOCs. Of particular note is that eutrophication remains to be a problem, despite the
present perception that conventional pollutants are no longer an issue and that toxic
chemicals must be the central focus for environmental researchers and managers. The
nearly singular discussion on toxic chemicals and Great Lakes environmental quality
minimizes the real threats posed by excessive nutrient and bacterial contamination and
neglects the overarching force of habitat despoilment in restructuring aquatic and nearshore terrestrial communities.
Table II.6.2 illustrates progress in restoring the impaired beneficial uses in the Canadian AOCs. In summarizing the current status for each impairment, Table II.6.2 notes in
parentheses the number of AOCs that, at the onset of the program, considered the beneficial use to be either impaired or potentially impaired. In several cases, it was uncertain
as to whether a particular beneficial use was impaired and further assessment was
required. The actual values we used to estimate progress in restoring beneficial uses only
considered cases where a change in the status of an impairment was due to implementing
a remedial action. The values, then, do not include uses that were considered not impaired
Chapter II.6:
451
Table II.6.1 Number of AOCs Reporting a Beneficial Use as Impaired

or Requiring Further Assessment as of 1997
Beneficial Use Impairment
Number of AOCs
Fish and wildlife consumption advisories

Tainting of fish and wildlife
Degraded fish and wildlife populations
Fish tumors or other deformities
Bird/animal deformities or reproduction problems
Degraded benthos
Restrictions on dredging
Eutrophication
Restrictions on drinking water
Beach closings
Degraded aesthetics
Added costs to agriculture and/or industry
Degraded plankton communities
9
2
9
5
1
12
11
10
3
10
9
2
4
12
Table II.6.2 Overall Progress on Restoring Restoration of Beneficial Users

Implementation and
Restoration <50%
No. AOCs Implementation and

Represented Restoration >50%
Fish tumors or other

deformities
Eutrophication of
undesirable algae
Beach closures
Degradation of
aesthetics
Degradation of
phytoplankton and
zooplankton
populations
(11), n = 5
(10), n = 10
(11), n = 10
(11), n = 0
(8), n = 4
Restrictions on fish and wildlife

consumption
Tainting of fish and wildlife
flavor
Degradation of fish and wildlife
populations
Bird or animal deformities,
reproductive problems
Restrictions on dredging
activities
Restrictions on drinking water
consumption; taste/odor
problems
No. AOCs Completely

Represented Restored
(15), n = 9
(5), n = 2
(13), n = 9
Added
cost to
agriculture
or industry
(4), n = 2
(16), n = 12
(8), n = 1
(14), n = 12
(17), n = 11
(5), n = 3
Note: Values in parentheses refer to the number of areas of concern that have addressed the beneficial use at
some point within the past 10 years. Where the value n used for statistical analysis is less than the
value in parentheses, it is due to clarifying whether or not a beneficial use was impairment and not a
result of remedial measures being taken.
based on a more in-depth assessment, instead of recovery in response to remedial actions.

As well, in some instances, some beneficial uses were originally designated as impaired
due to a misinterpretation of the intent of the GLWQA. This was the case, for example,
where RAPs assumed that exceeding a sediment chemical guideline meant that there were
restrictions on dredging, yet no dredging was necessary or predicted to take place at that
location. (Contaminants in sediment are considered in the context of impairments such
as degraded benthos, consumption advisories, and others.)
As illustrated in Table II.6.2, RAPs have made substantial progress in habitat rehabilitation. This is partly due to the tenet that informed individuals, groups, and clubs are
interested in getting involved and making a difference. Public involvement in RAPs has
452
July 1997
Beach closures
Restriction on drinking water; taste and odour problems
Bird or animal deformities, reproduction problems
Fish tumours or other deformities
Tainting of fish and wildlife flavour
0
Implementation
20
40
60
80
100
Restoration
Figure II.6.4 Mean degree of progress on restoring beneficial uses in the Canadian areas of concern.
(From Krantzberg, G., Aquatic Restoration in Canada, Backhuys Publications, 1999. With permission.)
been a major breakthrough for ecosystem management and recovery (Hartig and Zarull,
1992). A lesson learned is that RAP success depends on accomplishing visible elements,
such as relatively short-term, focused projects that contribute to habitat rehabilitation
(Figure II.6.4). A step-wise approach to restoring habitat achieves progressive gains in
ecological integrity.
With governments focusing on reducing loadings of chemicals from industrial sources,
extensive abatement activities have advanced point source control, resulting in declining
levels of contamination in water, sediment, and biota, as well as concurrent improvements
in benthic community structure, declines in deformities in wildlife, and fewer restrictions
on fish consumption in several AOCs (MOE, 1997).
Conversely, the diffuse nature of nutrient and bacterial inputs, from nonpoint source
inputs and combined sewer overflows, continues to require considerable effort. Without
exception, funding is the major concern of agencies and the public involved in implementing RAP recommendations (MacKenzie, 1996). The availability of adequate financial
resources to address infrastructure problems is a common concern across the RAPs. The
costs to municipalities of remediating and preventing further adverse stormwater quantity
and quality impacts are substantial. Although no province-wide expenditure estimates
are available, clean-up costs of stormwater-induced impacts at Ontarios 16 RAP sites were
estimated to be $2.5 billion. Local stormwater management programs can have annual
costs in excess of $5 million (Cameron et al., 2001). The current climate of provincial and
federal resource restraints means that limited resources are available to implement storm
and wastewater controls at a rate anticipated when RAPs began. This obstacle to progress
points to the need for research into new financing methodologies such as reported by
Apogee Research (1991, 1997) and Water Environment Federation (1994). Not withstanding
the above, considerable progress has been made in demonstrating new, cost-effective
technologies that are effective at controlling nutrients and bacterial loadings (COA, 1997).
Chapter II.6:
453
Development of an environmental quality index to measure change

As an alternative measure of environmental quality at the AOCs, we adapted a water
quality index based on an approach developed in British Columbia and described by
Rocchini and Swain (1995). The water quality index has features that quantify the number
of parameters that exceed a standard and the extent to which the standard is exceeded.
We substituted the number of parameters exceeded by the number of beneficial uses
impaired, and the extent by a qualitative measure of the degree to which the use is
impaired, based on a consensus among RAP practitioners for each of the Areas of Concern.
The higher the value of the index, the greater the degree of impairment. The index is
calculated as follows:
EQI =
(N 10 14)2 + (Din N )
where EQI is the environmental quality index; N is the number of beneficial uses impaired,
normalized to a scale of 1 to 10; D is the degree to which each of the uses is impaired, on
a scale of 1 to 10, with 10 being completely impaired. The maximum value for the index
is 14. Where an impairment is present but not due to local sources (such as fish consumption advisories), a full value of 10 (fully impaired) was assigned. Similarly, where further
assessment is still required to determine whether a beneficial use is impaired, we made
the assumption that the use was fully impaired.
For the impaired beneficial uses, the extent of recovery was based on our interview
data and consensus-based system. We used the assumption that at the point of beginning
the RAP in 1987, the average degree of impairment was 10 (fully impaired). If by 1997,
the degree of impairment was 40%, then the value for D was 4. With no clear rationale
for ranking the environmental or social importance of the beneficial uses, all were weighed
equally. Similarly, we considered the number of impairments to be equally important as
the extent to which they are impaired.
As illustrated in Figure II.6.5, considerable improvements in ecosystem quality have
been observed across the Basin. The mean environmental quality index in 1987 was 11.7,
with an improvement to 7.7 by 1997. Based on our knowledge of environmental conditions
at the AOCs, values less than 6.0 represent considerable recovery. Some extensive improvements are reflected by the index as noted at a number of AOCs.
For example, in Nipigon Bay, source control at Dow Chemical has virtually restored
benthic communities. Industrial upgrades have removed tainting problems, and the habitat strategy has dramatically improved fish populations (Figure II.6.6). The Environmental
Quality Index reflects these improvements, declining from 12 to 4.5. Due to substantial
investments and improvements in industrial effluents discharged by Omstead Foods to
Wheatley Harbour, beneficial uses are responding and the Environmental Quality Index
has declined from 10.5 to 2. Similarly, to infrastructure improvements, control of industrial
point sources and habitat enhancement in Spanish Harbour is mirrored by a decline in
the index from 11.2 to less than 2.
The Environmental Quality Index is just one further tool that can be used to portray
progress in restoring the AOCs. It diverts the focus from the singular milestone of delisting
to a continuous scale where actions that result in real environmental change can be documented.
Summary and recommendations

Measuring, recording, and reporting on progress in restoring beneficial use impairments
are a dynamic process, requiring continual reassessment as implementation status
454
St.Lawrence River
Bay of Quinte
Port Hope
Metro Toronto
Hamilton Harbour
Niagara River
Wheatley Harbour
Detroit River
St. Clair River
Collingwood Harbour
Severn Sound
Spanish Harbour
St. Marys River
Peninsula Harbour
Jackfish Bay
Nipigon Bay
Thunder Bay
10
12
14
degree of impairment
restored
impaired
index 87
index 97
Figure II.6.5 Environmental quality index for the areas of concern showing improvements from
1987 to 1997. Please refer to the text for an explanation of the calculation. (From Krantzberg, G.,
Aquatic Restoration in Canada, Backhuys Publications, 1999. With permission.)
changes. For example, in some of the AOCs, major projects are just getting under way
now that the environmental planning, assessment, or approval process has been completed. These activities were only marginally accounted for in this analysis, because actual
implementation had not yet proceeded and the substantial improvements to environmental quality have yet to be realized; however, the fact that a RAP has reached the point at
which a major project can be launched is itself a clear success. Other forms of progress
include the establishment of local, sustainable implementation teams to continue rehabilitation and protection initiatives for the long term. Multiagency and binational cooperative
arrangements are also markers of achievement that the RAP process enables.
This has been the first attempt to quantify environmental change across the Canadian
AOCs. It is intended in part to gives credit to thousands of individuals for their accomplishments. A record of success is an important element in maintaining momentum. For
many AOCs, the long-term strategic planning necessary to tackle complex problems, such
as infrastructure and sediment remediation, builds on the record of success. It also provides a view to where obstacles to progress lie.
It is important to provide an opportunity to motivate all partners to overcome the
challenges that remain to reach the goals and targets for delisting. The ability to demonstrate progress sustains public and political confidence and support (IJC, 1989). We recommend the public, the agencies, and the IJC press for such accounting.
Chapter II.6:
455
NIPIGON BAY
July 1997
Not
Impaired
Not
Impaired
Not
Beach closures
Impaired
Not
Restrictions on drinking water; taste and odour problems
Impaired
Not Impaired (ID)*
Not
Bird or animal deformities, reproduction problems
Impaired
Not Impaired
Fish tumours or other deformities
(AA)**
Tainting of fish and wildlife flavour
Impaired (NL)***
20
Implementation
Restoration
40
60
80
100
Percent Implementation Complete/Percent Restored
*RAP delisting targets are determined locally, and may or may not
* Not Impaired (ID) = Inappropriately designated 'Impaired'
be equivalent to complete restoration of beneficial uses
** Not Impaired (AA) = Redesignated based on additional

assessment
*** Impaired (NL) = not due to local sources
Figure II.6.6 Progress on restoring beneficial uses in Nipigon Bay. (From Krantzberg, G., Aquatic
Restoration in Canada, Backhuys Publications, 1999. With permission.)
Acknowledgments
The authors gratefully acknowledge the willingness of all those contacted to be candid in
their perspectives and for the wealth of information they provided. Their expert opinions
were invaluable in attempting to measure progress toward restoration of environmental
quality and identifying future challenges to restore and protect the Great Lakes Basin. RAP
Teams and Public Advisory Committee members all assisted in evaluating how they had
advanced over the past 10 years. The values assigned are theirs and not those of the authors.
Literature cited
Apogee Research, The user pay approach to stormwater management and its potential application
in Ontario, prepared for Environment Canada, 1991.
Apogee Research, Financing options for stormwater quantity and quality management in the RMOC,
prepared for the RMOC, Environment Canada, and the Ontario Min. Environ. Energy, 1997.
456
CanadaOntario Agreement (COA) respecting the Great Lakes Basin ecosystem. 1994 Cleanup Fund,
Great Lakes 2000 Cleanup Fund Project summaries report, Environment Canada, Downsview, ON, 1997.
COA RAP Steering Committee, The remedial action plan update, available from Ontario Ministry
of Environment, Great Lakes Programs, 40 St. Clair Ave. W., Toronto, ON, www.cciw.ca/
glimr/raps/intro.html, 1997.
Cameron, J. et al., User pay financing of stormwater management: A case study in Ottawa
Carleton, Ontario, submitted for publication in J. Environ. Manage., 2001.
Environment Canada and U.S. EPA, Progress in Great Lakes remedial action plans: Implementing
the ecosystem approach in Great Lakes areas of concern, Toronto and Chicago, 1994.
Hartig, J. and Zarull, M., Under RAPs, Towards Grassroots Ecological Democracy in the Great Lakes Basin,
Univ. Michigan Press, 1992.
International Joint Comm. (IJC), Revised Great Lakes Water Quality Agreement of 1978, as amended
by protocol signed in Ottawa, ON, Nov. 18, 1987, consolidated by the IJC, Jan. 1988.
IJC, Report on Great Lakes water quality, Great Lakes Water Quality Board, Windsor, ON, 1989.
Krantzberg, G., Ali, H., and Barnes, J., Status report on incremental progress in restoring beneficial
uses at the Canadian areas of concern, a site specific analysis, Ontario Min. Environ., Toronto,
ON, 1987.
MacKenzie, S.H., Integrated Resource Planning and Management, Island Press, 1996.
MOE, Guide to eating Ontario sport fish, Ontario Min. Environ. and Energy, and Ontario Min. Nat.
Resour., PIBS 590B07, 1997.
Rocchini, R. and Swain, L.G., The British Columbia water quality index, Water Quality Branch,
Environ. Protection Dept., B.C. Mini. Environ., Lands and Parks, 1995.
Water Environment Federation, User-Fee-Funded Stormwater Utilities, Alexandria, Va, 1994.
Water Quality Board (WQB), Review and Evaluation of the Great Lakes Remedial Action Plan Program,
IJC, Windsor, ON, 1991.
WQB, Position Statement on the Future of Great Lakes Remedial Action Plans, IJC, Windsor, ON, 1996.
Chapter II.6:
457
Response
Measuring recovery of impaired waters
The Great Lakes Basin is the largest freshwater body in the world and is one of the most
highly populated areas on the North American continent. As a result, the specific locations
where the two meet the industrial harbors have become degraded. This study demonstrates that innovative solutions are needed and are possible in even the most complicated of situations; complicated in terms of jurisdictional complexity involving multiple
agencies across binational borders.
One of the take-away lessons from this chapter by Gail Krantzberg and Judi Barnes
is in identifying the benefits ensuing from prioritizing sites to be targeted for recovery
efforts the identified areas of concern. By concentrating efforts on those particular
sites of greatest impaired beneficial use, a step-wise approach can be adopted that can
lead to progressive gains in ecological integrity. These improvements, first felt for the
immediate sites, will, with time, spread to the entire system due to the ability of water to
carry benefits (as well as, of course, pollutants) far. The importance of this message cannot
be overemphasized. Given the magnitude of ailments plaguing our waterbodies, it is all
too easy to become disheartened and dismayed at the restorative jobs that need to be
done. This chapter by Krantzberg and Barnes provides a vivid demonstration supporting
one of the cardinal tenets of the EPAs Top 10 Watershed Lessons Learned namely, watershed management works best when undertaken in incremental steps (e.g., see Nicholas
Pouder and Robert France in Chapter I.16).
Similar to many developed watersheds, those of the Great Lakes Basin faces problems
of waterfront habitat rehabilitation, aquatic restoration, and pollutant source reduction.
As also observed by Kelly Cave in Chapter II.9 for Detroit, Krantzberg and Barnes record
that the slowest progress has been made in dealing with the elusive or occult in the
true sense of the word, nonpoint source pollutants. Nevertheless, the common sense
approach they describe for measuring progress in their efforts toward recovery, termed
RAPs, or remedial actions plans, is commendable and could easily be adapted elsewhere: problem definition restoration plan implementation framework development
adoption of proposed actions monitoring recovery. And finally, the generation of an
environmental quality index is an important tool to help document and communicate
accomplishments and impediments to the public.
chapter II.7
Watershed management plans:

Bridging from science to policy to
operations (San Francisco, California)
David Blau
Abstract
Watershed management plans are nearing completion for San Franciscos two urban
watersheds the 23,000-acre Peninsula watershed and the 40,000-acre Alameda watershed. The watershed lands surround five storage reservoirs that provide drinking water
to 2.3 million Bay Area customers.
The watershed plans are the result of an in-depth, 9-year planning process that has
included extensive community outreach and agency involvement. A comprehensive database was assembled for each watershed and mapped using Geographic Information System (GIS) technology. This data set was then used to create a watershed tool kit comprised
of five parts:
1.
2.
3.
4.
5.
Water quality vulnerability

Ecological sensitivity
Erosion sensitivity
Cultural resource sensitivity
Fire hazard
The tool kit is serving as the foundation for the long-range management plan and will
also serve as a day-to-day reference base for decisions on the compatibility of land-use
proposals in the future. The planning process has included cutting-edge work on such
subjects as water quality vulnerability analysis and modeling; fire hazard analysis and
modeling; and grazing management. The grazing management plan was a particularly
challenging component of the work, considering the sensitivity of Bay Area residents to
risks of exposure to Cryptosporidium.
Unlike many watershed management plans that stop at the policy statement level,
this plan attempts to translate broad policy goals and objectives into specific management
actions and then down to day-to-day operation and maintenance practices. One product
of the effort, for example, is a users manual for the watershed keepers in the field.
The plan document is structured in such a way that a reader can easily track from
policy to actions to operations. The document is generously illustrated with photographs
1-56670-562-2/02/$0.00+$1.50
459
460
of watershed conditions, GIS data maps and analytical tools, planning process diagrams,
and plan recommendations.
Background
More than 130 years ago, the predecessor of the San Francisco Public Utilities Commission
(SFPUC), Spring Valley Water Works, had a vision of protected watershed lands that would
provide a pure and reliable drinking water supply from the developing economy of San
Francisco. In the last half of the 19th century, Spring Valley Water Works began purchasing
the watershed lands that are now managed by the SFPUC. They first acquired the
23,000-acre San Francisco Peninsula and then the 40,000-acre Alameda Creek watershed
in the East Bay. Today, these two watersheds remain largely protected and continue to
serve their primary purpose to collect and store a reliable supply of high-quality water
for the homes and businesses in the San Francisco Bay Area.
The SFPUCs mission for managing their two urban watersheds the Peninsula and
Alameda watersheds is to provide the best environment for the production, collection,
and storage of the highest quality water for the City and County of San Francisco and
suburban customers. The SFPUC seeks to accomplish this by developing, implementing
and monitoring a resource management program which addresses all watershed activities. The watershed management plan applies best management practices for the protection of water and natural resources and their conservation, enhancement restoration,
and maintenance while balancing financial costs and benefits. The Peninsula and
Alameda watershed management plans were prepared in response to this mission statement and because existing SFPUC policies do not address the management of watershed
lands in a comprehensive or integrated manner.
The plans provide a policy framework for the SFPUC to make consistent decisions
about the activities, practices, and procedures that are appropriate on SFPUC watershed
lands. To aid the SFPUC in its decision making, the plans provide a comprehensive set of
goals, policies, and management actions that integrate all watershed resources and reflect
the unique qualities of the watersheds.
In addition to serving as a long-term regulatory framework for decision making by
the SFPUC, the plans are also intended to be used as watershed implementation guides
by the SFPUCs Land and Resources Management Section (LRMS) staff. The plans provide
the LMRS manager and staff with management actions designed to implement and establish the goals and policies for water quality, water supply, ecological and cultural resource
protection, fire and safety management, watershed activities, public awareness, and revenue enhancement. The plans also enable LRMS staff to address and plan for future
management issues such as fire management, erosion control, range management, public
access, security, development encroachment, and ecological resource management.
In these plans, the SFPUC has taken an approach to watershed management that
considers water quality protection as the first and foremost goal. The primary goal is to:
Maintain and improve source water quality to protect public health and safety
In addition to the primary goal, the following six secondary goals are also supported
by the plans policies and management actions:
Maximize water supply
Preserve and enhance the ecological and cultural resources of the watershed
Protect the watersheds, adjacent urban areas, and the public from fire and other
hazards
Chapter II.7:
Watershed management plans: Bridging from science to policy to operations 461
Continue existing compatible uses and provide opportunities for potential compatible
uses on watershed lands, including educational, recreational, and scientific uses
Provide a fiscal framework that balances the financial resources, revenue-generating activities and overall benefits, and an administrative framework that allows
implementation of the watershed management plans
Enhance public awareness of the water quality, water supply, conservation, and
watershed protection issues.
The planning process and public involvement

The watershed management planning process commenced in August 1992 and, when
complete, it will have spanned nine years. The process addressed planning for both the
Peninsula and Alameda watersheds simultaneously, allowing for similar goals and policies
to be established for all of the SFPUCs local watershed lands. The process will culminate
with the completion of the final Peninsula and Alameda watershed management plans,
as well as the EIRs that will evaluate the environmental impacts of the plans in compliance
with the California Environmental Quality Act (CEQA). The planning process consists of
the seven stages described next, as well as an extensive, ongoing public and agency
participation program.
Establish goals
One primary and six secondary goals for watershed management were established at the
outset of the project by the Watershed Planning Committee (WPC), a group of SFPUC
division and department representatives who assisted the planning team with the plan
development and review. These goals are listed above and were used by the planning
team throughout the planning process to provide direction for alternative and plan development. The goals serve as a foundation for the plans policies and management actions
and will also serve as a basis for ongoing evaluation of plan implementation.
Assemble database and prepare resource vulnerability maps

Mapping of watershed resource information was conducted on the SFPUCs GIS. Each
resource type (e.g., vegetation, wildlife, etc.) was entered into the GIS system and became
a separate map (or layer). Selected layers were then sandwiched together to provide
information-rich composite maps. A set of resource vulnerability/sensitivity maps was
also created for each watershed.
Formulate alternatives
The analysis of water quality, natural resources, cultural resources, and wildfire severity
data gathered for the watershed was incorporated with the public comments and public
survey results to form three watershed management alternatives. These alternatives
applied to the management of both watersheds and were used to explore the range of
options between a totally closed and totally open watershed.
Evaluate alternatives and select preferred plan

The three alternatives were evaluated against the primary and secondary goals and the
requirements set by the various agencies with jurisdiction over the watershed. The alternatives were also presented at public, agency, and staff workshops. Although those in
462
attendance at the public workshops initially preferred a more permissive approach to the
management of the watersheds, the resource data, the public survey, and input from the
agencies pushed the preferred plan toward a more protectionist approach that stresses
controlled access to the watersheds, provides an improvement in water quality, and gives
a balance among ecological resource protection, water supply needs, and watershed activities.
As a result of the alternatives evaluation process, the watershed management plan
was formed and approved through and SFPUC resolution in January 1995. The preferred
plan, as it was approved in January 1995, applied to both watersheds. Subsequent amendments to the Peninsula watershed management plan include the Southern Peninsula
Watershed Golf Course Element (March 1997) and the Fifield/Cahill Ridge Trail Element
(June 1997). Subsequent amendments to the Alameda watershed management plan
included the Sunol Valley Resources Management Element (May 1996) and the Alameda
Watershed Grazing Resources Management Element (July 1997).
Prepare management plans and EIRs

The general direction provided by the SFPUC in the preferred plan has been developed
into two specific watershed management plans the Peninsula watershed management
plan and the Alameda watershed management plan. Each plan provides policies for
decision making and actions for day-to-day management, which are specific to the character and resources of each watershed.
The environmental impacts of each plan were evaluated in a programmatic EIR, along
with an evaluation of the other alternatives. The results of the EIRs necessitated some
revisions to the plans and required that mitigation measures be incorporated into the final
plans.
Public, agency, and staff participation program

An extensive outreach program was developed at the outset of the project and included
interviews with key stakeholders; public, agency, and SFPUC staff workshops; newsletters;
a public opinion survey; and development of a watershed Web page. The program was
designed to elicit information and opinions from the public, SFPUC staff, and agency
representatives, as well as to keep these interested parties abreast of the process as it
proceeded.
Interviews
Early in the planning process, 60 individual interviews were conducted with various
public agencies, local water agencies, adjacent cities and counties, neighborhood organizations, and special interest groups (e.g., golf, fishing, trails, mountain biking, and equestrian). The issues and concerns raised at these interviews were carried forward as the key
issues to be addressed during the planning process.
Public, agency, and staff workshops

At six key points during the planning process, a series of workshops were held with the
public. An agency workshop and a workshop with SFPUC staff were also held at these
points in the planning process. These workshops allowed the planning team to present
new information to the public, staff, and agencies, while receiving comments at each stage
in the process. During the course of the project, meetings focused on the finding of the
resource inventories and the GIS mapping, the management alternatives developed for
each watershed, identification of a preferred alternative, and EIR scoping. Additional
Chapter II.7:
workshops were held to address issues pertaining to specific plan elements, including the
management of the Sunol Valley resources, the Southern Peninsula Watershed Golf
Course, the Fifield/Cahill Ridge Trail, equestrian uses, and grazing on the Alameda
watershed.
Newsletters
Six newsletters were prepared to announce upcoming workshops and to summarize key
issues related to watershed planning. Newsletters may be requested from the LRMS
division of the PUC. Topics addressed include a summary of public opinions heard at the
first series of workshops; an overview of data collected on each watershed; presentation
of the watershed management alternatives; and a presentation of the preferred alternative,
as well as articles on critical watershed issues, such as quality protection and fire.
Public opinion survey

A statistically valid telephone survey of 578 randomly selected households, all of which
receive water from the SFPUC, was conducted in September 1993. The purpose of the
survey was to ascertain the opinions of these SFPUC consumers on issues of water quality,
the watershed management goals, recreational access to the watersheds, environmental
protection, financing, and other issues related to watershed management. The survey
results were used in conjunction with all of the other data and information collected to
develop the watershed management alternatives.
The results of the customer survey tended to be more protective in terms of water
quality and watershed management than the opinions voiced at the workshops. For
example, 71% of the respondents stated that ensuring water quality was the most important goal of watershed management, while only 3% indicated that providing recreational
access was the most important goal; 85% of the respondents wanted the same or less
public access than is currently provided. Respondents were relatively restrictive in the
type of public access they would allow. A majority of respondents supported natural
resource studies (92%), jogging (81%), hiking (75%), and guided tours (75%). Less than
half of the survey respondents would allow mountain biking (41%), fishing (28%), or golf
courses (26%).
Web page
A watershed Web page (www.ci.sf.ca.us/puclrms) was developed, allowing users to
review the plan online and download portions of the plans that are of interest to them.
The technology also allows the Web page user to view the GIS maps. The Web page
includes information on upcoming public involvement opportunities, directions for applying for access permits, and general information on the watersheds.
Technical work: Baseline and analysis

A tremendous array of resource data was compiled as input to the watershed management
plans. The information served as the basis for the goals, policies, and management actions
set forth in the plans.
GIS database
During the spring and summer of 1993, scientists from the EDAW team surveyed the
Peninsula and Alameda watersheds to identify and map the existing natural and cultural
resources. In addition to the field studies, existing data prior to watershed studies were
464
Figure II.7.1 Example of GIS database surveying the Peninsula watershed. (Color version available
also reviewed. Data maps were prepared that identified the location and the extent of
each resource, and data tables were created to indicate the special characteristics of each
resource. GIS layers prepared as part of this project included vegetation, soils, slopes,
aspect, geology and faults, mineral resources, landslide susceptibility, protected species
habitat, water contamination sources, cultural resources, aquatic habitat, wildlife habitat,
wildfire fuels, wildfire hazards, land use, lease areas, and protected plans. In addition,
specific maps were created to analyze and convey information on specific topics such as
grazing and grazing management on the Alameda watershed and trail access on the
Peninsula watershed (see Figure II.7.1 as an example of the GIS database).
Land-use/water quality relationship models

Because the primary goal of the watershed management plans is to protect and enhance
water quality, a tool to convey and better understand the relationship between land-use
activities and water quality was developed. A series of land-use/water quality impact
models was developed for each existing and anticipated activity, and the potential construction, maintenance, and operational impacts were defined. Activities that were delineated in this manner included golf courses, trails, grazing, and fishing.
Chapter II.7:
Ultimate water quality impacts were divided into six general categories turbidity
and sediment; microorganisms; THM precursors; VOCs, SOC, and pesticides; nitrogen,
phosphorus, and algae; and heavy metals.
One of the most important outcomes of these models was the realization that the risk
to water quality associated with various watershed activities was in certain cases very
high but very difficult to quantify. The outcome of this analysis was either to prohibit
activities that posed too great a risk to water quality or to develop policies and management actions that reduce the impacts of a particular activity to an acceptable level of risk.
Composite vulnerability maps and the watershed tool kit

Resource vulnerability and sensitivity zones were identified using the GIS and compiled
into a tool kit for each watershed. Each tool kit consisted of a set of five maps: the water
quality vulnerability map, the composite ecological sensitivity map, the cultural resource
sensitivity map, the erosion and land instability map, and the wildfire severity map. These
maps were used to identify those areas of the watersheds most sensitive or vulnerable to
disturbance and therefore least suitable for accommodating watershed activities. Zones
of high vulnerability/sensitivity are areas in which an activity, use, or facility is most likely
to have a negative impact on a resource.
The high-vulnerability/sensitivity zones from each of the individual tool kit maps
were combined to create the composite high-sensitivity zones shown in Figure II.7.2 for
the Peninsula watershed and Figure II.7.3 for the Alameda watershed. The composite maps
illustrate that few places within either watershed are completely free of resource vulnerability/sensitivity, and in many areas two or more of the high-resource sensitivity areas
overlap, indicating that activities in most locations on the watershed are likely to have an
impact on at least one, if not multiple, resources.
Development of the water quality vulnerability zones

A system for identifying those areas of the watershed where activities or disturbance have
the greatest potential to impact water quality was developed specifically for this project.
Development of these zones, termed water quality vulnerability zones (WQVZs), involved
extensive use of the GIS. The key criteria used to develop these zones were proximity to
water varied by intensity of rainfall, wildlife concentrations, vegetation, slope, and soils.
Initially, individual water quality vulnerability zone maps were prepared for microorganisms, particulates, and other groups of substances known to impact water quality. These
individual water quality vulnerability zone maps were then combined to create a composite WQVZ map for each watershed (see Figure II.7.4 for an example).
Development of the resource sensitivity zones

The sensitivity maps for ecological resources, cultural resources, erosion and land instability, and wildfire severity were developed in much of the same manner as the WQVZs.
The information for each individual data layer (e.g., vegetation) was assigned a ranking
of high, medium, or low sensitivity. All areas of high-sensitivity were assigned a high
sensitivity ranking on the composite maps, while a numeric ranking system was designed
for the areas of medium and low sensitivity, which, depending on the density of resources
in one area, ended up with an overall ranking of high, medium, or low sensitivity in the
composite map. These composite maps indicate that in most cases, the watersheds include
a significant amount of sensitive lands for at least one of the resource types (see Figure II.7.5
for the wildfire severity map).
466
Figure II.7.2 GIS analysis of composite high-sensitivity zones of the Peninsula watershed. (Color
Key planning issues

The key watershed management issues facing the SFPUC include fire hazard management,
road management, trail management, golf course management, and control over upstream
activities and urbanization. The watershed management plans strive to address these
issues through the development of sound management actions, which SFPUC staff can
implement.
Fire hazard management

Key issues related to fire hazard management include a significant buildup of fuels, the
fear held by neighboring communities of controlled burns, the stringent air quality requirements established by the Regional Air Quality Board for controlled burns, the difficulty
in balancing watershed activities with the need for fire protection, and the extensive fuel
management activities that must take place to reduce the fire hazard on the watersheds.
Because fire is an extremely complex issue, fire management plans were completed
for each watershed. These plans provided detailed actions developed to address the above
issues and included a prioritized list of necessary treatments and treatment intervals.
Recommendations for staff training and equipment purchase were also included. The GIS
Chapter II.7:
Figure II.7.3 GIS analysis of composite high-sensitivity zones of the Alameda watershed. (Color
was used extensively to prepare models of fires and fire movement, given certain temperatures and wind conditions.
Road management
At present, roads are the single greatest, ongoing contributor of the sediment on the
watersheds. Sedimentation and related erosion lead to treatment difficulties and may
result in drinking water taste and odor problems. The plans include extensive management
actions aimed at limiting the construction of new roads, closing and revegetating unnecessary roads, and establishing best management practices (BMPs) for road maintenance
and management. Because of the impacts roads can have on water quality, the road-related
management actions were given first priority in the phasing of the plan actions.
Trail management
Key trail management issues are related to the Fifield/Cahill Ridge Trail, which was
recently incorporated into the Peninsula watershed management plan based on a request
468
Figure II.7.4 Composite water quality vulnerability zones. (Color version available at www.gsd.
from the San Francisco Board of Supervisors. Because the trail passes through a high fire
hazard area of the watershed, provides access to many linking watershed service roads, and
passes through potential significant habitat areas, the trail poses significant management
challenges. The trail proposal put forth by the trail proponents calls for unrestricted access
for hikers, bicyclists, and equestrians. The Peninsula watershed management plan includes
two alternatives to the unrestricted access proposal: docent-led access or access by annual
permit, with an electronic gate tracking system. Both alternatives allow the SFPUC greater
control over the number of users and greater control over off-trail use. These two alternatives
also provide the SFPUC with the opportunity to close the trail or reduce the number of trail
users, based on monitoring results, to protect the resources from damage. All three alternatives were analyzed for their impacts in the Peninsula watershed management plan EIR.
Golf course management

In January 1995, the SFPUC adopted the watershed management preferred alternative for
the Peninsula and Alameda watersheds. The preferred alternative was based on the studies
of the watersheds, their resources, and the sensitivity and vulnerability of these resources.
The watershed management preferred alternative, approved in 1995, called for retention
of existing golf courses, consideration of golf course expansion in areas of low sensitivity
and vulnerability, and no new golf courses.
Chapter II.7:
Figure II.7.5 Wildfire severity map. (Color version available at www.gsd.harvard.edu/watercolors.

On March 4, 1997, the SFPUC held a public hearing to consider revisions to the preferred
alternative. After extensive discussion and review of citizen testimony, the SFPUC revised
the preferred alternative to consider a golf course on the southern Peninsula watershed.
Potential issues and concerns related to the development of the golf course are its
construction and operation, including the potential impacts to water quality from pesticides, herbicides, and fertilizers; the location of the site in the high- and medium-water
quality vulnerability zones; increased paved and impervious surfaces, as well as extensive
earthmoving, which may contribute particulates to water bodies. Of perhaps greatest
concern from a water quality standpoint is the fact that human error and failure of the
systems in place to protect water quality can, and do, occur and can lead to direct
contamination of drinking water supplies.
Water supply concerns include the large amount of water traditionally used by golf
courses, which reduces the supply available for drinking water. Ecological resource concerns include the loss of complexity of the original ecosystem due to extensive grading;
the disruption of habitat that impacts migration corridors; and the impacts of golf course
chemicals on aquatic species. The additional traffic generated by the golf course could
generate safety, air quality, and traffic concerns.
The Peninsula watershed management plan includes a number of policies that require
the golf course to adhere to strict standards in its development, construction, and operation.
470
These policies are intended to guide SFPUC decision making regarding the golf course
and assure that if a golf course is ultimately approved, water quality and ecological
resources are protected during its construction and operation. Key policies relate to the
containment of runoff, the lining of water impoundments, the implementation of integrated pest management strategies, and the requirements to adhere to environmental
design standards for golf courses. Other key policies require that the costs of operating
and monitoring the golf course are not borne by the rate-payers, and that adequate staff
and equipment be allocated to adequately monitor the construction and operation of the
course. Management actions were developed and intended for use if the project is
approved and related primarily to project-level studies and ongoing monitoring.
The golf course was analyzed at the programmatic level in the Peninsula watershed
management plan EIR. Should the impacts at this level be found to be insignificant, the
golf course may proceed to project-level design and environmental review.
Upstream activities
Upstream activities are of great concern on the Alameda watershed, as the upper twothirds of the watershed is not owned by the SFPUC. Current uses on these lands include
cattle grazing and residential development. Cattle grazing raises concerns about pathogens
entering the creeks and eventually the SFPUC water bodies. The lack of streamside protection and the potential for erosion and particulates to enter the creeks and water bodies
is also of concern. Residential development issues are related to septic systems and their
possible failure, and the improper use and disposal of household chemicals, both of which
raise water quality concerns. Domesticated pets are also a concern because of the disruption of wildlife and their habitats.
Because the SFPUC has no direct control over these lands, management actions were
developed to encourage cooperation and collaboration among the various agencies involved
in managing the upstream lands. Key agencies include the two counties with jurisdiction
of the upper watershed and the local resource conservation district. Actions were also
developed to assist in the development of educational materials to help upstream landowners understand the relationship between the activities they undertake and water
quality.
Management challenges
Three major challenges face the land managers at the SFPUC as they strive to manage the
watersheds in an integrated and responsible manner, described next.
Adjusting to changes in policy direction over time

Comprehensive watershed management planning is a multiyear process. At the staff level,
flexibility, continuity, and endurance are needed to adjust to changing decisions and to
keep various components of the plan in sync as pieces change.
As previously discussed, the development of watershed plans will span 9 years at the
completion of the process. During this time, major components of the plans were reconsidered, resulting in a number of significant changes. For example, the watershed management preferred plan, approved by the SFPUC in January of 1995, was developed to
provide the greatest protection to water quality and watershed resources. This plan prohibited new golf courses and unsupervised access to internal trails; however, the prohibitions on new golf courses and access to internal trails were reconsidered based on the
long history of these issues, direction from the San Francisco Board of Supervisors, and
Chapter II.7:
additional public testimony. This resulted in amendments to the preferred alternative to

include the southern Peninsula watershed golf course and the Fifield/Cahill Ridge Trail
in the Peninsula watershed management plan.
A number of implications are associated with this level of change:
1. It cannot be assumed that because the players, the opinions, and the facts remain
the same, the policy decisions will be stable. Instead, as the process proceeds, the
arena where decisions are made changes. In one setting, technical data and public
opinion surveys may carry a lot of weight, while in another, the relative political
power of various parties will have more of an impact.
2. Schedules are impacted as proposed changes to the plan are analyzed and subsequently made.
3. Changes to one part of the plan have implications for other parts of the plan. Some
of these are not immediately obvious and take time to consider.
4. Inevitably the process moves from stages dominated by technical data and staff
to arenas where political and financial factors become more significant. Whereas
presumably all changes stay consistent with the overall goals of the plan, the
specifics may or may not match the recommendations of the technical staff, who
support and manage the process. At different stages, staff at different levels of the
organization become actively involved in the process. This also results in changes
of direction. While the plan itself is changing, staff turnover also occurs. One of
the challenges of managing this kind of project is to maintain staff-level commitment to the process and product when continuity is lacking in both the staffing of
the project and significant aspects of the direction of the project.
5. The cycle of a utility planning project begins with the utility staff, goes through a
series of stages, and returns to the staff for implementation. If, during the interim
stages, the staff loses a sense of ownership in the plan, implementation will be
impaired.
Short-term decision making versus long-term planning

Although it takes years to finalize and publish a watershed management plan, the planning process itself affects the way the watershed is managed. Change does not begin when
the plan is published. If staff are adequately involved in the planning process, change is
initiated by the planning process.
Early in the watershed planning process, it was felt that leasing decisions should be
deferred until the plan was finalized. However, as the process moved forward, which
involved delays to handle difficult decisions and changes in direction, it became apparent
that many decisions needed to be made prior to the publication of the final plan. For
example, the Crystal Springs Golf Course lease was due to expire and new lease terms and
conditions were needed. The information developed through the planning process was used
to create a new lease with new provisions relating to water quality and watershed protection.
Grazing on the Alameda watershed also became an issue during the planning process,
with members of the public and certain decision makers requesting an end to grazing on
the watershed. This request was due to concerns over the pathogen Cryptosporidium, which
causes severe stomach distress and is of concern in San Francisco due to its high percentage
of individuals with compromised immune systems.
Resource specialists and the grazing community felt that grazing was vital to the
health of the ecosystem, because it reduced fire hazard on the watershed and could be
managed in such a way as to significantly reduce the risk of pathogens entering the water
bodies.
472
Working closely with the local resource conservation district and members of the
community opposed to grazing, SFPUC staff and the consultant team developed a grazing
management plan that retained grazing while introducing lease provisions, which provided much greater protection for water quality. Timing of calving, development of exclosures around streams and waterbodies, and limits on the number of cattle grazed were
all included in the new lease provisions. Significant penalties were also included if the
lease provisions were violated. The plan was especially significant in that it brought
together a number of opposing groups to develop a constructive solution that was agreeable to all parties.
Implementation and funding

Watershed management is not only about what you allow others to do to your watershed;
it is also about what you do to your watershed. Therefore, you need to provide the tools
and guidance that will enable staff to provide a higher level of watershed protection and
maintenance.
Implementation of the plans will require additional funding, additional staffing, additional training, and development procedure manuals for staff. This commitment is essential to achieve the goals of the plan. A number of steps have been taken to ensure that
this occurs.
A training manual for the staff has been prepared based on the plan recommendations
and will be provided to all staff involved in watershed work. The goal of the manual is
to educate staff so that they may better protect water quality and the watersheds natural
resources as part of their day-to-day activities. It is intended for use by nonresource specialists
and will assist watershed workers in defining areas of high-sensitivity/vulnerability,
locations of special status species that should not be disturbed, etc. (Figure II.7.6 is an
illustration from the training manual.)
Training sessions are also planned for staff, which will familiarize watershed workers
with the plans and provide them with the tools necessary to do their jobs more effectively,
while protecting water quality and the natural resources.
Within each management plan, the management actions have been prioritized to assist
staff with implementation. Priorities were based on staff input and the actions that were
determined to be the most essential to water quality protection and improvement. The
staff will be able to include several of these prioritized actions in each years budget
request, which ensures that plan implementation continues.
Certain implementation steps began early in the project. One of these was the purchase
of a GIS system for the Land Resources Management Section (LRMS). The consultant team
helped the LRMS select a watershed GIS system best suited to its needs, as well as
interfaced well with both the consultants GIS and other GIS systems within the city. The
consultant then set up the system and trained staff to work on the GIS in order to respond
to map queries and update the LRMS internal database. This system now resides with
the LRMS and continues to be used extensively to respond to data requests, to provide
maps, and to periodically update the resources layers based on new information.
Conclusions
The key to a successful watershed management plan is implementation. Implementation
generally occurs following plan adoption, but often the need to respond to an urgent issue
or problem requires action during the planning process. This is especially true in a multiyear watershed management planning process, where new issues and concerns arise
frequently.
Chapter II.7:
Figure II.7.6 Illustration from the training manual based on plan recommendations. (Color version available at www.gsd.
474
It is important to decide which issues can and should be addressed immediately, and
which should wait until completion of the plan and the environmental review process.
Issues that pose a threat to the health and safety of the consumer, or another significant
risk, are best addressed as they arise. Issues that do not pose a health or safety risk and
that may have significant negative environmental consequences, or may be highly controversial, may not be appropriate to address prior to plan completion and environmental
review. Often, these issues require additional time to build consensus, collect additional
data, develop alternatives, and identify a final solution.
Implementation during the course of the planning process is most successful once
data collection is complete and the key issues have been identified. This information can
be used to develop interim plans to address urgent health and safety issues that must be
addressed prior to plan completion.
Implementation following plan adoption requires ongoing dedication from the staff
assigned to implementing the plan. Staff and managers must be provocative in securing
funds, training staff in new activities and methods to execute plan recommendations, and
follow-through to get the job done.
A system to evaluate plan recommendations against accomplishments on an ongoing
basis is also critical to successful implementation. The plans policies, because they provide
guidance for long-term decision making, are generally fixed. The actions to implement
the policies and the means and methods to implement them, however, are changing as
technology, legislation, and understanding of the resource change. Ongoing evaluation of
plan actions and accomplishments allows watershed staff to adjust the actions, and the
means and methods to accomplish them, in order to fit the current state of the art.
Chapter II.7:
Response
Sociology of implementing adaptive management
A primary goal of many watershed management plans is to protect and enhance water quality
by using models that analyze the risks associated with various land-use activities (see, for
example, Chapter II.8 by Dennis Haag, Stephen Hurst, and Bryan Bear, Chapter II.12 by
Jeffrey Schloss, Chapter II.13 by Margot Cantwell, and Chapter II.17 by Neil Hutchinson).
This chapter by David Blau identifies a critical aspect in implementing successful watershed plans namely, that comprehensive water sensitive planning is a multiyear process
that needs to accommodate potential organizational changes by factoring in flexibility
through adaptive management. Although many watershed management plans stop at the
policy statement level, the one that Blau describes shows how production of a users
manual tool kit provides one means of translating policy goals and objectives into specific
management actions and operations. Two cardinal tenets are included in Blaus message:
first, clear goals are essential at the get-go; second, successful watershed planning is more
than just science (i.e., it is policy and implementation as well).
This watershed tool kit is spatially based, wherein water quality, ecological, and
erosion sensitivities are considered along with cultural resource sites and fire hazards. All
are integrated with various land uses to prioritize areas of vulnerability. The tool kit is
designed to be used for both long-term forecasting as well as day-to-day decision making.
A final element of interest in this particular study includes the attention paid to arrive
at balanced social goals for the watersheds. Approaches such as use of a Web site, interviews with stakeholders, newsletters, special-interest workshops, and general public surveys go far toward setting the stage for adaptively managing these watersheds through
time in a truly sustainable fashion, as, for example, that outlined by Daniel Williams in
Chapter II.11.
chapter II.8
Watershed assessment planning process

(Johnson County, Kansas)
Dennis A. Haag, Stephen A. Hurst, and Bryan J. Bear
Abstract
The Kansas Urban Resource Assessment Project (URAMP) was established in 1997 to
evaluate the effects of urban development policies and practices on riparian and wetland
areas at the watershed level. URAMP was a pilot study funded by a U.S. EPA grant administered by the Kansas Water Office, with assistance from George Butler Associates, a hired
consultant, and support from several state and federal agencies. The study involved a comparison of two watersheds, one urban and the other rural, located in Johnson County, KS.
Many of the URAMP conclusions and recommendations have been incorporated into
new watershed planning and development policies and practices being implemented at
the state and local levels of government. This paper discusses the current status of watershed planning in Johnson County. It presents a case history of a developing suburban
residential community that required stormwater variances to allow for the construction
of adjacent wetland, riparian, and park lands. The wetland and riparian areas are being
installed as part of a residential housing development to provide multiple uses that include
mitigation bank, stormwater treatment, science education, public recreation, and residential housing amenity. The project will be discussed from three planning and permitting
perspectives: (1) private developer/engineer; (2) state water resource planning/Kansas
Water Office; and (3) local planning/city of Overland Park, KS.
Introduction and background

In 1996 and 1997, the Kansas Water Office, the states water policy and planning agency,
completed an innovative project sponsored by the U.S. EPA Wetland Protection Grant Program. This project, the Urban Resource Assessment and Management Project (URAMP), was
conducted in Johnson County, KS, a fast-growing suburban area near Kansas City, MO,
(Figure II.8.1). It brought together federal, state, and city planners and private developers to
explore the impacts of urbanization on wetland and riparian resources.
URAMP examined two watersheds Tomahawk Creek, an almost fully developed
urban watershed, and Wolf Creek, a rural watershed that was just beginning development
(Figure II.8.2). The goal was to bring together a group of diverse stakeholders and, through
a collaborative, team-building process, explore alternatives to the we have always done
it this way approach to natural resource management.
1-56670-562-2/02/$0.00+$1.50
477
478
Johnson County
Kansas
Figure II.8.1 The project is located in Johnson County, KS, one of the fastest growing suburban
communities in the Kansas City, MO, metro area.
Johnson County Watershed Map

BONNER
SPRINGS
CedarMill
Creek
ROELAND
PARK
LAKE
QUIVRA
Todley
Creek
Turkey
Creek
MERRIAM
SHAWNEE
Mill Creek
CaptainKill
Creek
MISSION
WESTWOOD
WESTWOOD
HILLS
MISSION
WOODS
FAIRWAY
COUNTRYSIDE
MISSION
Brush
HILLS
Creek
PRAIRIE
VILLAGE
DeSOTO
LENEXA
OVERLAND PARK
Middle
Indian
Creek
Captain
Creek
Lower
Indian
Creek
Cedar Creek
TOMAHAWK CREEK
Kill Creek
OLATHE
Upper
Indian
Creek
GARDNER
Bull Creek
EDGERTON
LEAWOOD
BLUE RIVER
Little Bull
Creek
PROJECT
SITE
WOLF CREEK
SPRING
HILL
Massey
Creek
Figure II.8.2 The URAMP study compared the impacts of urbanization on wetland and riparian
areas in two watersheds: Tomahawk Creek, an almost fully developed urban watershed; and Wolf
Creek, a rural watershed just beginning development.
Chapter II.8:
Watershed assessment planning process (Johnson County, Kansas)
479
The study was divided into three phases: phase I, biological assessment; phase II,
watershed assessment; and phase III, workshop and demonstration. Phase I was performed by the Kansas Dept. of Wildlife and Parks, Pratt, KS, with species sampling,
identification, and enumeration; water sampling and testing; habitat and land-use evaluation, and stream flow measurements. Identification of major causes of stream degradation
was also a major component of phase I. Phase II was conducted by George Butler Associates, a local planning and engineering firm located in Lenexa, KS. This phase consisted
of assessing urban development impacts and their causes, reviewing urban development
policies, and providing alternative methods of resource management that may reduce
negative impacts. This phase also included a series of stakeholder meetings to address
the above concerns. Phase III consisted of a regional conference for local decision makers
on recommended alternative wetland and riparian area management strategies. Robbin
Sotir, of Robbin Sotir and Associates, also conducted a soil bioengineering demonstration
project with stakeholders and local planners.
The Kansas Water Office envisioned its role in URAMP as one of a cultural change
agent. The project afforded diverse stakeholders such as engineering and consulting firms,
homebuilders, developers, and local and state planners the opportunity to explore their
attitudes toward riparian and wetland resources. They were able to examine the values
and benefits, perception versus reality, of these resources and educate themselves on
innovative management alternatives. Planning and resource agencies within Johnson
County, and generally throughout the region, as well as developers are initiating many
of these management alternatives.
The Wilderness Valley and Mitigation Bank project

One of the post-project outcomes was a project that designed and constructed a wetlands
and riparian mitigation development plan, which used many of the concepts and recommendations of URAMP (KS Water Office, 1997). In particular, this project implemented
two concepts and one recommendation: using wetlands and riparian areas to convey and
treat urban stormwater runoff and (2) creating a wetland mitigation bank. With the support
of a local school district, county park district, and the city of Overland Park, the project
has integrated a wetlands park and science center into a new, upscale residential community. The project is called Wilderness Valley and Johnson County Wetlands Mitigation
Bank (Figure II.8.3). It is located in the fringe of the floodplain of the Blue River watershed,
Overland Park, KS (Figure II.8.4).
Johnson County, KS, is the states most populated county. It currently has 21 cities
and a population in excess of 460,000. Significant demand exists for new development of
all types, including residential communities, which is a primary cause of urban sprawl.
The demand for land has led to numerous conflicts with environmental concerns such as
the use of floodplains, destruction of wetlands and riparian zones, and impacts on the
quality of surface waters. A comparison of 1935 and 1991 USGS maps for Lenexa, KS,
which is located adjacent to Overland Park, indicates that many of the wooded riparian
areas have been eliminated by development. In addition, over 50% of Kansas wetlands
have been lost due to development (Kansas Wetlands and Riparian Areas Alliance,
undated). Figures II.8.5 and II.8.6 provide examples of the major impact to streams and
riparian areas typically caused by stormwater channel improvements, which most cities
require. URAMP addressed this type of problem.
The original scope of Wilderness Valley was to construct a typical urban-sprawl
residential community with concrete stormwater channels. The ultimate goal was to
maximize the use of the Blue River floodplain by excavating and filling for development
of lots (Figure II.8.7). After a review of the needs URAMP identified, however, the scope
Figure II.8.3 The Wilderness Valley and Johnson County Wetlands Mitigation Bank Project is located just west of the MissouriKansas state line
in Overland Park, KS.
480
Project Location
Chapter II.8:
481
Blue River Watershed Map

BONNER
SPRINGS
CedarMill
Creek
ROELAND
PARK
LAKE
QUIVRA
Todley
Creek
Turkey
Creek
MERRIAM
SHAWNEE
Mill Creek
MISSION
WESTWOOD
WESTWOOD
HILLS
MISSION
WOODS
FAIRWAY
COUNTRYSIDE
MISSION
Brush
HILLS
Creek
PRAIRIE
VILLAGE
CaptainKill
Creek DeSOTO
LENEXA
OVERLAND PARK
Middle
Indian
Creek
Captain
Creek
Lower
Indian
Creek
Cedar Creek
TOMAHAWK CREEK
Kill Creek
OLATHE
Upper
Indian
Creek
GARDNER
Bull Creek
EDGERTON
LEAWOOD
BLUE RIVER
Little Bull
Creek
PROJECT
SITE
WOLF CREEK
SPRING
HILL
Massey
Creek
Figure II.8.4 The project is located in the floodplain of the Blue River, Johnson County, KS.
of the project was expanded to include the creation of a wetland and riparian ecosystem
as a wetlands mitigation bank for 404 wetland permits, which would also be used as:
An amenity for the Wilderness Valley residential development

A regional park for public recreation
A regional science center for public education
A demonstration of using wetland systems for improvement of water quality from
urban runoff
Project formulation
The wetlands mitigation bank concept was formulated to provide a source of revenue
through the sale of credits to fund the other uses. According to URAMP (KS Water Office,
1997), most wetlands in Johnson County are man-induced, or created from the activities
of farming and ranching. Wetlands were typically associated with the construction of
ponds and erosional deposition areas along streams. Therefore, they were of relatively
low value and typically in the way of development. Although low in value, many are
482
Figure II.8.5 A typical gabion basket storm drainage project located in Lenexa, KS.
Figure II.8.6 A typical concrete-lined channel storm drainage project located in Overland Park.
classified as jurisdictional wetlands and require permitting by the U.S. Army Corps of
Engineers (USACE), under Section 404 of the Clean Water Act.
Often, a problem exists with on-site mitigation of these wetlands because of the lack
of long-term maintenance (USACE, 1992). Therefore, USACE, which was a participant in
the URAMP meetings, had expressed an interest in using a wetlands mitigation bank to
mitigate this problem. It was believed that a mitigation bank would provide a higher
value for replacement wetlands in the region. The Johnson County Wetlands Mitigation
Bank (Bank) was specifically designed to provide (1) an area to replace wetland disturbances and (2) high-quality wetlands and riparian ecosystems. It was designed to sell
credits to other developers who needed a place to mitigate wetland disturbances off-site.
Revised / Project 100 Yr.
CorrectedEffective and
Floodplain Fill
Floodway Limit
Effective 100 Yr.
Floodplain Boundary
Floodplain Boundary
Revised 100 Yr.
100 Yr. Floodplain Boundary
Fully Developed Revised
Existing Grade
Floodway Limit
Effective 100 Yr.

Vert. 1 = 10
SCALE: Horz. 1 = 100
SECTION 211237
870
880
890
900
910
Floodplain Boundary
100 Yr. Floodplain Boundary

Existing, CorrectedEffective
Floodway Limit
Floodway Limit
Rosewood
Figure II.8.7 Typical cross-section of the Blue River floodplain that shows the wetlands mitigation bank portion of the
development provided to elevate the adjoining residential final grades above the elevation of the fully developed 100-year
flood event.
870
880
890
900
Proposed Residential Grade
Fully Developed 100 Yr.

and Revised / Project
Chapter II.8:
910
483
484
PROJECT SITE
BLUE VALLEY
ELEMENTARY SCHOOL
PHASE 2
163rd TERRACE
NALL AVENUE
WILDERNESS
SCIENCE
CENTER
PHASE 1
PROJECT SITE
PRAIRIE
RESTORATION
AREA
Figure II.8.8 Project development includes two phases: phase 1 construction to be completed by
2002 with approximately 53 acres of wetlands; and phase 2 construction to be completed by 2005
with approximately 10 acres of wetlands.
Project design
The Bank is approximately 63 acres consisting of two phases Figure II.8.8. Phase 1 is
approximately 53 acres, and phase 2 is 10 acres. Phase 1 is separated into two construction
periods, with approximately 30 acres completed in 2000 and the remaining 23 acres to be
completed in 2001. Phase 2 will be completed in approximately five years following the
completion of phase 1. The residential community will be completed in conjunction with
the construction of each Bank phase.
The proposed wetland and riparian habitat types are shown in Figure II.8.9. Habitat
development includes various wetland types, including wooded, shrub, and herbaceous,
as well as pools of open water (Table II.8.1). Phase 1 also includes a preserved riparian
habitat located along the Blue River. Habitat units are randomly distributed in swales,
causeways, peninsulas, islands, and shallow zones around the edge of open water. Openwater areas will be approximately 8 or 10 ft deep and are expected to support fish and
other aquatic life.
A watershed analysis and water budget were performed to establish optimum hydroperiods. These data are presented in Table II.8.2. One flashboard outlet structure and an
Chapter II.8:
485
Komper Bern
Silo 8
Farm House
(Private Ownership)
LEGEND
Preserved Riparian Habitat
Created Riparian Habitat
Created Aquatic Habitat
Created Herbaceous Wetlands
Created Herbaceous
Shrub Wetlands
Created Wooded Wetlands
SCALE: 1" 400'
Figure II.8.9 Phase 1 includes a diverse mixture of wetland and riparian habitat types including
preserved and created woodland, shrub, and emergent and open-water aquatic areas.
emergency spillway will control water levels. Equalization pipes connect several of the
wetland cells that are separated by earthen causeways.
Wetland plantings include mixtures of native herbaceous and woody species, as well
as recruitment of native seeds and rootstock. Native plant materials will be captured in
topsoil that is salvaged and redistributed during construction of wetland cells. Swales are
designed to function as natural stormwater channels and include rock weirs for erosion
486

Table II.8.1 Wetland and Riparian Habitat Acres
Habitat Types
Riparian (preserved)
Riparian (created)
Aquatic (created)
Herbaceous wetlands (created)
Herbaceous shrub wetlands (created)
Created wooded wetlands (created)
Total acres
Number of Acres
8.3
5.0
4.6
18.0
4.9
12.7
53.5
Note: Phase 1 of the project included a total of 53.5 acres of a

diverse mixture of preserved and created wetland, riparian,
and open-water aquatic habitats.
control. The residential-lot fills located at the fringe of the floodplain adjacent to the
wetlands contain a buried rock berm to protect against slope failure and erosion during
flood events. A series of trails and boardwalks is to be installed for maintenance and to
provide access for the general public and students. The wetland and riparian zones located
adjacent to residential lots will be separated by a wooden rail fence and 10-ft-wide native
grass buffer.
Bank instrument review and approval

A bank instrument was prepared in accordance with USACE guidelines (1995) and submitted in October 1998. A mitigation bank review team (MBRT) was established to review
the instrument and banking proposal. The MBRT included representatives from the U.S.
EPA, U.S. Fish and Wildlife Service, U.S. Natural Resources Conservation Service, KS Dept.
of Health and Environment, and KS Dept. of Wildlife and Parks. In addition, several local
interests, including the Blue Valley School District (BVSD), the Johnson County Park and
Recreation District (JCPRD), and the city of Overland Park, were involved in the review
of the instrument.
The MBRT review required preparation of two revised documents that addressed
concerns and interests of the various agencies. The final bank instrument was approved
in April 2001, with a 401 Water Quality Certification received in February 2001. The final
approval and certification required preparation of a water quality protection and monitoring plan. The wetland development must protect three water quality standards
wildlife, secondary contact recreation, and groundwater protection. The water quality
protection and monitoring plan includes water quality monitoring stations at stormwater
inlets and the principal spillway flashboard structure outlet.
The Bank service area includes all of Johnson County and portions of the Blue River
watershed located outside the county. Approximately 53 wetland acre credits are available
with various thresholds of project completion. Thresholds include permanent easements,
grading as-built plans, documentation of hydrology, initial planting, replacement plantings, and final vegetation success. Vegetation success is measured by permanent transects
that have been established within each wetland habitat.
The final bank instrument also includes intent-to-donate agreements with the BVSD
and the JCPRD. The BVSD will receive the portion of the Bank that adjoins the existing
Wilderness Science Center. The JCPRD will receive the remaining portion of the Bank,
which is the first major land acquisition for the Blue River Streamway, a planned county
regional park. The JCPRD is responsible for installing trails, boardwalks, and observation
areas.
1.14
1.43
2.41
3.91
5.87
5.73
4.12
3.9
4.56
3.49
3.03
1.66
January
February
March
April
May
June
July
August
September
October
November
December
0.27
0.72
1.89
4.35
8.64
12.79
15.45
17.90
20.96
23.05
24.73
25.33
= 12.45 acres
1,784,495
2,925,837
7,677,887
16,193,322
28,170,118
27,309,488
17,450,827
16,130,550
20,076,061
13,713,684
11,089,899
3,933,619
Estimated
Runoff
(gal)
= 888 ft
339,671
251,291
321,878
741,164
1,099,746
1,736,806
2,673,032
3,005,927
2,718,650
2,064,263
1,061,577
528,693
293,393
306,708
564,738
1,342,326
1,624,855
1,911,996
2,143,936
1,854,568
1,453,683
1,063,272
596,015
361,989
Outflows
Estimated
Estimated
Evaporation Transpiration
(gal)
(gal)
Starting elevation
Baseflow (gal/day) =
Wetland area
2,312
2,312
2,312
2,312
2,312
2,312
2,312
2,312
2,312
2,312
2,312
2,312
Estimated
Seepage
(gal)
= 3.5 acres
= 18.8 acres
= 6.74 acres
= 15.56 acres
= 12.45 acres
1,149,119
2,365,525
6,788,959
14,107,520
25,443,206
23,658,373
12,631,546
11,267,744
15,901,415
10,583,837
9,429,994
3,040,625
888.00
888.00
888.00
888.00
888.00
888.00
888.00
888.00
888.00
888.00
888.00
888.00
Results
Water
Balance
Pond
(gal)
Elevation
Wooded area
Herbaceous
Water
Total wetland area
80% of wetland
Note: Data from Hillsdale Lake Reservoir (elevation = 918), water year 1997. A watershed and water budget analysis was performed for each
wetland cell. The analysis of rainfall/runoff, evaporation, transpiration, and seepage values established the water-surface elevation for
the optimum wetland hydro period.
Precipitation
(in.)
Event Date
Actual
Runoff
(in.)
6.74 acres
888 ft
0.0002 in./hr
242 acres
86
1.63
Inflows
=
=
=
=
=
=
Chapter II.8:
Pond Data
Design surface area
D.W.S. elevation
Seepage
Drainage area
Conservation number
Slope of channel
Table II.8.2 Monthly Water Budget Calculations Wetland Cell Elevation 888.00

487
488

Overview of Approval Process
Zoning obstacles
Wetlands
Platting required
Rezoning required
Subdivision design
Single-family lots not allowed in the floodplain
Open drainage channels violate citys stormwater
ordinances
Protests from neighbors
Figure II.8.10 Many city zoning and planning standards had to be addressed by detailed biological
and engineering studies as well as variance requests that were reviewed and approved by the
Overland Park Planning and Engineering staff, Planning and Zoning Committee, and City Council.
The final bank instrument also includes a trust fund that will provide long-term
operation money to be transferred to the BVSD and JCPRD at the time of the property
transfers. Money for the fund is obtained through a percentage of each Bank credit sale.
City and state review and approval

The design of the project included many unusual components that did not comply with
city ordinances related to subdivision design and stormwater management. Most notably,
the ordinances do not address or allow for the construction of wetlands systems and open,
natural stormwater drainages (Figure II.8.10). Numerous revisions were made to the initial
project to accommodate the zoning and planning restrictions. Nevertheless, the following
key approvals were still required for the project to be successful:
The wetland and mitigation bank area were rezoned Agricultural, which eliminated paying a significant excise tax.
A stormwater variance was obtained for construction and maintenance of open
drainage channels.
The citys official floodplain maps were revised in order to allow fill in the fringe
of the floodplain.
The subdivision design included stringent compaction and erosion controls.
The homeowners association restrictions and covenants included measures to
protect the adjacent wetlands from nutrients, trash, etc., from homes.
Sales agreements included cautions about wetland pests such as mosquitoes,
snakes, rodents, and unpleasant odors.
Exemptions were required from the citys weed ordinance.
After extensive engineering and biological studies, including hydrology and hydraulic
modeling, watershed and wetland function and values analysis, geotechnical study, and
performance and construction specifications were submitted in support of the project, the
city planning and engineering staff supported the project and the recommended approval
of the required variances and applications.
Chapter II.8:
489
Numerous state and county approvals were also required, including:
Water rights permit

Floodplain certificate
Stream disturbance permit
Sediment and erosion control plan and NPDES permit
Cultural resource survey and clearance
Threatened and endangered species clearance
Many state and local agencies considered this project an excellent opportunity to demonstrate the use of innovative wetland technologies to control water pollution from urban
runoff. Therefore, many of the approvals and variances granted were considered a first
for KS.
Summary and conclusion

The joint Wilderness Valley and Johnson County Wetlands Mitigation Bank Project provides many benefits. It is being constructed by a private partnership to create a public
trust. In addition to preserving and enhancing a major reach of the Blue River, the wetlands
serve as an excellent amenity for the adjoining housing development as well as an important regional county park and school science center. The county park will provide numerous recreation opportunities including birding, hiking, fishing, and nature study. The
Wilderness Science Center will be expanded to include open-water studies and several
new wetland types. The Bank also serves a critical need for other developers who need a
quality wetland mitigation site.
When completed, the project is expected to provide excellent treatment of urban
stormwater runoff, which will help to establish local guidelines for the use of similar
treatment technologies to meet the new phase II Water Pollution Control Standards. The
project has received much attention over the past two years. It received the 2000 Innovative
Technology Award presented by the Kansas City Chapter of the American Public Works
Association.
Acknowledgments
Project sponsors include Clay Blair Services Corporation, Developer; George Butler Associates, Engineer; Blue Valley School District, Donatee; Johnson County Park and Recreation
District, Donatee; city of Overland Park, Project Reviewer; and Kansas Water Office, Project
Reviewer.
Literature cited
Kansas Water Office and George Butler Associates, Inc., The urban resource assessment and management project (URAMP), Topeka, KS, 1997.
Final Banking Instrument, Johnson County Wetlands Mitigation Bank, U.S. Army Corps of Engineers, Kansas City District, MO, 1999.
Kansas Wetlands and Riparian Areas Alliance, Brochure regarding the background of wetlands in
Kansas, Manhattan, KS, undated.
U.S. Army Corps of Engineers (USACE), Wetland mitigation banking concepts, prepared by Richard
Reppert, Inst. for Water Resour., Alexandria, VA, 1992.
USACE, Federal guidance for the establishment, use, and operation of mitigation banks, Fed.
Register, Nov. 28, 1995.
490
Response
Managing suburban watersheds for multiple objectives
The suburban edge is where development pressures are often the heaviest. As such, it is
also the location where elements of water sensitive planning and design can be most
beneficial in mitigating these pressures. Other chapters is this book highlight the merits
of utilizing constructed wetlands for the treatment and management of stormwater runoff
(Robert France and Philip Craul in Chapter I.7, Diana Balmori in Chapter I.8, Catherine
Berris in Chapter I.9, and Glenn Allen in Chapter I.13). This chapter by Dennis Haag,
Stephen Hurst, and Bryan Bear is significant in that it demonstrates how site-specific
concepts of wetland design are effected through comprehensive watershed planning.
As Haag and his co-authors state, it is imperative to integrate different perspectives
at the very start of water sensitive planning projects. The views and aspirations of such
players as private developers, civil engineers, and state, federal, and local city water
resource planners all need to be recognized, understood, and then balanced with the local
communitys wishes and desires. Watershed management planning, especially that which
pushes the local wisdom base in its recommendations for innovative programs and
policies as was accomplished at this location in Kansas can often be just as much
about sociology as about ecology. This was also the case found in those studies described
by Thomas Benjamin in Chapter II.4, Gail Krantzberg and Judi Barnes in Chapter II.6, David
Blau in Chapter II.7, Kelly Cave in Chapter II.9, and Daniel Williams in Chapter II.11.
Another instructive element in this chapter is the presentation of a series of frameworks outlining straightforward, logical procedures to arrive at the goal of water sensitive
planning. For example, it makes good heuristic sense to first assess the effects of urban
development, then review existing urban development policies, and then provide alternative methods of resource management. Also, taking the phased approach of biological
assessment, followed by watershed assessment, and then followed by public participatory
workshops and discussions is also similar to the strategy that Benjamin adopts in Chapter II.4.
This chapter demonstrates the strength and importance of developing high-visibility
demonstration projects as sentinel education tools for water sensitive planning. Finally,
as Haag and associates imply, such projects will meet with the greatest success if they are
truly motivated by encouraging and enthusiastically supporting the need for multiple
objectives in addition to simply supporting stormwater treatment. In this case, the ancillary objectives were wetland mitigation banking, science education, public recreation,
and a residential housing amenity creation.
chapter II.9
Urban watershed management

(Detroit, Michigan)
Kelly A. Cave
Abstract
The Rouge River National Wet Weather Demonstration Project (Rouge Project) is a watershed-based effort, sponsored by the U.S. Environmental Protection Agency (U.S. EPA), to
manage wet weather pollution to the Rouge River, a tributary to the Detroit River in
southeast Michigan, which is designated as a significant source of pollution to the Great
Lakes system by the International Joint Commission (IJC). The Rouge River watershed is
largely urbanized, spans approximately 438 square miles, and is home to over 1.5 million
people in 48 communities and 3 counties. Sources of pollution to the river include industrial and municipal point sources, combined sewer overflows (CSOs), stormwater runoff,
interflow from abandoned dumps, discharges from illicit connections, discharges from
failed on-site septic systems, and resuspension of contaminated sediment. The Rouge
Project has expanded from a program to build and evaluate alternative approaches to
control CSOs to a comprehensive watershed-based pollution abatement initiative.
The Rouge Project is demonstrating that a watershed-based pollution management
program, which provides flexibility and real delegation of authority to local stakeholder
agencies to decide how to achieve water quality goals, is achieving faster and more costeffective restoration and protection of water resources. In addition, local involvement in
addressing water quality problems is resulting in alternative designs for pollution controls,
which incorporate multipurpose and aesthetic features that facilitate their acceptance by
the general public. The Rouge Project is designing, constructing, and evaluating over 200
full-scale pilot pollution control and watershed restoration projects, including CSO control
basins, stormwater best management practices (BMPs), wetlands, abandoned dump cleanups, habitat protection and restoration, and management of on-site sewage disposal systems. This chapter provides a summary of these pilot projects.
Addressing all of the sources of impairment in the Rouge River, however, necessitated
strong consensus building among the 48 community governments, 3 county governments,
state and federal governments, industries, environmental groups, and private citizens to
show that they had a stake in restoring the Rouge River and that their participation was
vital. This chapter also describes the Rouge Project efforts to build institutional and
regulatory frameworks necessary to accommodate a watershed approach to wet weather
pollution management. Consensus-building strategies, critical to the success of this effort,
are also described here and were used to engage numerous stakeholders, gain their
1-56670-562-2/02/$0.00+$1.50
491
492
Figure II.9.1 Location of Rouge River Watershed in southeast MI. (Courtesy of Rouge River National
Wet Weather Demonstration Project, Detroit.)
support, provide them opportunities to influence decisions, and participate in actions to

restore and sustain the Rouge River as a valuable community asset.
Introduction
In order to achieve water quality standards and associated designated uses within surface
waters, it has become abundantly clear that pollution management must be addressed
through a watershed approach. The watershed approach is a holistic approach that considers the impacts from all sources of pollution and use impairment in a receiving water. The
historic implementation of water quality management programs in the United States at
the federal and state levels has been to focus on point sources, which are the most obvious
sources of pollution to water bodies. This program has worked well to control pollution
from (large) point sources but has also left a patchwork of regulated and unregulated
discharges of stormwater and nonpoint source pollution to surface waters. This patchwork
is especially true in most urbanized areas where multiple local jurisdictions are located
in the same watershed. More subtle sources of pollution, such as stormwater, are now
emerging as the next priority for attention. The challenge is to develop innovative solutions
to achieve water quality objectives that may be (1) more cost-effective, (2) implemented
in a more timely fashion, and (3) better able to meet local needs. It has also become clear
that water resources management must have the support of the general public in order
to be effective and to become self-sustaining. A locally driven watershed approach to
pollution management as a means to achieve management goals is an exciting concept
that many have discussed but for which there is limited practical experience. This is
particularly true in urban situations where there are multiple sources of impairment to a
water body and stiff competition for limited local resources to address the pollution
sources. The Rouge River National Wet Weather Demonstration Project (Rouge Project)
in southeastern MI (see Figure II.9.1) has provided a unique opportunity for a watershedwide approach to restoring and protecting an urban river system by using a cooperative,
locally based approach to pollution control.
The Rouge Project has learned a great deal on what it takes to restore an urban
waterway to its beneficial uses. The purpose of this document is to present some of the
lessons learned to date. Each of these lessons learned was developed based on the
extensive experience of the Rouge Project. By availing themselves of the information
available from the Rouge Project, others will be able to save considerable time and money
in the implementation of their own pollution control programs. Comprehensive information
Chapter II.9:
Urban watershed management (Detroit, Michigan)
493
Main 1-2
Upper
Middle 1
Main 3-4
Middle 3
Lower 1
Lower 2
Figure II.9.2 Subwatersheds and communities of the Rouge River Watershed. (Courtesy of Rouge
River National Wet Weather Demonstration Project, Detroit.)
on the Rouge Project, including technical reports and other materials, is available from
the Web site http:\\www.rougeriver.com.
Background on the Rouge Project

The Rouge Project was initiated in 1992 by the Wayne County, MI, Dept. of the Environment. The project is a U.S. EPA grant-funded comprehensive program to manage wet
weather pollution to restore the water quality of the Rouge River, a tributary of the Detroit
River in southeast MI. The Rouge River has been designated as a significant source of
pollution to the Great Lakes system. The Rouge River watershed is largely urbanized,
spans approximately 438 square miles, and is home to over 1.5 million people in 48
communities and 3 counties. The communities and subwatersheds comprising the Rouge
River watershed are shown in Figure II.9.2.
The eastern side of the watershed consists of much of the older industrial areas in
southeast MI. The western side of the watershed consists of newer suburban development
and areas under heavy development pressure. The diverse nature of the Rouge River watershed can be seen in Figure II.9.3. The Rouge River consists of 4 main branches totaling
approximately 130 miles. All sanitary and combined sewers in the watershed are connected
to the Detroit Wastewater Treatment Plant, which discharges outside the watershed into the
Detroit River. Combined sewers serve 20% of the watershed. Separate sanitary and storm
sewers serve most of the remaining areas of the watershed, with the exception of isolated
pockets and rural areas in the headwaters that still have on-site septic systems. Historically,
the major sources of pollution to the river were industrial and municipal point sources, wet
weather sanitary sewer bypasses, and CSOs. The point sources have been successfully
controlled by an aggressive National Pollutant Discharge Elimination System (NPDES)
permitting process administered by the state regulatory agency (MI Dept. of Environmental Quality, MDEQ); however, the river still failed to meet water quality standards
due to a wide range of sources such as CSOs, sanitary sewer overflows (SSOs), stormwater
494
Figure II.9.3 Diverse characteristics of Rouge River Watershed. (Courtesy of Rouge River National
Wet Weather Demonstration Project, Detroit.)
runoff, illicit connections, failing septic systems, interflow from abandoned dumps, and
resuspension of contaminated river bottom sediment.
The Rouge River had been identified as one of 43 tributary areas of concern (AOC)
in the Great Lakes system by the IJC Water Quality Board in 1985. A remedial action plan
(RAP) documenting water pollution problems and proposing corrective actions was prepared for the Rouge River in 1989 and updated in 1994. The Rouge River RAP cited the
remediation of the CSO discharges in the combined sewer area of the lower watershed as
a priority but also recognized the importance of controlling sources of pollution emanating
from nonpoint and stormwater discharges in the upper watershed served by separate
storm and sanitary sewers, and on-site septic systems (Bean et al., 1994). The Rouge Project
was born out of a desire and critical need to manage the multiple sources of pollution in
this large, urban watershed in a prioritized, comprehensive manner. The U.S. federal
government began sponsoring this effort in 1992 to develop and demonstrate technical,
institutional, and regulatory options and processes to protect and restore a large, multijurisdictional, urban watershed.
Early on, the Rouge Project focused on the control of CSOs in the older urban core
portion of the downstream areas of the Rouge watershed. The Rouge Project initiated the
watershed-wide management approach in southeast MI by facilitating CSO control and
permitting based on common requirements throughout the watershed. Rouge communities served by combined sewers have entered into permits with the MDEQ and the EPA
requiring a base level of abatement construction throughout the watershed followed by
assessment of water quality impacts and future construction phases to meet public health
and water quality standards. This approach is a significant departure from previous point
source permits, which were typically issued on an individual basis and often not coordinated with other pollution control activities affecting the same water body.
CSO control is being implemented in phases, with phase 1 recently completed. Eight
communities constructed 10 retention treatment basins to serve 35.1 square miles of combined sewer systems. Each of these basins is sized for different design storms and many
also employ innovative technology. In addition, many of the CSO control basins are multipurpose facilities, such as the basin shown in Figure II.9.4, which provides a community park
Chapter II.9:
495
Figure II.9.4 Aerial view of Inkster combined sewer overflow basin in Rouge River Watershed,
Inkster, MI. (Courtesy of Rouge River National Wet Weather Demonstration Project, Detroit.)
and playground on the top of the basin. One retention/treatment tunnel is under construction in one community containing 3.2 square miles of combined sewers. Several
communities separated their sewers; the drainage area of these projects totaled 3.4 square
miles. A two-year evaluation study of the CSO control program was recently completed
(Hufnagel et al., 2000; Kluitenberg et al., 2000). The design, operation, and cost information
gained from the evaluation of phase 1 control facilities, coupled with efforts to control
stormwater and other pollution sources in the watershed, will provide the basis for the
phase 2 CSO control efforts.
Concurrent with the initial development of the CSO control strategy, the Rouge Project
initiated a comprehensive data collection effort, which included gathering information
about watershed features and characterizing existing water quality and ecosystem health.
The first water quality sampling under the project began in 1993, and by the end of 1994
a supporting geographic information system and watershed modeling effort were in full
operation with baseline, automated water quality monitoring sites located throughout the
watershed (Mullett et al., 1994). Water quality and ecosystem health monitoring has
involved an extensive effort in the collection, management, and analysis of data on rainfall,
stream flow, in-stream water quality, CSO and stormwater quality, biological communities
and habitat, in-stream bottom sediment, air deposition, and aesthetic conditions. In addition, the monitoring program includes measurement of the performance of various structural pollution controls, wetlands, and pollution prevention activities. The initial
sampling later confirmed during subsequent sampling in 1995 through the present
documented significant pollution problems in the Rouge River watershed upstream from
the CSO discharges. State water quality limits for bacteria and dissolved oxygen were
regularly exceeded even in dry weather periods in the upper watershed, and highly
variable flows caused flooding, exacerbated bank erosion, and increased sedimentation
that affected the lower river. This information, shown by the example in Figure II.9.5,
confirmed the suspicions of many that the discharges from separated storm systems in
heavily urbanized areas can be a significant sources of pollution, including coliform
bacteria. The Rouge watershed assessment tools have proved to be critical in garnering
public support for the river restoration efforts and have provided the general public, local
496
E. coli/100 ml
100,000
10,000
Uncontrolled CSO (Phase 2)
Storm Drain with Illicit Discharges
Typical Storm Drain

1,000
1,000 E. Coli
Phase 1 CSO Basin Effluent

130 E. Coli Full Body Contact Standard
100
Figure II.9.5 Comparison of sources of E. coli bacteria. (Courtesy of Rouge River National Wet
Weather Demonstration Project, Detroit.)
decision makers, and regulatory agencies with information to prioritize and tailor pollution control solutions to specific river reaches while coordinating efforts throughout the
watershed.
Based on what was learned, the focus of the Rouge Project became more holistic to
consider the impacts from all sources of pollution and use impairments in a receiving water.
The Rouge Project began to identify the most efficient and cost-effective controls of wet
weather pollution, while assuring maximum use of the resource. For example, over 60 pilot
innovative stormwater control technologies are also being evaluated under the Rouge Project
by 25 different communities and agencies. Categories of pilot stormwater management
projects currently underway include wetlands creation and restoration, structural practices
such as grassed swales and detention ponds, erosion controls, stream bank stabilization, and
habitat restoration, to name a few. Figure II.9.6 shows a linear sand filter, one of the pilot
stormwater management projects constructed and evaluated under the Rouge Project.
The Rouge Project has learned that illicit connections and failing septic tanks are major
sources of pollution problems in the Detroit urban area (Johnson and Tuomari, 1998).
Innovative ways to deal with these sources of pollution have been initiated.
A suite of computer models has been developed by the Rouge Project to simulate the
water quality and quantity response of the Rouge River during wet weather events for
existing and future conditions under various CSO and stormwater runoff management
alternatives. This effort has led to a very useful public communication tool on water quality
indices tied to actions needed to restore the Rouge River. A comprehensive geographic
information system (GIS) and relational databases were designed and implemented to
manage the wealth of data collected under the project. In addition, a special data exploration tool, DataView, was developed to support routine analyses of large time-series data
sets (Rood et al., 1995). DataView is user-friendly and readily transferable to other locations. Related to DataView is the Rouge Information Manager, also a user-friendly, readily
transferable tool (an electronic file cabinet) for accessing multimedia information about
the Rouge River restoration effort.
These tools have been vital in the success of the strong consensus-building activities
necessary to show the 48 community governments, 3 county governments, state and
federal governments, industries, environmental and community groups, and private citizens that they had a stake in restoring the Rouge River and that their participation was
vital to the success of the comprehensive watershed management program. These consensus-building strategies were used to engage numerous stakeholders, gain their support,
Chapter II.9:
497
Figure II.9.6 Linear sand filter retrofit to urban area for stormwater management, Wayne, MI. (Courtesy of Rouge River National Wet Weather Demonstration Project, Detroit.)
provide them opportunities to influence decisions, and participate in actions to restore

and sustain the Rouge River as a valuable community asset. The Rouge Projects public
education/public information program has been demonstrated to be a very effective
component of the consensus-building process (Powell and Bails, 2000). Although many
elements are included in the public education program, as illustrated in Figure II.9.7, one
very important element is the involvement of youth in this effort. As one example, in the
annual Rouge Water Festival held each May, 52 fifth-grade classes from over 27 schools
(nearly 1,500 students) participate in this hands-on festival where they learn about the
importance of water in all aspects of their lives. Over 100 schools are currently involved
in education efforts of the project.
The Rouge Project has spent considerable effort to build institutional and regulatory
frameworks necessary to accommodate a watershed approach to wet weather pollution
management. Part of this framework is a watershed-based general permit for municipal
stormwater discharges issued under the National Pollutant Discharge Elimination System
(NPDES) program. This stormwater permit program was developed jointly by the Rouge
communities and the MDEQ and is based on the concept of cooperative, locally based
watershed management (Cave and Bails, 1998). Communities and agencies in over 95%
of the watershed have applied for coverage under this innovative, watershed-based permit
program. The MDEQ permit requires permittees to participate in watershed management
planning for a self-determined subwatershed unit. The subwatershed management plans
form the basis for implementing watershed goals and objectives that will result in
498
Figure II.9.7 Public education and involvement materials developed by the Rouge River National
Wet Weather Demonstration Project. (Courtesy of Rouge River National Wet Weather Demonstration
Project, Detroit.)
improved water quality and pollution control. The Rouge communities will also use these
watershed management plans to achieve other program objectives, such as those under
the federal TMDL program and the state Clean Michigan Initiative (Cave et al., 2000).
The local communities, agencies, industries, and citizens have been working together
in seven subwatershed advisory groups to develop and implement management strategies
for various segments of the river. The Rouge Project has recently been the catalyst for the
Rouge Gateway Partnership, a collaborative effort among county government, corporations, local communities, and academic and cultural institutions, which is guiding redevelopment to restore the 7-mile section of the lower Rouge River. This section of the river
includes a 4-mile concrete channel and a 3-mile section of navigable dredged waterway
downstream of the channelized section. As shown in Figures II.9.8 and II.9.9, the Rouge
Gateway Project will create a 7-mile greenway link for this area, providing the public
access and linking the park system along Hines Drive to the Detroit River waterfront
(Rouge Project, 2000). This effort will demonstrate to other communities how to reclaim
waterways that have been essentially removed from public access and use.
Lessons learned to date

As stated earlier, the project participants have learned a great deal on what it takes to
restore an urban waterway to its beneficial uses. The purpose of this chapter is to present
some of the lessons learned so far. The approach used in determining the lessons learned
was built on the idea of asking, What do we know now that would have saved us time
and money had we known at the start of the project? Each of these 13 Lessons Learned
Chapter II.9:
Figure II.9.8 Rouge Gateway Master Plan, a comprehensive destination and interpretive program plan. (Courtesy of Rouge River National Wet Weather
Demonstration Project, Detroit.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)

499
500
Figure II.9.9 Conceptual greenway along lower Rouge River, part of Rouge Gateway Master Plan.
(Courtesy of Rouge River National Wet Weather Demonstration Project, Detroit.) (Color version
was developed from the vast experience of the Rouge Project. The background for each
of these lessons is explained, citing specific experiences related to the lesson learned.
The watershed approach

Lesson learned
We have known for years that the only way to effectively achieve water quality and ecosystem
protection in an urban river system is to look at all problems and their solutions from a
holistic watershed perspective. The Rouge River Project experience gives us tangible
measures of the benefits of this approach hundreds of millions of dollars being saved.
The Rouge River Project has undertaken a number of actions to demonstrate how to
restore an urban river. Those actions include, but are not limited to CSO controls, stormwater management (including flow reductions), stream restoration, and wetland creation
and enhancement. The largest traditional point sources of pollution, CSOs, have been
controlled or are programmed for control.
When the Rouge Project began, its main focus was on the control of CSOs because
the perception was if CSOs the largest point sources of pollution were controlled,
most of the river would meet water quality standards. The monitoring and modeling
program quickly demonstrated that even if all of the CSOs were totally eliminated from
discharging to the Rouge River, the designated uses of the river would still not be met
due to other sources of pollution such as stormwater runoff and to lack of environmental
habitat. In order to achieve water quality standards and associated designated uses within
the Rouge River, it has become very clear that pollution management must be addressed
Chapter II.9:
501
in a holistic fashion that considers the interrelationship between the impacts from all
sources of pollution and use impairments in a receiving water. Water quality protection
at the watershed level is neither a single capital project, nor a series of projects. In essence,
it is a way of life, and the problems must be approached that way. The EPAs Watershed
Approach Framework also discusses this philosophy (U.S. EPA, 1996).
The project has enough preliminary data to make a rough cost comparison between
utilizing a watershed approach to achieve desired water quality objectives versus using
the historical approach of addressing the causes of water quality degradation individually.
These preliminary data indicate that cost savings for the Rouge River watershed citizens
could easily approach several hundred million dollars.
Lesson learned
In a large watershed, it is most effective to restore or protect water quality and ecosystem
health by looking at subwatersheds within the overall watershed. In the Rouge River
Project, the focus on 20- to 50-square mile subwatersheds has allowed the project to move
from a purely regulatory-driven CSO program to voluntary community-based efforts for
stormwater management.
Environmental protection relies on a mix of federal and state resources as well as
increasing levels of local resources. Use of local resources requires a search for and identification of common environmental interests. Most people have a strong environmental
interest in their community; fewer have a strong interest in an environment with which
they cannot identify. Therefore, subwatersheds give a means for focusing the local
resources to address local problems due to the interest people have in their immediate
surroundings. The EPAs Top Ten Watershed Lessons Learned document also discusses the
need for identifiable goals to focus local resources (U.S. EPA, 1997).
Focusing on subwatersheds has many advantages. First, smaller areas are more manageable in terms of addressing water quality problems. Second, people identify more with
a subwatershed than with a larger watershed. Local ownership of pollution problems and
their solution is critical. In his article, Crafting better urban watershed protection plans,
Schueler (1996) also suggests that subwatersheds having a drainage area of 518 to 3,885
hectares (2 to 15 square miles) in size provide the best scale from both a technical and a
political basis on which to base management plans. Third, it is easier to analyze the various
sources of water quality problems in the subwatershed and decide how to get a handle
on the priority of dealing with those problems. It is critical to establish a hierarchy of
pollution sources in subwatershed (point sources and nonpoint sources) based on the
adverse water quality impacts of those sources. It is very important to keep reinforcing,
at a subwatershed level, the concept of not randomly addressing pollution sources but to
prioritize the control of those sources to get desired environmental protection. It may take
a long time to correct some of these pollution sources, so it is important to prioritize the
control programs. Fourth, it is critical to assess the cumulative watershed impacts to
quantitatively assess the physical and biological processes and then fashion the subwatershed solutions. Fifth, before river restoration can be initiated, it is critical to understand
the cause of stream disturbance and disequilibrium conditions. Without this understanding, the restoration often treats the symptoms instead of effecting a cure. Sixth, it is easier
to manage a process that has a small set of stakeholders and competing interests. Seventh,
it is easier to convince people that water quality improvements will require many little
and possibly inexpensive actions instead of massive capital programs exclusively. Eighth,
and finally, the tools needed to solve subwatershed water quality problems must be geared
to that subwatershed. The management plan developed must be tailored to address
subwatershed-specific problems.
502
Each of these advantages has associated cost implications. It is very important to use
innovation in the watershed approach. By a consensus-based approach, focusing the local
resources to address the local water quality problems will solve them faster and more
inexpensively.
The control of combined sewer overflows (CSOs)

Lesson learned
The control program for CSOs must be flexible and tailored to the site-specific situation
in order to achieve the most timely and cost-effective solution(s).
Michigan has had design standard requirements for the control of CSOs for a number
of years. On the basis of those design standards, the MDEQ issued NPDES permits to the
appropriate communities. Certain of the issued permits were challenged. There followed
a period of negotiation between the permittees and MDEQ. The permittees, supported by
the Rouge River Project, entered into negotiations with the state, which resulted in alternative design standards for certain CSO control facilities. These alternative design standards reflect site-specific considerations and will result in the same level of water quality
protection. The modified design standards resulted in a savings of over $300 million to
the CSO communities in the initial phases of the project.
CSOs are brief, intermittent pollution sources with long-term water quality consequences. The Rouge River Project has demonstrated that after sewer systems are optimized, CSO outfalls generally discharge for less than 100 hr/year. Most other communities
in the United States will have similar CSO discharge duration. Cost-effective CSO control
for the Rouge River has reduced the duration of discharge to 20 to 40 hr/year. This
discharge has received first-flush capture, and the more dilute flow that follows the first
flush is screened and disinfected, and has some degree of solids and biochemical oxygen
demand (BOD) removal. These same CSO controls have provided an extra degree of
insurance for water quality protection by virtually eliminating the chance for any dryweather discharge from the combined sewer system.
One can look at the cost of CSO control in different ways. Using the Rouge River
Project phase 1 experience, the capital cost is $15,000/acre, while the annualized cost
amounts to $10,000/day of water quality protection. Either way, these represent extremely
high fixed costs to deal with an intermittent source. The high fixed cost requires flexibility
in the development of CSO controls because of the wide array of local conditions and
priorities within a community. As stated previously, with design standards that reflect
site-specific considerations, the potential for cost savings is enormous. Therefore, by developing a site-specific CSO control program that is integrated into an overall watershed
management plan that prioritizes the pollution problems to be addressed, considerable
cost savings can be realized.
Lesson learned
For large and costly CSO control programs, implementing the program in phases can save
time and money and foster critical support from local communities.
The installation of CSO controls is very expensive. It is therefore critical that decisions
on those controls be based on the best available information. Phasing the installation of
controls in complex CSO situations allows for reasonable steady progress to achieve water
quality goals in the shortest possible time in a cost-effective manner. Phasing allows time
to understand and develop control programs on related sources of pollution that are
adversely affecting water quality; to develop and implement sewer system optimization
efforts; to determine what is necessary for public health protection; and to determine what
Chapter II.9:
503
is necessary for the achievement of water quality standards, considering all sources of
impairment to the river system, in the most cost- and time-effective manner by allowing
time to develop better baseline information for designing subsequent phases. The phased
approach has some CSO controls installed on part of the overall system, a period of
evaluating the effectiveness of the control technology used to meet the design standards,
including evaluating the impact on the resource, and then making the decisions on the
design standards and treatment technology to be used on the rest of the CSO system.
A fundamental premise of the Rouge River Project was to demonstrate various design
standards and control technologies for CSO controls. This demonstration would occur in
two phases. Phase 1 specified 17 communities to install CSO controls. Under phase 1, 6
communities have separated their sewers and 8 communities have constructed 10 retention
treatment basins utilizing various treatment technologies and design standards. A twoyear evaluation study of the completed CSO control basins was recently completed to
assess compliance with the design standard used for each basin and to assess the positive
water quality impacts on the receiving stream of the controlled CSO discharge. The results
of the evaluation study, coupled with efforts to control stormwater and other pollution
sources in the watershed, will provide the basis for the phase 2 CSO control program on
the remaining CSO sources in the watershed. Phase 2 will require all of the CSO communities to install CSO controls necessary to protect public health by a certain date. Phase 2
controls will be based in part upon the information learned in phase 1. Phasing of the
program also creates opportunities for communities to work together to develop joint
projects that are mutually beneficial and more cost-effective in correcting pollution problems.
Preliminary results from the CSO basin evaluation study completed to date are providing extremely useful information. The information gained from the evaluation of design
storms and control technologies will be extremely useful nationwide in determining costeffective CSO controls to meet water quality standards. Information on the size of the CSO
basins and their operation has the potential for savings hundreds of millions of dollars in
the Rouge watershed. The nationwide savings will even be more substantial.
As stated earlier, it is critical to have strong local support for CSO control programs
if they are to succeed. The best way to generate and sustain that support is to clearly
demonstrate the benefits of the CSO control program. Those benefits are already becoming
evident in the Rouge watershed. For example, over 100 miles of river are now CSO-free.
This translates into the water quality downstream of controlled CSOs being as good as
the water quality upstream. The result is that the natural beauty of the river, along with
its ecological health, is being restored. In addition, the economic benefits are also becoming
more evident because of the increased recreational usage of the river.
The control of stormwater

Lesson learned
In urban areas, stormwater discharges have just as large adverse water quality impacts
as CSO discharges, but the control of stormwater sources is more politically and technically
complex. The Rouge River Project recognizes these differences and is demonstrating the
necessity of using a different approach to managing stormwater.
Stormwater runoff occurs during and after each rainfall event, whereas CSOs may
occur only during the heavier storms. As has been discussed extensively in the literature
(e.g., CWP, 2000), the sources of stormwater in urban areas are many and varied, as are
the resulting adverse water quality impacts. The Rouge Project has documented that water
quality standards are regularly exceeded because of discharges from stormwater in the
504
watershed. Unlike CSOs, where the ownership of the outfall pipe is known, often the
ownership of the stormwater source is not clear. The watershed approach has proven
to be the most effective way to address stormwater issues. The watershed approach is
locally driven, encourages voluntary participation by communities and other public agencies, requires the development of a comprehensive plan to address the problems, and
defines actions and iterative steps leading to comprehensive, watershed-wide stormwater
management. In the Rouge watershed, the local communities, working through the Rouge
Project, assisted the MDEQ in the development of a watershed-based general stormwater
NPDES permit (Cave and Bails, 1998). This regulatory framework encourages communities to cooperate in a watershed approach to address pollution problems. As discussed in
the EPAs Top Ten Watershed Lessons Learned, Watershed work is about partnerships . . .
because partnerships equal power (U.S. EPA, 1997).
Overall stormwater management includes structural, vegetative, or management practices to treat, prevent, or reduce stormwater runoff. Therefore, the solutions to stormwater
problems are difficult and time-consuming to develop and implement.
The cost implications of this lesson are many. The associated potential for cost savings
is therefore also evident. For example, use of the watershed approach to prioritize the
water quality problems and then their solution has the potential to save a great deal of
money. As will be discussed in the next lesson, tying stormwater control to other public
works projects is a sure winner and money saver.
Lesson learned
Integrating stormwater control projects into ongoing public works plans and actions of
local governments results in more timely management of stormwater.
Local governments often have a number of projects or actions planned or underway
in the community such as street sweeping, drainage control programs, and related efforts.
By showing the local officials the benefits of integrating the stormwater control program
into other planned projects or actions, implementation of the stormwater program will
proceed more quickly. This fosters local control over the outcome of the overall project or
action, which also is a key to overall acceptance and then success.
This concept of integrating efforts is closely tied to the lesson learned on combining
stormwater work with the phasing of work. For example, the Rouge River Project has
demonstrated this by creating and then seeing the success of the three-county Roads and
Drainage Roundtable, which has resulted in the road commissioners integrating into their
day-to-day thinking the needs of watershed protection as impacted by road and drainage
work. In summary, local units of government can substantially improve in-stream water
quality by reviewing and adjusting how they perform a number of their daily public
works activities.
This lesson learned has obvious implications for substantial cost savings. Communities need to think of and create opportunities to marry public works and stormwater work,
which results in significant cost savings and more timely environmental protection. Such
a marriage involves other agencies whose prime mission is not water quality protection.
This combining helps everyone involved.
General public support

Lesson learned
Achieving pollution abatement in a more timely, cost-effective fashion must have general
public support, which can be assisted by combining the watershed approach with the
concept of phasing needed pollution controls.
Chapter II.9:
505
The public needs time to understand the complexities of restoring a degraded river
and to then respond with the needed support, including the commitment will and financial
support. The Rouge River Project has learned that the general public may not fully
understand or appreciate a goal of meeting water quality standards. What they do
understand and appreciate is whether a waterway is fishable and swimmable. They accept
that, if all of the sewage is not removed from the water, it is not swimmable; if toxics
preclude fish consumption, the water is not fishable; and if habitat is destroyed, there will
be no fish, and therefore, the water is not fishable. In summary, if the conditions of the
river discourage fishing, swimming, and other recreation, attention must be directed at
correcting the problems.
Generally, local commitments to address pollution problems in a watershed will come
in small increments with the demand to demonstrate the value of those increments if
support is to be sustained. Therefore, a phased program and the pursuit of a multitude
of pollution sources/problems are essential. The Rouge River Project has taken this lesson
to heart and has undertaken a number of locally driven watershed-based projects in order
to demonstrate how to restore an urban river. Those projects include CSO controls, stormwater management, stream restoration, addressing failed septic tank problems, correcting
illicit connections, wetlands creation and enhancement, and stormwater flow reductions.
The project also has implemented a phased approach to the installation of controls.
The Rouge Project has demonstrated that it is cost-effective to spend adequate funds
on developing base public support. Without that support, it will not be possible to achieve
the needed water pollution control program to restore beneficial uses. This public support
also will save money in the long run.
Public education and involvement

Lesson learned
Broad-based public education and involvement programs are critical to the overall success
of watershed projects, particularly in urban areas.
One of the major goals of the Rouge River Project is public involvement and education.
Each person who lives in the watershed needs to be educated as to how his or her
individual daily actions affect the conditions of the river. Public support depends on public
awareness and education. Education and involvement drive action. Therefore, public
education efforts should be viewed as a campaign with intensity and high stakes. The
approach employed should leverage the tremendous environmental education and media
movement already in place.
The Rouge River Project solicits community input through a number of directed
activities. Some successful programs are (1) getting education programs under way in the
area schools, (2) getting people out to the stream to look at and experience it, (3) showing
people some obvious things that have been done to get successes in improved water
quality, and then widely publicizing the results so people see that things are being implemented instead of just planning and talking taking place, (4) making people aware of the
river, which renews their interest in it, and (5) placing signs at each river crossing to
increase awareness. A technical project such as this always faces the challenge of conveying
data in a way so that the layperson understands the issues. The project has stressed the
focus of restoring uses versus raising water quality. People understand restoring uses to
the river versus the abstract concept of improving water quality. The project developed a
graphical presentation on water quality indices, which has fostered a good focused debate
on some key issues of what is the end objective to be achieved (Smith et al., 1997). This
communication tool became the heart of the development of watershed management plans
for subwatersheds.
506
Another important lesson has been on the need to measure, communicate, and account
for progress. The EPAs Top Ten Watershed Lessons Learned states, Having good data
systems in place to measure and communicate progress is a critical part of watershed
work (U.S. EPA, 1997). They not only keep watershed issues on peoples radar screens
but assist in sharing successes and facing new challenges to the watershed.
Information systems
Lesson learned
A data management and information system that can effectively communicate to the broad
public is critical to achieving success in watershed/water quality restoration projects.
In order to make informed decisions on water quality improvements in a watershed
system, it is necessary to have access to, and be able to process, large amounts of data. It
is important to consider very early in the process how the data will be compiled and
analyzed. In other words, consider what type of data management and information system
is needed to accomplish the desired objective of analysis and communication of results.
What the data system demonstrates needs to come through as a clear message. The
involvement of local officials in the process is critical to the long-term success of the
information system. Most local governments have, or soon will have, geographic information systems (GIS) to aid them in decision making. Every effort should be made to
bring together the local units of government in the watershed to establish ways to share
information among data systems and to have commonality in information spreading into
the various data/information systems. This cooperation fosters watershed-based decisions
versus individual community-based decisions.
The Rouge River Project has developed two effective data management and information systems that are easily transferable to others for use. DataView is a user-friendly data
exploration tool developed to support routine analysis of large time-series data sets. The
system allows for tabular data viewing, data plotting, generating summary statistics,
spatial display, and data export. The Rouge Information Manager is a user-friendly tool for
accessing and displaying information. It serves as a portable electronic filing cabinet that
anyone can use to access information on monitoring data, GIS maps, modeling alternatives,
public involvement, wetlands, stream restoration, illicit connections, and other technical data.
It contains maps, technical reports, photos, and videos, which can be displayed.
Information systems can be very costly. By shopping around it should be possible to
adapt an already existing system to meet your needs. This will result in considerable cost
savings. As stated earlier, the Rouge Project information system is easily transferable to
others.
Also, cost savings will result by having a well-informed public that supports the water
quality goals trying to be achieved.
Monitoring versus modeling

Lesson learned
Achieving a balance between monitoring and modeling can save money and time in
implementing a wet weather water quality management program.
When addressing wet weather impacts on water quality in waterways, the natural
tendency is to want to establish an extensive monitoring program to fully document the
cause and effect of the water quality problems. Monitoring programs are time-consuming
and expensive. Related to monitoring is the use of models to define present conditions
and to project future options. The objective should be to achieve a balance between
Chapter II.9:
507
monitoring and modeling. The Rouge River Project has demonstrated several key aspects
related to that balance.
First, think through how you will use any data collected from a monitoring program
to make a subsequent decision. The natural tendency is to collect much too much data on
a large suite of chemicals. After a series of hard questions, the amount of monitoring data
collected can probably be reduced significantly, with the concomitant cost savings. Some
of those hard questions are:
How will I specifically use the data on this specific chemical to reach a decision?
How much data do I need on this specific chemical to reach a decision?
Do I need screening level data or extensive data to reach a decision?
Are models available to help predict the future water quality conditions in this water
body based on changed waste inputs?
If so, what are the minimum data needed to support the model so that the prediction
is within reasonable bounds, i.e., within a reasonable percentage of accuracy?
How can I use the model to the maximum extent possible to save costs and only use
limited monitoring to fill in gaps as needed?
By the same token, a project can be consumed by modeling and the need to develop
the perfect model. Be careful not to continue to gather data to continually feed a model
mindlessly. By asking a series of hard questions on modeling, you can have significant
cost savings. Some of those hard questions are:
Are models currently available to help predict the future water quality conditions in
this water body based on changed waste inputs so that I do not have to develop
a new model from scratch?
If so, what are the minimum data needed to support the model, and what is the
minimum modeling effort needed so that the prediction is within reasonable
bounds, i.e., within a reasonable percentage of accuracy?
How can I use the model to the maximum extent possible to save costs, remembering
the concept of getting answers within reasonable bounds?
Remember to keep asking yourself, What do I want to model? If you do not keep
asking that, your model could grow topsy. You must always remember that monitoring
and modeling are tools to get to the end, but these tools should never be confused as being
the end.
The idea is to use simple models at first, drawing on available water quality sampling
information. Let the modeling show where additional data are needed to define location
and parameters. Then continue a combination of modeling and monitoring to achieve the
desired objectives.
As an example to demonstrate the economics of monitoring and modeling, the amount
of money the Rouge River Project has spent to characterize the resource has been approximately $5,000 to $7,500 per mile/year (calculations based on total monitoring costs over
the life of the project and the number of main stream miles). A long-term monitoring
program is estimated to be approximately $2,500 per mile/year. Compare that to modeling
costs, which can be calculated to be $2,000 to $3,000 per mile/year. The long-term modeling
costs are $750 per mile/year. The ratio of monitoring costs to modeling costs increases
from 2.5:1 for the short-term work to 3.3:1 for the long-term work.
In summary, local monitoring programs are needed. They can be minimized and still
provide a good basis for planning and decision making. Some key steps to remember are:
508

1. Collect and analyze existing data.
2. Compare existing data to findings elsewhere such as the Rouge River Project and
determine what conclusions can be drawn. Use simple models at this point to help
make sense of the data.
3. Plan new data collection and monitoring to fill gaps in information and to further
investigate known pollution problems in the watershed.
An overarching lesson associated with monitoring and modeling is the need to avoid
paralysis by analysis. Monitoring and modeling can do much to prioritize restoration
activities, but certain common-sense activities should be implemented given only minimal
data. Existing national data are often sufficient to flag those items. Some examples include
septic field inspection and correction of problems, illicit connection removal, and downspout disconnection.
As stated in the previous example, the potential for cost savings through a mix of
monitoring and modeling is enormous. In addition, a sound mix of modeling and monitoring will save time, which also translates into money saved.
Institutional and financial arrangements

Lesson learned
The toughest problems to be addressed and solved in wet weather and watershed protection programs are developing and implementing the institutional and financial arrangements needed to sustain the program. The technical issues are easy by comparison. Early
and continued efforts should be directed toward developing workable institutional and
financial arrangements.
The sources of stormwater runoff to surface waterways are many and varied, with
ownership often unclear. CSO discharges are identifiable but can be costly to control. Other
sources of waste that cause water quality degradation such as dry-weather pollutant sources
(sediments, septic tank leakage, illicit connections) are a considerable problem in urban areas.
Other considerations such as quantity of flow in an urban stream and the flashiness of that
flow can adversely impact water quality and the program to restore beneficial uses. Habitat
issues are an important consideration related to beneficial uses of the waterways. Urban
growth and related land-use patterns can have major impacts on water quality restoration
programs. All these challenges are related to the institutional and financial arrangements
needed to accomplish the desired end objective of a restored waterway.
The Rouge River Project has learned some important lessons in this area. For example,
it must be demonstrated to the residents of a watershed/urban area that all are paying
their fair shares of the waterway restoration costs. It is generally accepted that every city
and county must do a set of minimum activities to address water quality problems and
that certain cities and counties will have to do more because of their specific set of water
quality problems/issues. Residents need to be shown that the benefits of water quality
improvement are worth the costs. Interjurisdictional agreements get local communities
committed to the projects objectives and garners the day-to-day support needed to accomplish the necessary objectives.
In order for a watershed project to be successful, an institution to oversee the
progress is not necessarily needed. What is critical is the need for effective institutional
arrangements. These can be as simple as utilizing forms of interjurisdictional cooperation
or remaking or combining existing institutional arrangements. These solutions will mirror
the complexity of the problem to be solved.
A key element of the institutional and financial concerns deals with the federal and state
regulatory agencies. Obtaining regulatory agency consensus on all aspects of a watershed
Chapter II.9:
509
project is critical to its overall success. Regulatory inertia is very strong toward continuing
down the historical paths that have been followed in the past. Therefore, trying new
methods of institutional and financial arrangements can be a tough sell. It is critical that
the federal and state regulatory agencies have staff and managers fully involved in all
aspects of a watershed project in order to witness the advantages of trying new approaches
to address problems.
Substantial cost implications are associated with institutional and financial arrangements. The project has demonstrated that it is well worth the time spent upfront looking
at these topics. Waiting until the end of a project and then using a piecemeal approach
will be a waste of time and money. It takes time to solve these issues. Many meetings are
necessary to address this topic, but they are worth the cost in the long run.
Flow as a water quality issue

Lesson learned
In urban areas, the quantity of stream flow is very important in assessing water quality and
ecosystem health and the subsequent correction of problems. By solving problems associated
with high flow, many water quality and ecosystem health problems will also be solved.
In urban areas, streams often experience high flows during wet weather events. These
high flows cause or aggravate ecosystem health problems for several reasons, including
scouring of banks and the resulting turbidity problems; subsequent sediment deposition
and its related adverse impacts on habitat and beneficial uses; adverse impacts on wetland
habitat; the physical destruction of the river banks, which adversely impacts land values;
and the wide variation in flows, which adversely impacts swimming, wading, boating,
and fishing. Wetlands in urban areas are very beneficial because they provide good storage
for water, which tempers flow variability, improves water quality, and provides good
habitat for aquatic life as well as other wildlife. Yet, highly variable flows adversely impact
these needed wetlands.
Often, the damage caused by flooding and stream bank erosion on private property,
the loss of recreational opportunities (i.e., flooded golf courses in flood plains), and traffic
disruptions caused by flooding will quickly gain the publics attention and will be a
stronger reason for citizens commitment to invest in pollution control. By building detention ponds or other improvements to solve flooding problems, the pollution problems
may also be largely solved. The Rouge River Project is showing the possibility of solving
a number of its wet weather pollution problems in the context of regional detention and
floodway improvement projects.
As stated in an earlier lesson, water quantity issues can often be integrated with other
public works plans and actions so that the solution positively impacts water quality as
well as water quantity. For example, cities need to control stream bank erosion in order
to protect property. Integrating water quality considerations into how the erosion control
is accomplished and maintained has a double benefit. Therefore, it is important to consider
water quantity issues in the development of the programs to address urban quantity of
flow issues.
Cost savings opportunities can be realized while addressing flow issues.
General
Lesson learned
Do not be afraid of taking advantage of dumb luck and legislative/political will.
This speaks for itself. The corollary is, bad luck also happens. When that occurs, learn
from the mistake or event and move on in a positive, smarter fashion.
510
Conclusion
The Rouge Project in southeastern MI is a working demonstration of a watershed-wide
approach to restoring and protecting an urban river system by using a cooperative, locally
based approach to pollution control. Since 1992, the Rouge Project has implemented over
$400 million of water pollution control, environmental restoration, and recreation projects.
Total expenditures under the project are expected to exceed $530 million by the year 2003.
(Note: these totals do not include related Rouge work by the city of Detroit for CSO
controls outside of the Rouge Watershed, which is estimated to be in excess of $2 billion.)
A number of successes have been achieved to date, but a number of challenges remain.
The year 1998 was a point when noticeable measurements of environmental progress along
the Rouge River could first be made. Dissolved oxygen was higher at several sites along
the river in the last two years, compared with earlier findings in 1994 and 1995. This
success can be attributed to CSO controls, illicit connection elimination, stormwater management, and better public, industry, and community awareness of pollution prevention.
Also, larger and more diverse species of fish are now sighted more often. The CSO basins
have collectively captured and sent to treatment an estimated 4 billion gal of overflow
per year since the first basins went online in 1997. The illicit discharge teams have eliminated 12.5 million gal/day of dry-weather discharges, and the Newburgh Lake project
has eliminated the PCB fish advisory and restored a major recreational resource in the
Rouge. The school-based education program operated by the Friends of the Rouge is in
nearly 100 schools, and the annual Rouge Water Festival draws over 1,400 students/year.
The CSO data suggest that large storms may have a different water quality impact
than small storms. This finding could result in significantly reduced costs when communities move from implementation of nine minimum controls to long-term control plans.
Recreational use in the watershed has increased when a stretch of the river was opened
to canoeing for the first time in 30 years. This action created new hopes and expectations
for further increased use of the river. The individual stormwater communities have begun
to understand how they are part of the water quality problems. Although they are traditionally very independent, they are finding ways to work together voluntarily to address
issues such as illicit connections, public education, and implementation of BMPs. The
Rouge Project is moving toward integration of all pollution sources and use attainment
into a unified, consistent watershed management approach, but this has not yet been
completely achieved.
The Rouge Watershed presents a unique management challenge because there are no
significant point sources that can be controlled by the action of a single agency, or from
which to readily establish an effluent trading scheme. Quite the opposite, the Rouge is
dozens of communities, hundreds of major commercial, industrial, and institutional properties, as well as hundreds of thousands of residential homeowners. The environmental
management goal of the project is the control of flow and wet weather pollution to achieve
flow and quality to meet water quality standards. The institutional management goal is
to find ways to effectively work with the parties within the watershed boundary to meet
the environmental goal.
The Rouge Project is an overwhelming success so far. Water quality is improving, and
the demonstration techniques have resulted not only in concrete and steel structures, but
in real institutional changes that integrate the work of stormwater and watershed improvement into the basic institutions of government. The Rouge Project approach demonstrates
that a watershed can be managed. Most important, the Rouge River is being restored.
It is hoped that by utilizing the information learned from the Rouge Project, others can
save considerable time and money as they implement a watershed management program.
Chapter II.9:
511
Acknowledgments
This chapter is a summary of select elements from the ongoing efforts of many individuals
and organizations involved in the restoration of the Rouge River. The work of Dale S.
Bryson to assist with development of the lessons learned is especially noted. The author
is honored to represent the numerous contributors of this successful partnership to restore
and protect a large urban waterway.
The Rouge River National Wet Weather Demonstration Project is funded, in part, by
the U.S. EPA Grant #XP995743-01, -02, -03, -04, -05, -06 and #C995743-01. The views
expressed by individual authors are their own and do not necessarily reflect those of the
EPA. Mention of tradenames, products, or services does not convey, and should not be
interpreted as conveying, official EPA approval, endorsement, or recommendation.
Literature cited
Bean, C., Schrameck, R., and Davidson, C., 1994 Rouge River remedial action plan update, MI Dept.
Natur. Resour. and Southeast MI Coun. Govts., Detroit, 1994.
Cave, K.A. and Bails, J.D., Implementing a model watershed approach through a state general
stormwater NPDES permit, Proc. WEFTEC 98 Conf., Orlando, FL, 1998.
Cave, K.A., Bryson, D.S., and Ridgway, J.W., Achieving multiple objectives through a single watershed plan, Proc. Watershed 2000 Conf., Vancouver, BC, 2000.
Center for Watershed Protection (CWP), The Practice of Watershed Protection: Techniques for Protecting
and Restoring Urban Watersheds, 2000.
Hufnagel, C.L. et al., What performance monitoring tells us about how to improve the design of
CSO storage/treatment basins, Proc. Watershed 2000 Conf., Vancouver, 2000.
Johnson, B.A. and Tuomari, D., Did you know . . . the impact of onsite sewage systems and illicit
connections on the Rouge River, Proc. Natl. Conf. Retrofit Opportunities for Water Resour.
Protection in Urban Environ., Chicago, 1998.
Kluitenberg, E.H., Kaunelis, V.P., and Johnson, C.R., Evaluation of in-stream impacts of CSO control
facilities, Proc. Watershed 2000 Conf., Vancouver, 2000.
Mullett, Jr., N., Bristol, C.R., and Koleda, K.P., Project technical support: GIS/sampling/modeling,
presented at 1994 Water Environ. Fed. Ann. Conf., Chicago, 1994.
Powell, J.A. and Bails, J.D., 2000. Measuring the soft stuff evaluation public involvement in urban
watershed restoration, Proc. Watershed 2000 Conf., Vancouver, 2000.
Rood, S., Koleda, K., and Bristol, C.R., Data access and Analysis using RPO DataView, Proc. WEFTEC
95, Miami Beach, 1995.
Rouge River National Wet Weather Demonstration Project, Wayne County, MI, http:\\www.rougeriver.
com, 19922001.
Rouge River National Wet Weather Demonstration Project, Rouge River Gateway Master Plan,
Wayne County, MI, 2001.
Schueler, T., Crafting better urban watershed protection plans, CWP, Watershed Protection Tech., 329,
1996.
Smith, E., Hughes, C., and Snyder, K., Communicating river water quality information to the public
using a graphical indicator approach, Proc. WEFTEC 97, Chicago, 1997.
U.S. EPA, Watershed approach framework, EPA840-S-96-001, Washington, DC, 1996.
U.S. EPA, Top 10 Watershed Lessons Learned, EPA840-F-97-001, Washington, DC, 1997.
512
Response
Looking beyond the end of the pipe
Perhaps the clearest message in this chapter by Kelly Cave is the concept that the effective
management of watersheds may often have just as much to do about managing people
as about managing pollution. As Cave describes, it is essential to gather the wide consortium of interested players, including governments, industries, environmental groups, and
private citizens, and include their viewpoints in consensus-building strategies. Community participation in identifying problems and suggesting solutions is often the key
attribute separating those watershed management projects achieving success such as
the Rouge River in Detroit and those that flounder. This is a recurrent theme discussed
elsewhere in this book: notably by Gail Krantzberg and Judi Barnes in Chapter II.6 and
by David Blau in Chapter II.7. Further, as Cave also points out, support from the general
public is essential for water sensitive planning to be self-sustaining through time, a topic
that Daniel Williams considers in Chapter II.11.
If such a process is to work, it is critical that the public become educated about the
nuances of watershed management. As Cave identifies, there is a need for effective transfer
of technology to the public in terms they can easily comprehend (e.g., converting phrases
such as meets water quality standards to something like fishable or swimmable).
The other major lesson from this chapter is the requirement to move beyond the belief
in end-of-pipe, cure-all solutions to water quality management. Instead, much more comprehensive planning can be achieved by focusing on subwatersheds, by recognizing that
advances are made incrementally in small steps, and by integrating stormwater control
projects into ongoing public works plans.
chapter II.10
Modeling a soil moisture index using

geographic information systems in a
developing country context (Thailand)
John S. Felkner and Michael W. Binford
Abstract
Land-cover change has important influences and impacts on biodiversity, climate, ecosystem structure, as well as the lands capacity for sustained use. This is a particular problem
in the developing world, where population, food, and land-use pressures are greatest.
The ability to characterize the spatial variability of soil water content in a simple yet
physically realistic way is highly valuable. Soil moisture is one component of soil quality
and agricultural suitability, and its measurement is important for land use, drought,
agriculture, irrigation, and construction planning, as well as for evaluation of ecological
system processes and health. Consequently, a need exists for techniques and models that
can rapidly estimate soil moisture across a landscape, particularly in a context where
access to rich data sources can be limited, and where inputs to spatially varying models
of general land-use change are needed.
This chapter describes a model for continuous soil moisture variation estimation in a
developing-country context, based on topographic, precipitation, land-cover, and soil
quality data inputs. Building on previous terrain-based indices, the model represents static,
or steady-state, conditions of relative soil water content and is thus not a water budget
estimation but provides an output that is an index correlating with soil moisture availability. The model calculates average, not actual, runoff and water availability and provides
a basic measure of general spatially heterogeneous soil moisture conditions in a developing-country context where data sources can be limited. The model inputs and outputs are
transformed into geographic information system (GIS) raster grids, facilitating future
incorporation into spatially explicit statistical or predictive land-use models.
The analysis was performed in two provinces of Thailand, which is an ideal country
for such a study because it has experienced rapid rates of land-use change as well as
strong population and economic growth in recent decades. Specifically, the model was
performed in the Thai provinces of Sisaket and Chachoengsao, selected because of their
high relative environmental and economic contrasts (Binford et al., 2002).
The model inputs are land cover, derived from classified satellite imagery, spatially
varying precipitation data, spatially varying soil quality and permeability conditions interpolated from point sources, and topography in the form of a digital elevation model (DEM).
1-56670-562-2/02/$0.00+$1.50
513
514
The land cover was derived using an ISODATA unsupervised clustering algorithm, with
clusters identified by comparison with ground-truth data as well as with other land-cover
images. The precipitation data were interpolated from a series of rainfall gauge stations
in Thailand and were averaged to obtain representative conditions. Soil sample data points
measuring soil field capacity were spatially interpolated and then converted to the U.S.
Natural Resources Conservation Service (NRCS) Hydrologic Soil Group (HSG) categories,
based on a observed relationship between field capacity and HSG categories. The HSG
output was combined with land cover to obtain NRCS runoff code number (RCN) values.
With these inputs, the GIS was used to measure the likely runoff and infiltration at each
pixel, and the surface runoff water flow and accumulation were modeled, using established mathematical relationships derived by the NRCS. Then a soil moisture index was
calculated as the tangent of the slope.
The results are consistent with expected soil moisture variation, considering land cover
(forest and agriculture), topography, and low-lying wetland areas. Finally, the output was
incorporated into a spatial statistical predictive model of land-use change and found to
act as a significant variable explaining change over time. In particular, areas of low soil
moisture were found to correlate with areas of deforestation due to agriculture, while
areas of high soil moisture values correlated with no land-use change.
Introduction
In the last 20 years, land-cover change whether it is deforestation in the tropics, urbanization, intensification of agriculture, or land degradation has accelerated as a result of
population pressure and economic development. Two hundred million hectares of land,
most classified as agriculturally marginal, were brought under cultivation for the first time
during the period from 1980 to 2000 (Dudal, 1980; Sanchez, 1976). Conversion of land on
this scale, unmatched in human history, is large enough to contribute significantly to
changes in global climate and global biogeochemical cycles (Turner et al., 1995; Woodward
et al., 1996). The effects of this global-scale change are significant, powerfully impacting
biodiversity, climate, ecosystem structure and the lands capacity for sustained use and
its ability to regain its original cover (F.A.O., 1981; Stern et al., 1992; WCED, 1987).
Major land changes were centered primarily in the mid-latitudes of the Northern
Hemisphere prior to the middle of the 20th century (Turner et al., 1990). In the last 50 years,
however, the major land-cover changes have occurred in the tropics, where cropland and
grassland/pasture expansion, deforestation, and urbanization, among other changes, are
increasing rapidly (Woodward et al., 1996).
Land-cover change has an important influence on biogeochemical cycling of gaseous
compounds of carbon, nitrogen, and other elements (Penner, 1994), and on water and
energy balances, all at regional to global scales. Changes in the distribution of land cover
alter regional and possibly global surface roughness, albedo, latent and sensible heat
flux, and actual evapotranspiration, all of which are important parameters for general
circulation models (Turner et al., 1995). Plants extract water from soils, returning it to the
atmosphere through transpiration. In the Amazon rainforest, for example, water recycling
by vegetation has had a direct impact on local and regional precipitation patterns (Victoria
et al., 1991). Thus, changes in land cover at regional scales can have direct impacts on
precipitation and local climate effects, and certainly land-use/cover change can affect
atmospheric conditions (Foley et al., 1994; Henderson-Sellers, 1990).
This relationship is a problem in the developing, tropical world, where the land-use
pressures of population increase are greatest. Consequently, a great need exists for the
Chapter II.10:
Modeling a soil moisture index using geographic information systems
515
monitoring, quantitative assessment, and predictive modeling of land-use change (Behera

et al., 1996; Rao, 1999; Robinson et al., 1994; Turner et al., 1995; Wald, 1999). Ideally, these
models would be able to predict land-use change patterns and processes at a regional or
local scale as well as continental or global, to quantify the changes and types of changes,
and to provide insight into the drivers of these changes (Bockstael and Irwin, 1999). The
derived information from such models can be used for more effective land-use planning,
preservation of ecologically fragile areas, and possibly aid in poverty alleviation.
Soil moisture estimation is important

The importance of the variation in surface soil moisture, including the location of areas
of very high water saturation, is well appreciated in disciplines including engineering,
soil conservation, forestry, agriculture, and hydrology because of the effect on soil strength,
erosion, soil aeration, plant ecology, secondary salinization, and storm runoff (OLoughlin,
1986). The ability to characterize the spatial variability of soil water content in a simple
yet physically realistic way is of major importance (Barling et al., 1994). There is strong
support in the land-use change modeling literature for correlations between soil quality
and the probability of land-use change. Superior soil quality tends to promote agriculture,
which initially causes deforestation. Over time, however, rich soil can act as a brake on
the loss of remaining forests because high agricultural productivity can reduce the economic need for deforestation (Crist, 1986; Katawatin et al., 1994, 1996a, 1996b; Rindfuss,
1998). Soil moisture is one component of soil quality and agricultural suitability. Thus, the
estimation of soil moisture in a landscape is an important goal for effective land-use change
modeling and land planning.
This goal is especially applicable in the developing world, which has the highest rates
of land-use change and land degradation, and more than 70% of the worlds total population (Skole, 1994). Soil moisture estimation in the developing world is important for
drought, agriculture, irrigation, and construction planning (Binnie, 1997; Chiew et al., 1995;
Damota et al., 1992; Fu and Gulinck, 1994; Kimmage and Adams, 1992; Kobayashi, 1996;
Winkler and Ulehla, 1992). For example, land conversion to agricultural production in
Thailand during the past two or three decades has resulted in severe soil erosion, land
degradation, and decrease of the soil moisture content due to the lack of implementation
of appropriate soil and water conservation measures (Kobayashi, 1996). Finally, soil moisture estimation can also be particularly useful in the delineation of wetland areas (Costanza
and Sklar, 1984), which are in turn integral to the health of hydrologic and biotic ecosystems (Schulze and Mooney, 1993).
Goals of this modeling effort

Given this context, there is a need for techniques and models that can rapidly estimate
soil moisture across a landscape, particularly in developing countries where both time
and data are likely to be scarce (Binnie, 1997; Florinsky, 1998; Rango and Shalaby, 1998;
Turner et al., 1995). The use of GIS and remote sensing tools offers a possible solution to
meet this need, as lower costs and increasing power have developed dramatically in recent
years (Goodchild, 1994; Goodchild et al., 2000; Robbins and Phipps, 1996). This includes
the development of raster-format hydrologic models and watershed databases (Julien
et al., 1995; Schloss, this volume; Mueller et al., this volume). In recent years, there has
been considerable effort devoted to utilizing GIS to provide inputs (soils, land use, and
topography) for comprehensive simulation models and to display model outputs spatially
516
(Arnold et al., 1998; Bonham-Carter, 1994; Cantwell, this volume; Julien et al., 1995; Moore
and Hutchinson, 1991; Rewerts and Engel, 1991; Robbins and Phipps, 1996; Srinivasan
and Engel, 1991; Srinivasan and Arnold, 1993; Tarboton et al., 1991).
The goal of this chapter is to describe the development and implementation of a model
that will estimate relative soil moisture availability across a large landscape in a developing-country context where data are scarce. The soil moisture estimation output in spatially
continuous raster GIS format would be useful as input to spatially explicit, statistical,
predictive land-use models that predict future land use at a provincial scale (Agarwal
et al., 2000; Chomitz and Gray, 1996; Dimyati et al., 1996; Felkner, 2000; Kaimowitz and
Angelsen, 1998). Specifically, some land-use prediction models require a relative soil
moisture estimate with values at all points on the landscape.
The soil moisture model described here relies primarily on topographic information,
which, in the form of contour data or a DEM may be the only type of data available in a
developing-country context (OLoughlin, 1986). In the same vein, innovative and alternative methods for deriving spatial variation in precipitation patterns and in soil characteristics may have to be devised, as point data may be the only available data sources for
these inputs. For this model, spatial variation in precipitation patterns and in soil qualities
were derived from point data sources.
Our model uses the following data inputs: land cover derived from satellite remote
sensing data; average annual precipitation data from meteorological stations; field samples
of soil conditions; and a DEM interpolated from contour elevation information. The model
is designed to produce raster GIS output. The approach is applicable at regional or
provincial scales, rather than solely country or continental ones. The approach thus fulfills
a perceived need in the literature of land-use change monitoring and modeling for
regional, rather than country or larger-scale, efforts (Bockstael and Irwin, 1999; Kaimowitz
and Angelsen, 1998; Lambin et al., 1999).
Previous research on soil moisture indices

The approach taken here relies heavily on topographic data to create a continuous grid
of relative soil moisture. The importance of topography has been recognized as the primary
determinant in regulating variation in surface soil moisture in several experimental studies
(Anderson and Burt, 1977; Anderson and Burt, 1978; Beven, 1978; Dunne and Black, 1970;
OLoughlin, 1986). Topography plays an important role in the hydrologic response of a
catchment to rainfall and can affect, for example, the location of zones of surface saturation
and the distribution of soil water. The automation of terrain analysis and the use of DEMs
have made it possible to quantify easily and accurately the topographic attributes of a
landscape.
A number of terrain-based indices have been derived and relationships have been
sought between these indices and a range of hydrologic processes (Barling et al., 1994;
Beven and Kirkby, 1979; Burt and Butcher, 1986; Moore et al., 1988, 1991, 1993). The soil
moisture index approach has typically been based on the use of simplified representations
of the underlying physical processes (soil qualities, precipitation amounts, landscape
cover, etc.) while at the same time striving to include most of the primary factors that
regulate the systems behavior. This approach sacrifices physical sophistication and resolution but enables relatively easily computed estimates of the spatial patterns in a landscape (Barling et al., 1994; Moore et al., 1991, 1993; Moore and Hutchinson, 1991).
In some cases, particularly in developing-country contexts, a topographic map or DEM
may be the only relevant information available for the area in question. Consequently, the
Chapter II.10:
517
terrain-based index approach may be best suited, and consistent, with the level of available
data and the precision with which many of the management and planning questions both
need, and can, be answered (Barling et al., 1994). Thus, this approach is logical for a
developing-country context when a rapid assessment is needed as a component or input
into larger land-use systems analysis.
This model represents a static, or steady-state, condition of relative soil water content
and the location and size of zones of surface saturation (Beven and Kirkby, 1977, 1979;
OLoughlin, 1986). Thus, the model is not a water budget estimation but provides an output
that is an index correlated with soil moisture availability. High values mean the soil is moist
or even saturated; low values mean the soil is dry. The model calculates average, not actual,
runoff and water availability and provides a basic measure of general spatially heterogeneous
soil moisture conditions considering topography, generalized spatially heterogeneous average precipitation amounts, generalized spatially heterogeneous soil types, and specific land
cover. The approach taken here modifies previous approaches to adjust to limited availability
of input data, specifically a lack of appropriate and comprehensive soil data, and a lack of
detailed, spatially heterogeneous precipitation data. The output is a GIS raster grid that
provides a generalized reflection of the ability of the soil to hold water.
Description of study area and data sources

Thailand is an ideal country for such a study, because it has experienced extensive deforestation, land-use change, and urbanization within the last 40 years (Lombardini, 1994).
Originally rich in tropical rainforests and high plant diversity, Thailand has become
heavily agricultural since the 1960s. Also, Thailand has a wide variation in regional
economic conditions, including income and education levels. Thus, analysis of different
regions of Thailand provides a cross-section of tropical developing economic conditions.
The study area consists of two provinces of Thailand: Chachoengsao (5,351 km2) and
Sisaket (8,840 km2). The larger land-use predictive model was applied to the two provinces
in order to provide contrasting case studies, both for analysis and verification purposes.
These two provinces were specifically selected to present a maximal contrast both economically and environmentally, based on a consideration of income levels and income
growth, agricultural output, precipitation, and soil fertility (Binford et al., 2002).
The province of Chachoengsao is located immediately east of the Bangkok metropolis,
while Sisaket is located in the poor northeast region, on the border with Cambodia (see
Figure II.10.1). Whereas Chachoengsao is highly developed, with extensive infrastructure
and industry, Sisaket is one of the poorest provinces in the country, with very limited
infrastructure and urbanization. Chachoengsao boasts a broad industrial base, with intensive and diverse agricultural practices and excellent proximity both to ocean ports and to
Bangkok. By 1990, more than 90% of the population of Sisaket was engaged in agriculture,
compared with approximately 66% of Chachoengsao (Thai Census Dept., 1990). Educational attainment and literacy rates are considerably lower, and child mortality rates are
higher, in Sisaket. Physically, the two provinces also contrast: Chachoengsao is on the flat
coastal plain and has much richer soils than Sisaket, which is inland and has significant
relief especially in the south (Daniere, 1999; Hirsch, 1993; Binford et al., 2002). Both
provinces have been extensively deforested in the last 40 years (Lombardini, 1994), mostly
for agriculture except in areas of topographic relief (Figures II.10.2 and II.10.3). Despite
their relatively equal amounts of deforestation, the two provinces provide excellent case
studies because of their simultaneous economic and environmental differences.
518
India
China
Bangladesh
Myanmar (Burma)
Hong Kong
Laos
Paracel Islands
Sisaket
Thailand
Cambodia
Bangkok
Vietnam
0
300
Kilometers
Chachoengsao
Spratly Islands
Malaysia
Brunei
Indonesia
Figure II.10.1 Location of Sisaket and Chachoengsao provinces, Thailand.
Much of the data for this study were gathered by a research group headed by Dr.
Robert Townsend of the University of Chicago, Dept. of Economics, with funding from
the National Science Foundation and the National Institutes of Health (Binford et al., 1998;
Townsend and Lim, 1998) in cooperation with the Thai government and the Thai Bank of
Agriculture and Agricultural Collectives (BAAC). Precipitation data were interpolated
from measurements taken by the Thai government at a series of rainfall gauges distributed
throughout the country (Figure II.10.4) that included monthly precipitation amounts from
the 1950s to 1986. Soil quality and conditions were also interpolated from samples taken
in each of the provinces in a number of villages as part of an extensive socioeconomic and
environmental survey conducted in 1997 (Binford et al., 1998) (Figure II.10.5). The soil
samples were analyzed for field capacity (the volume of water remaining in a soil after
gravity drainage of saturated soils, representing the maximum amount of water available
for plants) among other soil qualities (see Figures II.10.6 and II.10.7).
Land-cover classes were derived from interpreted Landsat Multi-Spectral Scanner
(MSS) and thematic mapper (TM) data (Figures II.10.2 and II.10.3). The DEM was interpolated from vectorized contour lines at 10-m intervals on 1:50,000 scale topographic
quadrangles produced by the Thailand Dept. of Mapping (maps covering the ThailandCambodia border in southern Sisaket were unavailable).
The soil moisture procedure estimates relative soil moisture content conditions on
average during the 1980s. This was done for purposes of a larger predictive land-cover
modeling work, which predicted 1999 land-cover conditions based on historical trends in
Chapter II.10:
519
Sisaket 1979
Land Cover
Study Area
Sisaket Provincial Boundary
Land Cover Categories
Water
Bright Soil/Bare Earth
Agriculture
Wetland
Secondary Forest Cover
Forest
Urban
30 Kilometers
Sisaket 1989
Land Cover
Figure II.10.2 1979 and 1989 Sisaket land cover derived from Landsat MSS and TM imagery.
land-use change from 1979 to 1989 (Felkner, 2000). Therefore, all inputs to the larger model
were designed to represent approximate average conditions in the 1980s. Consequently,
precipitation data for the soil moisture model was based on monthly averages from the
1980s. Also, land-cover classifications were derived from Landsat MSS images from 1979
and 1982, to approximate conditions in the early 1980s. Although the soil samples were
obtained in 1997, we assume that the general spatial arrangement of intraprovincial soil
conditions would not have changed appreciably since the early 1980s.
Methods
The model algorithm is shown in two flow charts in Figures II.10.8 and II.10.9. Figure II.10.8
depicts the process of preparing the four primary inputs to the soil moisture index calculation: a runoff curve number (RCN) grid reclassified from the HSG grid, in turn derived
from a soil field capacity grid interpolated from the original soil samples; a precipitation
grid interpolated from Thai rain-gauge measurements; and two input grids derived from
520

Chachoengsao 1979
Land Cover
Study Area Boundary

Chachoengsao Provincial Boundary
No Data Areas
Urban
Water
Agriculture
Primary Forest
Secondary Forest
Chachoengsao 1989
Land Cover
Study Area Boundary

No Data Areas
Water
Wetland Areas
Agriculture
Primary Forest
Urban
30 Kilometers
Figure II.10.3 1979 and 1989 Chachoengsao land cover derived from Landsat MSS and TM imagery.
the DEM a slope grid (with every cell having the value of its slope in degrees) and a
flow-direction grid (every cell contains a numerical representation of the direction vector
of water flow across the surface) (Jenson and Domingue, 1988). Figure II.10.9 depicts the
soil moisture modeling process once the input grids (which are rectified and have equal
numbers and sizes of grid cells) are created. All grids used cell sizes of 60 by 60 m.
Creation of the DEM

Topography is the primary determinant of relative soil moisture content in the model.
Specifically, the model requires the use of a DEM in order to measure slope, flow direction,
and flow accumulation (the uphill area in number of grid cells from which water
would flow into any particular cell) (Jenson and Domingue, 1988; Tarboton et al., 1991).
The DEM was interpolated from vectorized contour lines, as described in the following
paragraphs. Because of the need for the measurement of flow direction and flow accumulation, a depressionless DEM was generated.
First, contour lines were digitally extracted and vectorized (represented as vector GIS
objects) from the Thai governments 50,000-scale topographic map quadrangles, using
automated vector extraction software operating on scanned versions of the maps. Once
the contour lines were vectorized, they were assigned appropriate elevation values. This
process of extraction was manually supervised during the automated extraction routines,
Chapter II.10:
521
Figure II.10.4 Locations of Thai rainfall-gauge stations, Thailand.
and then the output was error-checked and cleaned by hand. The interpolation of a
continuous raster surface was performed with the topogrid command in ArcInfo GIS
software that is based on the ANUDEM program developed by Michael Hutchinson
(Hutchinson, 1988, 1989, 1993, 1996; Hutchinson and Dowling, 1991). The method uses an
iterative finite-difference interpolation technique that is essentially a discretized thin-plate
spline approach (Wahba, 1990), where the roughness penalty has been modified to allow
the fitted DEM to follow abrupt changes in terrain, such as streams and ridges. The
procedure is specifically designed to consider hydrologic drainage features, which are the
primary erosive and shaping force on a landscape.
The DEMs were used in the soil moisture index model for hydrologic analysis, so it
was necessary to process the DEM surfaces iteratively to remove sinks. For many hydrologic analyses run on a GIS grid, the flow of water across the surface is often simulated
as a function of gravity (a function of the elevation differences between grid cells). A sink
is a location that will trap water running into it essentially a well, depression, or pit
on the surface of the DEM. Simulated water movement across the DEM surface is prevented from running uphill and will thus get stuck in the sink. Often, the sinks are
522
30 Kilometers
30 Kilometers
Figure II.10.5 1997 soil sampling sites, Sisaket and Chachoengsao provinces, Thailand.
artifacts of data processing errors and can be quite small (Mark, 1988). The process used
for creating the depressionless grid is shown in Figure II.10.12.
The final output of the DEM creation process was smoothed to remove aberrant
artifacts using cubic convolution smoothing (Hunter and Goodchild, 1997; Jenson and
Domingue, 1988; Kyriakidis et al., 1999). Once depressionless DEMs were created, surfaceanalysis algorithms in ArcInfo GIS software were used to calculate slope and flow direction
grids (Greenlee, 1987; Jensen and Domingue, 1988).
Land-cover classification
Land cover for the early 1980s for both provinces was classified from Landsat MSS data
using the ISODATA unsupervised clustering approach implemented in the software program ERDAS Imagine, in combination with ground-truth data we collected in Thailand
in 1998 (Sabins, 1987; Tou and Gonzalez, 1977). Two sequential Landsat MSS 1979 images,
acquired on October 6, 1979, were a mosaic map that provided almost complete coverage
of Sisaket. For Chachoengsao, a single Landsat MSS scene covered the southern two-thirds
of the province, acquired on October 10, 1979. The satellite data were rectified to the UTM
coordinate system, using the Indian 1975 datum and the Everest 1830 spheroid.
In August 1998, 80 and 40 widely dispersed ground-control and land-cover data points
were collected in Sisaket and Chachoengsao, respectively, using a Global Positioning
System (GPS) receiver, to allow geo-rectification and classification. The geographic locations
of these points were tied in the field to their visible locations on the images, and then they
were used as tie-point inputs into a bilinear interpolation for rectification (Jensen, 1996),
with approximately a one-pixel root mean squared error (RMS).
Chapter II.10:
523
Figure II.10.6 1997 Sisaket Province soil field capacity sample locations and values.
Although the 1979 Sisaket MSS scenes were acquired on the same day, a multipledate empirical radiometric normalization was performed using a regression approach with
pseudoinvariant ground targets (water bodies; dry, barren soil; etc.) to minimize differences in atmospheric conditions, satellite acquisition angle, lighting conditions, and detector calibration errors (Jensen, 1996). The images were then were made into a mosaic map
with a feathering interpolation system along the overlapping edges.
To aid in the process of assigning land covers to each cluster (agriculture, forest, built,
or other), we used natural-color composite images in conjunction with the groundcontrol and ground-truth data. Normalized difference vegetation indices (NDVI) images
helped distinguish forested areas (Rouse et al., 1973). Tasseled cap, or KauthThomas
transforms (Crist, 1986; Crist and Kauth, 1986), were calculated for the raw imagery to
identify the soil brightness index (SBI), the green vegetation index (GVI), the yellow stuff
index (YVI), and the nonesuch index (NSI). In addition, de-haze images were generated
as a product of a transform similar to KauthThomas (Richards, 1993).
All these derived data layers were consulted closely in assigning labels to each cluster.
Initially, 100 clusters were created for each image, and each was assigned a land-cover
category. Extensive cluster busting (Jensen, 1996) separated the clusters that could not be
clearly identified (e.g., extending both into built areas and into agriculture). The results
are displayed in Figures II.10.2 and II.10.3.
524
Figure II.10.7 1997 Chachoengsao Province soil field capacity sample locations and values.
Creation of the precipitation grid

Our model works most effectively with accurate spatially heterogeneous precipitation
data, but these data were not available at the subprovincial level in Thailand. Rainfall
data were obtained from a network of rain-gauge stations distributed throughout Thailand
(Figure II.10.4). Monthly precipitation values during the 1980s were averaged for each
precipitation station. Then, a simple inverse distance weighted (IDW) interpolation was
performed from all the national rain-gauge stations to create a continuous grid of spatially
varying precipitation values across each province. This method was the best option given
readily available data inputs, although it yielded a poor degree of spatial heterogeneity
for each study province.
Estimation of hydrologic soil groups for Thailand using field capacity

measurements
Previous static soil moisture approaches have taken advantage of an empirical method
and several derived mathematical relationships developed by the U.S. Natural Resources
Conservation Service (NRCS, formerly Soil Conservation Service, or SCS). This approach
uses HSG and empirically derived RCNs to estimate the ratio of runoff to infiltration for
a given type of land-cover/soil-type combination (U.S. Dept. of Agriculture, 1972). The
ratio of infiltration to runoff depends on both the degree of imperviousness of the land
cover and the permeability of the soil type. Pavement or building roofs are practically
impervious, while certain kinds of forest, prairie, or agricultural lands are highly permeable.
The HSG soil classification system was developed as a way of summarizing a soil
series hydrologic effects, and NRCS categorized every soil series in the United States into
Chapter II.10:
Soil Samples:
Field Capacity
Interpolate
Precipitation
Data
Interpolate
Field
Capacity
Grid
Digital
Elevation
Model
Precipitation
Grid
GIS
Surface
Analysis
Established
Relationship
Hydrologic
Soil Group
Grid
HSG
Look-up
Table
525
Land
Cover
Grid
Slope
Grid
FlowDirection
Grid
Recode
Runoff Curve
Number Grid
Soil Moisture
Index Process
(See Figure
II.10.9)
Figure II.10.8 Soil moisture model flowchart: Building of inputs.
four groups, lettered A through D (see Figure II.10.13). Group A has the highest infiltration
capacity and is the least likely to generate runoff, while group D has the lowest infiltration
capacity and highest runoff. Given a soil series hydrologic soil group, the RCN is derived
from a combination of HSG and land cover, according to published NRCS lookup tables
(Ferguson, 1998). Thus, the NRCS curve number summarizes the combined effects of soil
and land cover. RCN was derived from experimental studies establishing specific relationships between rainfall and runoff for different types of soil and land cover and can
vary, theoretically, from 0 to 100 as soil and land-cover conditions generate increasing
amounts of runoff. In practice, however, RCN is seldom less than 30, and the highest RCN,
for impervious pavement and roofs, is 98.
HSG soil classifications do not exist for Thailand. Furthermore, accurate soil maps for
each province were not obtainable; although, soil samples taken in 1997 in both provinces
(Figure II.10.5) included a measurement of soil field capacity (among other soil variables),
shown for each province in Figures II.10.6 and II.10.7. A method of estimating HSG for
the entire area of each province was devised using field capacity measurements as a proxy
for hydrologic group, based on an observed relationship between field capacity and HSG
observed in soils from Puerto Rico (Soil Conservation Service, 1975), as shown in
Figure II.10.14. Soils in Puerto Rico and Thailand, both tropical climates, share many
similar properties, and the USDA conducts soil surveys in Puerto Rico, which include
measurements of field capacity and simultaneous designation of hydrologic groups. HSG
is primarily a classification of the soils tendency to allow water permeability, and field
capacity is related to soil texture and infiltration capacity (Dunne and Leopold, 1978).
526
Process from Figure II.10.8

RCN
Precip
Flow
Slope
Grid
Grid
Direct
Grid
SGrid = 254(100/RCNGrid-1)
Sgrid
RunoffGrid = ((PrecipGrid 0.2 *

Sgrid)^2) / (PrecipGrid + 0.8 * Sgrid)
Runoff
Grid
M3RunoffGrid = RunoffGrid /
3
1000 * Cell Area (3600 m )
M3Runoff
Grid
flow accumulation
M3Accum
Grid
SMIndex = ln (M3AccumGrid /
tan (SlopeGrid))
Soil
Moisture
Index Grid
Figure II.10.9 Soil moisture model flowchart: Calculation of a relative soil moisture index.
Because the modeling process required a continuous and complete raster grid of the
entire process, however, the first step in the estimation of HSG was the interpolation of
the field capacity measurements to derive a continuous field for the entire provincial areas.
This step was performed with a full recognition that soil spatial variation is by no means
always continuous, yet without such an interpolation the model would have no soil spatial
variation, whereas such an interpolation gives at least a rough approximation of provincial
soil variation. Based on previous studies into the effectiveness of different point interpolation methods for soil samples (Laslett et al., 1987; McBratney and Webster, 1986), a
kriging approach was used (Bonham-Carter, 1994; Myer, 1991). Once the point samples
were effectively kriged to obtain complete raster surfaces, these were converted to HSG
grids using a lookup table derived from the observed Puerto Rican relationship. The final
Chapter II.10:
527
Study Area Boundary

Elevation in Meters
119-120
121-126
127-130
131-135
136-139
140-145
146-151
152-160
161-175
176-342
20 Kilometers
Figure II.10.10 Sisaket Province digital elevation model (DEM).
step in the process was to convert the HSG grids into RCN grids with the addition of the
1979 land-cover grids (see Figures II.10.2 and II.10.3): each pixel was assigned an RCN
value as a function of both its HSG designation and its corresponding land cover according
to the lookup tables (Ferguson, 1998). The resulting RCN grid was input into the soil
moisture index calculation process, along with precipitation, slope, and flow direction
grids. The final RCN grids are shown in Figure II.10.15.
Soil moisture index estimation process

The process of calculating the soil moisture index from the input grids is depicted in
Figure II.10.9. The process was automated into Arc Macro Language (AML) and run in
ArcInfo software. The approach combines the NRCS methodology for estimating runoff
based on RCN and the calculation of the accumulation of uphill water flow into each
downhill grid cell on a depressionless DEM in order to obtain the amount of runoff
528
30 Kilometers
Study Area Boundary

Elevation in Meters
1-3
4
5-30
31-48
49-62
63-80
81-98
99-120
121-149
150-718
Figure II.10.11 Chachoengsao Province digital elevation model (DEM).
flowing through each grid cell. This flow accumulation grid (Jenson and Domingue,
1988) is divided by the tangent of the slope, and the natural log is taken of that quotient
to produce the index. The NRCS methodology specifically calculates a retention factor, s,
defined as (for precipitation in mm)
s = 254
100
1
CN
where CN is the runoff curve number. The retention Sgrid is next combined with the
precipitation grid in the SCS curve number, equation which predicts surface runoff:
Runoffgrid =
(precipgrid 0.2 sgrid)2

(precipgrid + 0.8 sgrid)
resulting in the creation of the runoffgrid. Finally, the soil moisture index is calculated as
a function of the tangent of the slope:
soil moisture index = ln m3accumgrid tan(slopegrid)
The final soil moisture index grids for Sisaket and Chachoengsao are shown in
Figure II.10.16.
Chapter II.10:
Thai 50,000
Scale Royal
Maps
Find
flow direction
for all grid cells
529
Depressionless
Digital
Elevation
Model (DEM)
Vectorization
of
contour lines
Locate all
DEM sinks
Interpolation
on contours
to create DEM
Find watershed
areas for each
sink
repeated
iteratively
until all
sinks
removed
Find total
depth of each
sink watershed
Depressionless
DEM
Fill sink to
watershed
depth
cubic
convolution
smoothing
Figure II.10.12 Process for creating a depressionless DEM.
Results
Land-cover classifications for each province for 1979 and 1989 are shown in Figures II.10.2
and II.10.3. For the most part, the major land-cover classes were easily differentiated,
especially forest and agriculture classes. Agriculture is the primary land use in both
provinces, and extensive deforestation between 1979 and 1989 occurred as forest was
converted to agricultural uses, as can be seen in the figures. Remaining forests tend to be
clustered in the high topographic relief areas, in the south of Sisaket and the east of
Chachoengsao. The densest urban areas tend to be clustered near the major rivers and, in
the case of Chachoengsao, along the ocean shorelines in the southwest of the province.
Despite the forest clustering in the higher relief areas, in general both provinces are
relatively flat topographically and conducive to extensive agriculture.
Classification problems certainly were encountered, however, in the delineation of
built land cover (urban, residential, dense settlement areas). It quickly became apparent
that these were by far the most problematic to identify and label, as clusters that certainly
appeared in areas that were clearly built (major urban areas) also often appeared in areas
that were clearly agriculture, or even forest. Cluster-busting help to split these clusters
along land-cover lines; although, certain clusters remained that were extremely difficult
530

Hydrologic Soil Group definitions, as used in the SCS method:
Group A:
Group A soils have low runoff potential. They have high infiltration rates even when thoroughly
wetted. They consist chiefly of deep, well to excessively drained sands or gravels. This group also
includes sand, loamy sand, and sandy-loam that have experienced urbanization but have not been
significantly compacted.
Group B:
Group B soils have moderate infiltration rates when thoroughly wetted. They consist chiefly of
moderately deep to deep, moderately well- to well-drained soils with moderately fine to moderately
coarse textures. This group also includes silt loam and loam that have experienced urbanization but
have not been significantly compacted.
Group C:
Group C soils have low infiltration rates when thoroughly wetted. They consist chiefly of soils with
a layer that impedes downward movement of water and soils with moderately fine to fine textures.
This group also includes sandy clay loam that has experienced urbanization but has not been
significantly compacted.
Group D:
Group D soils have high runoff potential. They have very low infiltration rates when thoroughly
wetted. They consist chiefly of clay soils with high swelling potential, soils with permanent high
water tables, soils with clay pans or clay layers at or near the surface, and shallow soils over nearly
impervious material. This group also includes clay loam, silty clay loam, sandy clay, silty clay, and
clay that have experienced urbanization but have not been significantly compacted.
(Source: U.S. Conservation Service, Urban Hydrology for Small Watersheds, 2nd ed., Technical Release
55, Washington, 1986.)
Figure II.10.13 Hydrologic Soil Group (HSG) definitions.
Relationship Between Soil Hydrologic Group and Field Capacity in Puerto Rico
Observed Field Capacity Values
25
20
15
y = 4.3312x + 3.6413
R2 = 0.6585
10
0
0
0.5
1.5
2.5
3.5
4.5
Soil Hydrologic Group: 1 = A, 2 = B, 3 = C, 4 = D
Figure II.10.14 Observed relationship between soil field capacity values and Hydrologic Soil Group
(HSG) categories in Puerto Rican soils.
Chapter II.10:
531
Study Area Boundary

Runoff Curve Number
57-61
62-65
66-69
70-74
75-78
79-82
83-87
88-91
92-95
96-100
30 Kilometers
Study Area Boundary

Runoff Curve Number
33-39
40-46
47-53
54-59
60-66
67-73
74-79
80-86
87-93
94-100
30 Kilometers
Figure II.10.15 Runoff curve number (RCN) grids for Sisaket and Chachoengsao provinces.
532
to classify. Thus, the classification of built land use in all of these classifications is certainly
the class with the lowest accuracy confidence. The identification of forest and agriculture
classes is much stronger.
The Sisaket and Chachoengsao RCN grids are depicted in Figure II.10.15. The influence of the major inputs that have delineated runoff code number can be seen in the
outputs, namely land-cover and soil characteristic patterns. The figures reveal individual
runoff/infiltration permeability patterns for each province. Sisaket shows higher degrees
of runoff in the low-lying river/wetland areas in the north and near reservoirs and lakes
in the south, as well as higher amounts of permeability in the southern forested areas,
and lower amounts of permeability with higher runoff in the central agricultural regions.
These trends are, however, mitigated in a number of cases by underlying soil quality
patterns, based on interpolations from soil sample measurements, with soils in the south
having less permeability than those in the central areas. This appears logical because the
central provincial areas have been extensively used for agriculture in preference to the
southern higher relief areas. The RCN models do not consider topography, however, and
thus this may indicate that the central-area soils are superior for agriculture due to higher
inherent permeability qualities, instead of only because they are topographically flat.
Chachoengsao shows generally higher runoff numbers in the western half of the
province, which corresponds to the half that is much more highly developed, urbanized,
and topographically flatter. Highest permeability appears as a central band running
northsouth in the center of the province and corresponds with areas used extensively for
agriculture with little remaining forest cover even as of 1979.
The final soil moisture model output grids are depicted in Figure II.10.16. The results
are consistent with a generalized representation of relative subprovincial soil moisture
levels. In Sisaket, very high soil moisture values clearly appear in the major river valley
in the northern portion of the province. At the same time, the higher-relief forested areas
in the south have considerably lower soil moisture values, as would be expected due to
less permeable soils and much higher slopes to encourage runoff. In general, the large
agricultural areas in the central and west-central portions of the province have higher soil
moisture values, corresponding to the general topographic pattern of the province as well
as the pattern of primary agricultural uses established by 1979. On a smaller scale, the
model has accurately increased soil moisture values in small valleys between ridges
throughout the province.
Chachoengsaos results are somewhat distorted by the extremely flat portions of the
DEM in the western part of the province, but generally the soil moisture output reveals
that the eastern half of the province is drier. A comparison of the soil moisture output
with the DEM for Chachoengsao reveals that the model has also accurately increased soil
moisture values in lower elevation, with ridgetops appearing drier.
These results suggest the potential utility of a spatially continuous soil moisture
output. Such information can be compared with data on land-use change, to look for
correlations: if deforestation or land-use change does correlate with higher soil moisture
values, then predictions can be made for future likely areas of deforestation. At the same
time, inferences can be made regarding the need that agriculture has for high soil moisture
values. In addition, if correlations between other types of land-use change such as
conversion of agriculture or forest to urban and high or low soil moisture values can
be established, such linkages could be useful for planning and management purposes.
However, a more rigorous method of assessment is needed beyond simply a visual
assessment. Such comparisons are possible using GIS outputs, such as this soil moisture
model grid output. The soil moisture models displayed in Figure II.10.16 were inputted
into a spatial statistical model of land-use change for both Sisaket and Chachoengsao
(Felkner, 2000). The approach carefully measured actual land-use change from 1979 to
Chapter II.10:

Study Area Boundary
Soft Moisture Index
33-39
40-46
47-53
54-59
60-66
67-73
74-79
80-86
87-93
94-100
30 Kilometers
Study Area Boundary

Soft Moisture Index
-523-2051
2051-4626
4627-7201
7202-9776
9777-12351
12352-14926
14927-17501
17502-20076
20077-22651
22652-25226
30 Kilometers
Figure II.10.16 Relative soil moisture index grids for Sisaket and Chachoengsao provinces.
533
534
1989 in both provinces along four basic change categories: conversion of forest to agriculture; conversion of forest to built uses; conversion of agriculture to built uses; and no
change. Using this record of actual change as the dependent variable, a selected set of
economic and environmental inputs were used as independent predictor variables using
a classification-tree statistical technique (Venables and Ripley, 1994; Clark and Pregibon,
1992) to establish correlations between specific types of change and specific input variables.
The soil moisture model constituted one of the environmental variables, along with
topography, a proximity to forest edge model, and a proximity to water model. All inputs
were weighted equally, and the statistical algorithm chose variables to partition the data
into homogenous partitions by calculating splits along specific variables that resulted in
the greatest reduction in variance between split data partitions.
The results revealed that the soil moisture index variable served as an important explanatory variable in deriving a model for land-use change, based on the inputs. While soil
moisture played less of a role than did elevation or proximity to forest edges in explaining
deforestation, it did factor significantly ahead of proximity to water bodies. The classification-tree algorithm chose soil moisture as a splitting variable by the fourth level of splits
in the land-use change model tree. Also, areas of lower-than-average soil moisture values
were specifically highly correlated with conversion of forest to agriculture change, while
areas of high soil moisture correlated highly with areas of no change. These results held
for both provinces, but the derived tree models revealed that soil moisture played a more
important role in explaining land-use change in Sisaket than in Chachoengsao and was
specifically more significant (especially with high correlations with deforestation to agriculture) in southern Sisaket. The derived tree statistical model, using the soil moisture
output displayed in Figure II.10.16, was then used to predict future land-use change with
over 80% accuracy (Felkner, 2000).
These results suggest that, during the 1980s, deforestation to agriculture tended to
occur in low soil moisture areas possibly because the relatively flat, low-lying high soil
moisture areas (most conducive to agriculture) had already been cleared for crops in
previous decades, and subsequent deforestation tended to occur in the drier, higher-relief
areas with relatively impermeable soils. This could have been more significant in Sisaket
than in Chachoengsao because, during the 1980s, Sisaket was relatively less deforested
than Chachoengsao. Also, the correlation between high soil moisture with no land-use
change could simply be a reflection of the clearing of high soil moisture areas for agriculture in previous decades: since these areas continue to be highly productive, they are not
converted to other uses. Thus, the results should not necessarily be interpreted as indicating that low soil moisture inherently correlates with deforestation, but instead particular
temporal and geographic economic conditions should be considered for context before
conclusions are formulated.
Further analysis to consider more directly specific correlations between soil moisture
variation and land use, or land-use change, could be achieved by running the classificationtree statistical algorithm using land use or land-use change as dependent variables and
the soil moisture grids as independent predictors. The results would give very specific
correlations between particular soil moisture index ranges and land use or land-use change
types.
Thus, the approach taken here allows for the relatively rapid estimation of relative
soil moisture amounts in a developing country context given limited data inputs, primary
reliance on topographic map or DEM elevation data, and reliance on remote sensing for
rapid and accurate land-cover derivation. Future research will focus on accuracy assessment and calibration of the model through actual on-site measurement and comparison
with the model output.
Chapter II.10:
535
Literature cited
Agarwal, C. et al., A review and assessment of land-use change models: Dynamics of space, time
and human choice (publication draft), U.S. Forest Service Southern and Northern Change
Program, 2000.
Anderson, M.G. and Burt, T.P., Automatic monitoring of soil moisture conditions in a hillslope spur
and hollow, J. Hydrol., 33, 27, 1977.
Anderson, M.G. and Burt, T.P., Toward more detailed field monitoring of variable source areas,
Water Resour. Res., 14, 1123, 1978.
Arnold, J.G. et al., Large area hydrologic modeling and assessment: Part I. Model development, J.
Am. Water Res. Assoc., 34, 73, 1998.
Barling, R.D., Moore, I.D., and Grayson, R.B., A quasi-dynamic wetness index for characterizing the
spatial distribution of zones of surface saturation and soil water content, Water Resour. Res.,
30, 1029, 1994.
Behera, G. et al., Interfacing remote sensing and GIS methods for a sustainable rural development,
Intl. J. Remote Sensing, 17, 3055, 1996.
Beven, K.J., The hydrological response of headwater and sideslope areas, Hydrol. Sci. Bull., 23, 419,
1978.
Beven, K.J. and Kirkby, M.J., Considerations in the development and validation of a simple physically
based, variable contributing area model of basin hydrology, 3rd Intl. Symp. Theor. Appl.
Hydrol., Fort Collins, CO, 1977.
Beven, K.J. and Kirkby, M.J., A physically based variable contributing area model of basin hydrology,
Hydrol. Sci. Bull., 24, 43, 1979.
Binford, M. and Ervin, S., A static soil moisture index approach, in C. Steinitz, Ed., Biodiversity and
Landscape Planning: Alternative Futures for the Region of Camp Pendleton, California, Harvard
Univ. Press, Cambridge, MA, 1996.
Binford, M.W., Lee, T.J., and Townsend, R.M., In prep. Sampling design for an integrated socioeconomic and ecological survey using satellite remote sensing and ordination. J. Am. Stat.
Assoc.
Binford, M.W. et al., Collaborative Research: Agroecological and Socioeconomic Systems Interactions in
Thailand, unpublished manuscript, 1998.
Binnie, C.J.A., Climate change and its potential effect on water resources in Asia, J. Water Suppl. Res.
Tech. Aqua, 46, 274, 1997.
Bockstael, N.E. and Irwin, E.G., Economics and the land use-environment link, in Folmer and
Tietenberg, Eds., Yearbook of Environmental Economics, 1999.
Bonham-Carter, G.F., Geographic Information Systems for Geoscientists: Modeling with GIS, Pergamon/Elsevier Science, New York, 1994.
Burt, T.P. and Butcher, D.P., Development of topographic indices for use in semi-distributed hillslope
runoff models, In O. Slaymaker and D. Balteanu, Eds., Geomorphology and Land Management,
Gebruder Borntraeger, Berlin, 1986.
Chiew, F.H.S. et al., Simulation of the impacts of climate-change on runoff and soil moisture in
Australian catchments, J. Hydrol., 167, 121, 1995.
Chomitz, K.M. and Gray, D.A., Roads, land use, and deforestation: A spatial model applied to Belize,
The World Bank Econ. Rev., 10, 487, 1996.
Clark, L.A. and Pregibon, D., Tree-based models, In J.M. Chambers and T.J. Hastie, Eds., Statistical
Models in S, Chapman and Hall, London, 1992.
Costanza, R. and Sklar, F.H., Articulation, accuracy and effectiveness of mathematical models: A
review of freshwater wetland applications, Ecol. Modeling, 27, 45, 1984.
Crist, E.P., Vegetation and soils information contained in transformed thematic mapper data, IGARSS 86
Symp., 1986.
Crist, E.P. and Kauth, R.J., The tasseled cap de-mystified, Photogrammetric Eng. Remote Sensing, 52,
81, 1986.
Damota, F.S. et al., Drought risks for the soybean crop in Rio-Grande-Do-Sul-State, Brazil, Pesquisa
Agropecuaria Brasileira, 27, 709, 1992.
536
Daniere, A.G., Environmental behavior in Bangkok, Thailand: A portrait of attitudes, values, and
behavior, Econ. Devel. & Cult. Change, 47, 525, 1999.
Dimyati, M. et al., An analysis of land use/cover using the combinations of Landsat and land use
map a case study in Yogyakarta, Indonesia, Intl. J. of Remote Sensing, 17, 931, 1996.
Dudal, R., Soil Related Constraints to Agricultural Development in the Tropics, IRRI, Los Banos, The
Philippines, 1980.
Dunne, T. and Black, R.G., An experimental investigation of runoff production in permeable soils,
Water Resour. Res., 6, 478, 1970.
Dunne, T. and Leopold, L.B., Water in Environmental Planning, W.H. Freeman, San Francisco, 1978.
F.A.O., Agriculture Towards 2000, F.A.O., Rome, 1981.
Felkner, J.S., A digital spatial predictive model of land-use change using economic and environmental inputs and a statistical tree classification approach: Thailand, 1970s1990s, unpublished
doctoral thesis, Harvard Univ., Cambridge, MA, 2000.
Ferguson, B.K., Introduction to Stormwater, John Wiley & Sons, New York, 1998.
Florinsky, I.V., Combined analysis of digital terrain models and remotely sensed data in landscape
investigations, Prog. in Phys. Geog., 22, 33, 1998.
Foley, J.A., Kutzbach, J.E., and Coe, M.T., Feedbacks between climate and boreal forests during the
Holocene Epoch, Nature, 371, 52, 1994.
Fu, B. and Gulinck, H., Land evaluation in an area of severe erosion the Loess Plateau of China,
Land Degradation and Rehabilitation, 5, 33, 1994.
Goodchild, M.F., Integrating GIS and remote-sensing for vegetation and analysis and modeling
methodological issues, J. Vegetation Sci., 5, 615, 1994.
Goodchild, M.F. et al., Toward spatially integrated social science, Intl. Reg. Sci. Rev., 23, 139, 2000.
Greenlee D.D., Raster and vector processing for scanned linework, Photogrammetric Eng. and Remote
Sensing, 53, 1383, October 1987.
Henderson-Sellers, A., Modeling and monitoring greenhouse warming, Trends in Ecol. and Evol., 5,
270, 1990.
Hirsch, P., Political Economy of Environment in Thailand, Journal of Contemporary Asia Publishers,
Manila, 1993.
Hunter, G.J. and Goodchild, M.F., Modeling the uncertainty of slope and aspect estimates derived
from spatial databases, Geograph. Anal., 29, 35, 1997.
Hutchinson, M.F., Calculation of hydrologically sound digital elevation models, 3rd Intl. Symp. on
Spatial Data Handling, Columbus, OH, 1988.
Hutchinson, M.F., A new procedure for gridding elevation and stream line data with automatic
removal of spurious pits, J. Hydrol., 106, 211, 1989.
Hutchinson, M.F., Development of a continent-wide DEM with applications to terrain and climate
analysis, Environmental Modeling with GIS, Goodchild, M. et al. (Eds.). Oxford Univ. Press,
New York, 1993.
Hutchinson, M.F., A locally adaptive approach to the interpolation of digital elevation models. 3rd
Intl. Conf. on Integrating GIS and Environ. Modeling, Santa Fe, NM, 1996.
Hutchinson, M.F. and Dowling, T.I., A continental hydrological assessment of a new grid-based
digital elevation model of Australia, Hydrol. Processes, 5, 45, 1991.
Jensen, J.R., Introductory Digital Image Processing, 2nd ed., Prentice-Hall, Upper Saddle River, NJ, 1996.
Jenson, S.K. and Domingue, J.O., Extracting topographic structure from digital elevation data for
geographic information system analysis, Photogrammetric Eng. Remote Sensing, 54, 1593, 1988.
Julien, P.Y., Saghafian, B., and Ogden, F.L., Raster-based hydrologic modeling of spatially-varied
surface runoff, Water Resour. Bull., 31, 523, 1995.
Kaimowitz, D. and Angelsen, A., Economic Models of Tropical Deforestation A Review, Center for
International Forestry Research, Jakarta, 1998.
Katawatin, R., Crown, P.H., and Grant, R.F., Simulation modeling of land suitability evaluation for
dry season peanut cropping based on water availability in northeast Thailand, Soil Use and
Manage., 12, 25, 1996a.
Katawatin, R., Crown, P.H., and Klita, D.L., Mapping dry season crops in Thailand using Landsat5 TM data, Canad. J. Remote Sensing, 22, 450, 1996b.
Katawatin, R. et al., Regional cropping potential in NE Thailand, Canad. J. of Soil Sci., 74, 358, 1994.
Chapter II.10:
537
Kimmage, K. and Adams, W.M., Wetland agricultural production and river basin development in
the Hadejia-Jama Are Valley, Nigeria, Geograph. J., 158, 1, 1992.
Kobayashi, H., Current approach to soil and water conservation for upland agriculture in Thailand,
Jarq-Japan Agricul. Res. Quart., 30, 43, 1996.
Kyriakidis, P.C., Shortridge, A.M., and Goodchild, M.F., Geostatistics for conflation and accuracy
assessment of digital elevation models, Intl. J. of Geograph. Inform. Sci., 13, 677, 1999.
Lambin, E.F. et al., Land-Use and Land-Cover Change (LUCC) Implementation Strategy (48), International
Geosphere-Biosphere Programme (IGBP), Stockholm, 1999.
Laslett, G.M. et al., Comparison of several spatial prediction methods for soil pH, J. of Soil Sci., 38,
325, 1987.
Lombardini, C., Deforestation in Thailand, in K.B.A.D. Pearce, Ed., The Causes of Tropical Deforestation,
the Economic and Statistical Analysis of Factors Giving Rise to the Loss of Tropical Forests, Univ.
College London Press, London, 1994.
Mark, D.M., Network models in geomorphology, Modeling in Geomorphological Syst., John Wiley & Sons,
New York, 1988.
McBratney, A.B. and Webster, R., Choosing functions for semi-variograms of soil properties and
fitting them to sampling estimates, J. Soil Sci., 37, 617, 1986.
Moore, I.D., Burch, G.J., and Mackenzie, D.H., Topographic effects on the distribution of surface soil
water and the location of ephemeral gullies, Trans. Amer. Soc. Agricul. Engineers, 31, 1098, 1988.
Moore, I.D., Grayson, R.B., and Ladson, A.R., Digital terrain modeling: A review of hydrological,
geomorphological and biological applications, Hydrol. Processes, 5, 3, 1991.
Moore, I.D. and Hutchinson, M.F., Spatial extension of hydrologic process modeling, Intl. Hydrol. and
Water Resour. Symp., Austria, 1991.
Moore, I.D., Norton, T.W., and Williams, J.E., Modeling environmental heterogeneity in forested
landscapes, J. Hydrol., 150, 717, 1993.
Myer, D.E., 1991, Interpolation and estimation with spatially located data, Chemometrics and Intelligent
Lab. Syst., 11, 209, 1991.
OLoughlin, E.M., Prediction of surface saturation zones in natural catchments by topographical
analysis, Water Resour. Res., 22, 794, 1986.
Penner, J., Atmospheric chemistry and air quality, in Meyer, W.B. and Turner, B.L., Eds., Changes in
Land Use and Landcover: A Global Perspective, Cambridge Univ. Press, Cambridge, 1994.
Rango, A. and Shalaby, A.I., Operational applications of remote sensing in hydrology: Success,
prospects and problems, Hydrol. Sci. J. (J. des Sci. Hydrologiques), 43, 947, 1998.
Rao, D.P., Sustainable development and remote sensing, Intl. Arch. of Photogrammetry and Remote
Sensing, Special Unispace Vol., 32, 1999.
Rewerts, C.C. and Engel, B.A., ANSWERS on GRASS: Integrating a watershed simulation with a
GIS, ASAE paper No. 91-2621, Amer. Soc. Agricul. Eng., St. Joseph, MI, 1991.
Richards, J.A., Remote Sensing Digital Image Analysis, 2nd ed., Springer-Verlag, Berlin, 1993.
Rindfuss, R., Soils, water, people and pixels: A study of Nang Rong, Univ. NC Chapel Hill, National
Institutes of Health (NIH), http://www.cpc.unc.edu/projects/nangrong/nangrong_
home.html, 1998.
Robbins, C. and Phipps, S.P., GIS/water resources tools for performing floodplain management
modeling analysis, AWRA Symp. on GIS and Water Resources, Ft. Lauderdale, FL, 1996.
Robinson, J. et al., The need for projections of land-use and land-cover change, in W.B. Meyer and
B.L.Turner, Eds., Changes in Land Use and Landcover: A Global Perspective, Vol. 4, Cambridge
Univ. Press, Cambridge, 1994.
Rouse, J.W. et al., Monitoring vegetation systems in the Great Plains with ERTS, 3rd ERTS Symp., 1973.
Sabins, M.J., Convergence and consistency of fuzzy c-Means/ISODATA algorithms, IEEE Trans.
Pattern Anal. Machine Intelligence, 9, 661, 1987.
Sanchez, P.A., Properties and Management of Soils in the Tropics, John Wiley & Sons, New York, 1976.
Schulze, E.D. and Mooney, H.A., Biodiversity and Ecosystem Function, Springer-Verlag, New York,
1993.
Skole, D.L., Data on global land-cover change: Acquisition, assessment, and analysis, in W.B. Meyer
and B.L. Turner, Eds., Changes in Land Use and Landcover: A Global Perspective, Cambridge
Univ. Press, Cambridge, 1994.
538
Soil Conservation Service, U.S.D.A., Soil Survey of Arecibo Area, Northern Puerto Rico (Soil Survey
Report), in cooperation with the College of Agricultural Sciences, Univ. PR, San Juan, 1975.
Srinivasan, R. and Arnold, J.G., Integration of a basin scale water quality model with GIS, Water
Resour. Bull. (accepted for publication), 1993.
Srinivasan, R. and Engel, B.A., A knowledge based approach to extract input data from GIS, ASAE
paper no. 91-7045, Amer. Soc. Agricul. Eng., St. Joseph, MI, 1991.
Stern, P.C., Young, O.R., and Druckman, D., Eds., Global Environmental Change: Understanding the
Human Dimensions, National Academy Press, Washington, DC, 1992.
Tarboton, D.G., Bras, R.L., and Rodriguez-Iturbe, I., On the extraction of channel networks from
digital elevation data, Hydrol. Processes, 5, 81, 1991.
Thai Census Dept., The Thai National Census, Bangkok, 1990.
Tou, J.T. and Gonzalez, R.C., Pattern Recognition Principles, Addison-Wesley, Reading, MA, 1977.
Townsend, R. and Lim, Y., General equilibrium models of financial systems: Theory and measurement in village economies, Rev. Econ. Dynamics, 1, 1998.
Turner, II, B.L. et al., Two types of global environmental change: Definitional and spatial-scale issues
in their human dimensions, Global Environ. Change, 1, 14, 1990.
Turner, B.L., Skole, D., Sanderson, S., Fischer, G., Fresco, L., and Leemans, R., Land-Use and LandCover Change (LUCC): Science/Research Plan (HDP Report 7), IGBP, Stockholm, 1995.
U.S. Dept. of Agricul., S.C.S., National Engineering Handbook (SCS/ENG/NEH-4), Soil Conservation
Service, Washington, DC, 1972.
U.S. Soil Conservation Service, Urban Hydrology for Small Watersheds, Technical Release 55, 2nd ed.,
Venables, W.N. and Ripley, B.D., Modern Applied Statistics with S-PLUS, 1st ed., Springer-Verlag, New
York, 1994.
Victoria, R.L. et al., Mechanisms of water recycling in the Amazon Basin: Isotropic insights, Ambio,
20, 384, 1991.
Wahba, G., Spline Models for Observational Data, Society of Independent Applied Mathematics, Philadelphia, 1990.
Wald, L., Data fusion for a better exploitation of data in environment and earth observation sciences,
in ISPRS, Ed., Unispace III ISPRS Workshop on Resource Mapping from Space, Vol. 32, ISPRS,
Vienna, 1999.
WCED, Our Common Future, World Comm. Environ. and Develop., 1987.
Winkler, L. and Ulehla, J., A comparison of the methods for computing soil-moisture under field
crops with actual soil-moisture data, Rostlinna Vyroba, 38, 885, 1992.
Woodward, F.I. et al., Natural Disturbances and Human Land Use in Dynamic Global Vegetation Models,
IGBP, Stockholm, 1996.
Chapter II.10:
539
Response
Incorporating scientific information into land-use planning
This chapter by John Felkner and Michael Binford, like that of Amir Mueller, Robert France,
and Carl Steinitz (Chapter II.15), is important in reminding us that water sensitive planning can involve concerns related to subsurface- as well as surface-water resources. Further, as Felkner and Binford state, given the rapid alterations in land-use patterns and
processes occurring in developing nations, an important need exists to create rapid
appraisal methods to be able to identify and measure repercussions while they are being
wrought instead of waiting to assess damage done afterward. Such concerns are particularly apt with respect to soil resources and agricultural planning.
This chapter provides an important lesson for water sensitive planners in clearly
instructing that comprehensive and realistic GIS analysis modeling should not be
approached as if it were a simple, completely honed tool to be blindly applied to any
situation without alteration. Instead, this example shows that in order to be effective, GIS
techniques should be tailored to the specifics of the study problem and region in question.
Felkner and Binfords soil moisture index was created in a challenging geographic context
through the innovative application of firm scientific knowledge and incorporation of local
specifics based on the best available data. This perhaps is the most useful take-away
message from this chapter namely, that the use of computers for developing spatially
explicit models should not occur at the expense of the inclusion of a healthy amount of
imagination on the part of the applicants involved in applying and testing those models.
The simple and unimaginative use of GIS analysis will generate nothing more than simple
and unimaginative results whose real-world significance may be marginal at best. We, as
water sensitive planners, should expect more, though, as this chapter demonstrates, the
way to that end will almost certainly be neither simple nor simplistic.
chapter II.11
The design of regions: a watershed

planning approach to sustainability
Daniel Williams
Abstract
The knowledge and understanding gained from the regions natural system and applied
to watershed planning creates an excellent model leading to water sustainability and urban
smart growth. The battle lines between the developers and the preservationists have been
drawn for years. The same old and tiresome conflicts will come up again if a consensus
cannot be reached between the two. The decisions clearly need to be made on an interactive
systems basis and an agreement of interdependency be declared. The water supplies,
demands, and reuses within a watershed can be accomplished with planning and design
criteria that are integrated in building and zoning codes while creating desirable regional
and neighborhood form.
The regional design projects described here are compelling examples. They connect
many disciplines working together to find a common vocabulary, one that is understood
by architects, landscape architects, planners, ecologists, economists, developers, clergy,
and neighbors. Regional watershed design will be of great interest to citizens in that it
can empower them to act, to plan for the next 100 years, and to create alternatives to the
present trends and growth patterns that have raped their land, segmented their neighborhoods, and stressed their economy.
For the last 50 years, lineal, nonsystems thinking has driven growth and development.
During this time, development has been about housing, not community and not about
the natural resources and processes replaced by the development. In these developments,
which represent an overwhelming majority of the housing stock in the country, the site
preparation permanently changed the vegetation, soil, geo-hydrological, and social patterns. Water has been rerouted from river to reservoir, mountain to desert, and region to
region for urban and agricultural consumption resulting in the drying up of groundwater aquifers, some of which have not been recharged in thousands of years. This water
is now transported for use to regions less abundant in precipitation and permanently lost
to systems such as estuaries and riparian zones instead irrigating deserts for golfing,
urban sprawl, and agricultural crops.
1-56670-562-2/02/$0.00+$1.50
541
542
No cities or towns within any watershed supply their own potable water without
pumping, diverting, or damming the regional watershed. It is not just necessary to plan
at the larger scales but essential for the preservation, protection, and increased quality of
life.
Introduction
All things arise from the principle of water.
Vitruvius
The U.S. Environmental Protection Agency has identified the need for a comprehensive
watershed approach to ecosystem management and protection (Light and Dineen, 1994).
In cities from Seattle to Miami, the impact of unplanned growth on the regions water
regime and habitat is severe, unnecessary, and expensive (Figure II.11.1). In the past, there
has been the perception of the availability of unlimited energy, water, land, and other
natural resources. This apparent abundance combined with the seemingly limitless powers
of technology was seen as sufficient to balance the recognized consequences of boundless
growth and development. When an environmental problem arose, a technological solution
was sought. The profession of planning, following this lead, became little more than
developing a patchwork of land uses, separated and segregated as color patches. Lacking
both vision and sense of place, this planning helped create the opportunity for sprawl
and placelessness.
The grand scale of degradation in recent years (Figure II.11.2) emphasizes the need to
recognize the limits of natural resources. Technological solutions often cause greater harm,
requiring even more technology to resolve them (Carter, 1974). The effects of unbridled,
poorly designed growth are unavoidable when the approach is piecemeal and nonsystemdefined. These effects have been revealed in relatively modest ways, such as water shortages, oil shortages, power outages, and higher taxes. It is, however, more profoundly
illustrated in the breakdown of whole biological systems, which threaten the extinction
of species and destruction of natural resources linked to short-term economic gain. These
are clear signs that the human process has significantly infringed on the processes of
nature. The natural production, cycling, and recycling of materials and energy is being
short-circuited. More is being taken from the environment than is being returned; the
natural capital that which sustains all life is declining (Hawken, 1995). Now, faced
with limited resources, a dwindling tax base, and the increasing costs of government
applying technology indiscriminately, we will only increase these negative attributes.
Designing regions afford the opportunity to refocus, to discover how to do more with
less, and how to successfully design within the natural system sciences. The new challenge
is to integrate urban design with natural sciences taking full advantage of the free work
of nature. Water is the common resource in this approach.
Connecting land use and water resources

. . . watershed design requires a change in the perception of what
we have thought land to be to what are the common functional,
ecological and regional connections that must be reconnected? (Williams, 1998a)
Water has the remarkable ability to be deaf to political boundaries (Figure II.11.3). This
uniqueness makes it the most effective common denominator of the elements that sustain
Chapter II.11:
The design of regions: a watershed planning approach to sustainability
543
Figure II.11.1 One hundred years of draining wetlands has created cheap land for sprawl and
permanently reduced the water storages.
life. Where water managers and planners typically get into trouble is when the lines of
water supply and demand do not reflect the boundaries of the users. Virtually all municipalities get their potable water supply from a watershed outside their own development
boundaries. New York City, as an example, has been purchasing land in the Catskills
watershed for decades. This action that protected its potable watershed has saved in the
order of $5 to $8 billion for additional treatment for their potable water supply (Robert,
Yaro, personal communication, 1996).
544
Figure II.11.2 The completed development of the area from the Everglades to the Atlantic will be
impervious surface by 2025, virtually eliminating recharge in the entire region.
Figure II.11.3 A hydrographic topo of 29-year rainfall data showing amorphic rainfall patterns.
Note the greater averages over the urban impervious area along the coastal zone.
Hawken states that . . . in any endeavor, good design resides in two principles. First,
it changes the least number of elements to achieve the greatest result. Second it removes
stress from a system rather than adding it (Hawken, 1995). In that urban settlements
have overwhelmingly disregarded these most basic concepts, the initial task is to reconnect
the natural systems principles to the natural systems regional and local patterns.
Chapter II.11:
545
Regional watershed planning takes the approach of reconnecting elements lost in

previous plans and healing the system while creating more livable communities. Designing local and regional watershed connections that reinforce the interaction of land use and
water is a first step in the design of regions. The objectives are:
To provide a comprehensive understanding of the relationships between land use
and regional water management objectives while building consensus among local,
state, and federal interests as to the actions necessary to achieve regional water
management objectives
Develop a regional design vision that illustrates the reconnection to natural resources and general systems ecology (Williams, 1992)
In the relationship of water quantity and quality, the relationship of water to land use
is the single relationship over which humans have some control. Because it is the land
that receives, stores, and distributes water, it is the land planning, particularly the master
planning, that decides the fate of the water crop. Only land, the watershed, can make this
claim.
If the water quantity is insufficient but a renewable supply exists, then the receiving
land area must be increased. If the quality is insufficient, then the land must store and
clean the water through biological uptake. If there is insufficient water for use and the
supply is insufficient, the water must be recycled to make up for the lack of supply. If
sustainability is desired all of the above must be accomplished. In other words, water
sustainability requires regional self-sufficiency the water that lands on the area is the
working budget for all uses additional consumption of water over that allotted amount
is deficit spending. This deficit spending reduces the net total capacity of the system and,
in that it is accommodating additional growth, the deficit can only get worse as the demand
grows while the recharge area is reduced. In the budgeting cycle, first the natural system
and hydrological integrity should be satisfied, then the combined needs of the agricultural
and urban systems. Without regional systems solutions, the loss to all systems will far
exceed the gain to one.
Case study: Florida

Water, not growth management controls, will be the ultimate factor
in determining how big Florida can grow. Water management is
part of the overall effort in Florida to get a handle on where this
state is, and where were headed to, and to get out ahead of our
tumultuous growth, once and for all, Lt. Gov. Buddy MacKay said.
It gives you an idea that whatever it is thats going to limit growth
in Florida, its not going to be growth management . . . the availability of that resource is probably going to be the ultimate limiting
factor. (The Ledger, 1994)
In Florida, there is a large section of uniquely beautiful yet dying piece of land known as
the Everglades (Figure II.11.4). The Native Americans called this area a river of grass,
which many years ago became the title of Marjory Stoneman Douglass book a work
that, until now, saved it from development.
The issues of the Florida Everglades and the drainage associated with the urban and
agriculture use in its surrounding watershed have been well studied (Carter, 1974). This
type of drainage has been done for the entire county and most of the state.
546
Figure II.11.4 The Everglades, a 5000-year gentle flow of water falling 20 ft in 120 miles.
Figure II.11.5 The Miami River, once flowing from the Everglades basin, now a drainage ditch
delivering stormwater from thousands of miles of drainage canals.
Figure II.11.5 is of the Miami River in South Florida in 1996; remarkably not a single
drawing was done to guide this citys future. More time is spent deciding on the quantity
of parking spaces than considering the future of the city for future generations. The issue
then is not if a place will change, but when it changes, can it be in a way that the citizens
desire. Planning, in a sense, is taking the necessary steps to assure one gets what is desired,
rather than what happens by default.
Water is a vital element of Floridas natural and human-made environments. Fresh
water underlies the state and periodically inundates much of its interior with an expansive
freshwater marsh, the Everglades, which covers much of the states southern interior. The
surface and underground freshwater reservoirs of the region are essential to the integrity
of near-shore ecosystems (Figure II.11.6). These waters flow into coastal estuaries, mixing
Chapter II.11:
547
Figure II.11.6 Illustration showing relationship between rainfall and surface- and groundwater recharge. Remove the recharge and saltwater intrusion occurs.
with the vast saltwater seas that surround the states 1,200 miles of coastline. At the
interface between salt and fresh water, sensitive and finely balanced estuaries provide
nurseries for abundant marine life and, consequently, commercial fisheries. As the foundation of Floridas natural environment and cornerstone of the states recreation and
tourism industries, water is essential to the states sustained health and prosperity. Due
to its seeming abundance, water, as Plato said, . . . gets the least amount of care.
Historically, wetlands were considered not only worthless, but also a threat to human
survival (Carter, 1974). Estuaries (Figure II.11.7) were mosquito-infested mangrove
swamps and in dire need of improvement. Thus, the last 100 years of Floridas history
have been characterized by attempts to tame nature, marked by endeavors to adapt the
natural landscape solely to human uses. A result of this is a drainage and flood control
system in South Florida that is inadequate for flood control, is inadequate for water supply,
and provides insufficient water quality protection. This is occurring at a time when
demand for flood protection, water supply, and water quality standards is increasing
exponentially while the water losses exponentially decline due to drainage. This demand
goes hand in hand with the pressures exerted upon the present system by the rapid and
unchecked outward expansion of urban settlements into the water recharge areas.
Opportunity
Following Hurricane Andrew on August 24, 1992, the opportunity arose to apply a watershed approach as a critical element of the rebuilding effort in South Dade County. This
adversity to opportunity approach seeks to successfully integrate land-use development
and natural resource protection by preserving, protecting, and, when necessary, reestablishing the connections on a regional scale. Thus, watershed planning would become the
conceptual foundation for the local, state, and federal planning initiatives on smart growth,
all connected to the need for potable water (Williams, 1993).
A watershed plan represents an initial step in developing a comprehensive development plan for the restoration and protection of Everglades National Park and Biscayne
Bay while future growth occurs. It will also begin the educational process, which is
necessary to build consensus within the local communities. This consensus provides the
power to the local agenda for developing and implementing watershed management
strategies.
548
Figure II.11.7 The South Dade regional system illustrating coastal and estuary systems connecting
to make a vital ecosystem.
The objectives of the watershed plan are:

1. To provide a comprehensive understanding of the relationships between land use
and regional water management objectives
2. To build consensus among local, state, and federal interests as to the actions
necessary to achieve regional water management objectives in South Dade County
3. To select a hydrologic basin within the South Dade County watershed for future
study and the development of future demonstration projects (Williams, 1994).
Watershed planning: A process

A first step is to find out where the community wants to be in 50 to 100 years. In Lewis
Carrolls Alice in Wonderland, Alice asked the Cheshire Cat, Would you please tell me
which way I ought to go from here? The cat replies, That depends a good deal on where
you want to get to. The second step is to find out how the system functions and what
elements are in the regional resources inventory.
Chapter II.11:
549
Figure II.11.8a Aerial photo showing development around low-lying wetlands and flood control
canals.
An additional million people will pave over an additional 21 square miles of the
remaining pervious area. This is 21 square miles (Figure II.11.4) of present recharge area
lost that once received, stored, and distributed water to sustain the regional water budget.
Use without recharge is not sustainable. More important, the water budget is truly a
function of a continuous supply. Water budgets that include the huge storages of the
aquifer are not budgets that can be sustained. In an economic analogy, this is similar to
living off the principal rather than the interest. With this agricultural recharge comes the
associated runoff and pollutants, but the urban system may be many times worse. Agriculture, while a major consumer of water, is also an important water recharge area. These
open agricultural land areas, which are pervious surfaces, allow percolation of resident
rainfall to add to and maintain the regional storage. It is estimated that up to 7 times more
runoff is associated with urban land use. Degradation of the water in the runoff is equivalent to secondarily treated effluent (Light and Dineen, 1994).
The plan
One of the questions citizens asked over and over is, How will this affect my daily
experience? What will sustainability look like? These greenway areas (Figure II.11.8a and
b, showing before and after) allow for the storing and release of water to the Biscayne
Bay, mimicking the historic natural system. These areas will also increase the watersheds
ability to store water, while providing additional flood protection. Because this water is
being stored, it can be cleaned up through biological uptake prior to its conveyance to the
canals, to recharge areas, and then on to estuaries.
Shown within this vision (Figure II.11.9) are areas composed of three land-use types:
urban, agricultural, and reclaimed wetland storages. In this concept, all of the land uses
are networked through the storage, distribution, and clean-up of water. This creates a
networking of water use and reuse. The wetlands represent local treatment plants while
the canal littoral areas and larger surface storage areas would be sized as localized storage
for clean-up as the pollutant loading would dictate.
Designs of the greenways can take many forms and be many sizes. The size would
depend on the pollutant loading of the land use within the watershed that drains to the
550
Figure II.11.8b The wetlands have been reclaimed, improving flood protection, adding open space,
and increasing property value.
Figure II.11.9 The Watershed Interactive Network Plan the WIN plan, illustrating systems planning and design that reconnects all components of land use into a working community of urban,
agricultural, and natural system functions.
canal. These greenways also act as habitat for various species, including humans. Creating
pathways rich in diversity and function including lineal parks, greenways, bike paths
and local flood protection (Figure II.11.10) are funded by stormwater management funding.
At the coastal zone, canals would no longer flow directly into the Biscayne Bay but
through an archipelago of sorts. By diverting the flow and creating storages upstream,
the water quantity and quality will be enhanced. Simultaneously, considerably more lineal
area is created for the food-chain nursery ground.
This new edge condition in the urban segment where the canal and littoral zone meet
would create identity and image for the community while increasing the flood protection
and neighborhood open-space amenities. The combined water storage areas and open
spaces not only help define better neighborhoods but increase property values at the same
time (Figure II.11.11a, b, and c).
Chapter II.11:
551
Figure II.11.10 A detail of the WIN plan illustrating linkages between urban design, hydric open
space, gray and black water reclamation, and transportation.
EXISTING STREET SECTION
PROPOSED STREET SECTION
PLAN VIEW
Figure II.11.11a Illustration of conversion of a typical road system to a WIN plan street recharges
the aquifer, reduces heat island impacts, creates desired microclimate, and increases property values.
The agricultural areas would again be defined by the natural water regime in the area.
Agriculture would employ reverse-flooding irrigation, which will simultaneously waste
less water and control nematodes and other pests while rebuilding soil. The flows of water
to Biscayne National Park would filter through these reclaimed wetlands and be better
552
Figure II.11.11b A hydric boulevard stormwater is gravity-fed to the center green area for storage,
clean-up, and aquifer recharge. Microclimate, quality of life, and property value all increased at a
cost lower than the typical street. (See Figure II.11.11c.)
Figure II.11.11c The typical, undesirable street same right-of-way as Figure II.11.11b but environmentally costly.
timed to the bay. This regional watershed plan illustrates all the interactive, networked,
and sustainable ideas incorporated into a single vision. It is a vision that the citizens,
scientists, and planners through consensus consider affordable, practical, and desirable.
Conclusion
The future . . . its not working! Even developing new and wonderful walkable communities will not have any significant change in the disastrous path being taken, nor will
Chapter II.11:
553
energy-saving homes, or commercial conservation, or drip irrigation, or hydrogen fuel

cells, or paying-our-way in fossil fuel costs none of it is enough. It all comes from the
same trend, the same flawed conceptual base individuals rule the land has no voice.
Typically, development is thought of as roads and houses neatly or not so neatly placed
upon the land. In most cases this development obscures the lands beauty and its deeper
value. The powerful ethic of house and home, that basic need to place a roof over ones
head and provide places to work and have community, has disregarded the value of the
regional systems connectivity, its place, and it is costing huge amounts.
The relationship between land use and water is the essential tool needed to design
and plan for the symbiotic relationships presented here. The missing links are the next
steps, which require the planning professions proactiveness on the regional scale. The
design of regions creates the drawings and makes the design decisions necessary to protect
the region so that it can work as an organism and applies the understanding of these
relationships to short- and long-term planning.
The watershed design approach develops many successful connections:
Protection of the natural systems functioning within regional and urban greenways
increases the value of the adjacent properties while providing for regional recharge.
Storing water within these hydric greenways will increase water supplies for future
use in an economic way while increasing the potable water supply and receiving
the bonus of community open space.
The preservation and protection of parks and conservation areas integrated within
the urban and regional pattern provide critical livability, health, and economy to
the entire region.
Planning and zoning must integrate water management with land-use decisions
and, while doing so, create strong growth management boundaries informed by
conservation science and design.
Ecosystem management and watershed planning simultaneously address the complex issues of urban quality of life and preservation of natural systems.
Developing a common vision, illustrating it, and establishing the incremental steps
for its implementation are essential to urban and regional sustainability. Watershed planning provides a process that informs the community-based vision while providing for an
economic and environmentally sustainable future.
Live locally, think globally, act regionally!
Literature cited
Blake, N.M., Land into Water Water into Land: A History of Water Management in Florida, Univ.
Presses of Florida, Tallahassee, 1980.
Carter, L.J., The Florida Experience: Land and Water Policy in a Growth State, John Hopkins Univ. Press,
Baltimore, 1974.
Douglas, M.S., The Everglades: River of Grass, 1996.
Hawken, P., The Ecology of Commerce: A Declaration of Sustainability, Island Press, Washington, DC,
1995.
The Ledger, Water limits may halt growth, Lakeland, FL, Sept. 26, 1994.
Light, S.S. and Dineen, J.W., Water control in the Everglades: A historical perspective, in Davis, S.M.,
and Ogden, J.C., Eds., Everglades: The System and Its Restoration, St. Lucie Press, Delray Beach,
FL, 1994.
Miami Herald, Andrew: Recovery by the numbers, August 24, 1994.
Williams, D., The Watershed Interactive Network: Towards a Sustainable Watershed Design Methodology,
Univ. Miami, Center for Urban and Community Design, Miami, 1992.
554
Williams, D., Regional Studies Introduction, The New South Dade Planning Charrette: From Adversity
to Opportunity, Univ. Miami Press, Miami, 1993.
Williams, D., South Dade Watershed Project Workshop report, Center for Urban and Community
Design, Miami, 1994.
Williams, D., South Dade Watershed Project Workshop report, Center for Urban and Community
Design, Miami, 1995.
Williams, D., Sustainability panel, APA: FL Conf., Tampa, 1998a.
Williams, D., The Eastward Ho vision, FL Dept. of Comm. Affairs, proposal, Univ. FL, Miami
Education and Research Center, Miami, 1998b.
Chapter II.11:
555
Response
Expanding planning vision in space and time
Sometimes even watersheds are not large enough for truly effective water sensitive planning. Due to the unnatural rerouting of water across watershed boundaries, from locations
where there is less infiltration to locations where the water is used in toilets and then
discharged into pipes and carried to a third location, Daniel Williams believes that planning at larger spatial scales is essential for human and environmental sustainability. Also
important is the need to broaden the time frame of planning given that aquifer recharge
may take place on the scale of centuries, not years or even decades. With these two
expanded visions, Williams asks the provocative question as to whether we can ever know
for sure if any of our development projects are truly sustainable.
Past planning has been predicated, Williams instructs, on the misconception of unlimited supply and an unhealthy faith in techno-fix solutions as the means to solve all
problems; that is, almost all large communities now draw water from outside their watersheds, and if the amount decreases, well, one can always build a longer intake pipe, cant
one? Such a planning approach has helped to create a lack of sense of place and has
promoted sprawl.
Modern water sensitive planning needs a much greater comprehensive understanding
of the relationships (both physical and social) between landscapes and water assets and
uses. Such understanding, Williams believes, can be found only at the regional scale.
Implicit in this chapter is a lesson of great importance that underlies this entire book;
namely, that water is not a renewable resource, and until it is stopped being treated as
such, we will continue to suffer from various, so-called water crises, both in North America
and abroad. Toward this end, Williams makes a plea for water sensitive planning to shift
its focus from housing and community development to natural resources and processes.
For, in the end, it is the water/landscape that should be regarded as the ultimate client,
not the humans who wish to reside there.
chapter II.12
GIS watershed mapping: Developing

and implementing a watershed natural
resources inventory (New Hampshire)
Jeffrey A. Schloss
Abstract
Watershed planning, management, and protection efforts have generally been driven by
more reactive approaches in which actions and concerns stem from perceived or actual
impairment and existing threats. Decisions are often made using a very limited set of data
that may or may not be directly related to the specific watershed of concern. Typically,
these decisions are directed with a bias toward land capability analysis. A more proactive
approach requires a sufficient watershed natural resources inventory that would allow for
the documentation and analysis of landscape activities and features of importance along
with information on the extent and condition of in-water and riparian resources and their
associated habitats. This then would allow for informed preemptive planning and protection. It is also the only way to approach community stewardship and planning for the
more pristine watersheds where limited degradation has occurred and the emphasis is to
maintain it that way.
The advent and refinement of geographic information systems (GIS) and related
spatial technologies, such as remote sensing and global positioning systems (GPS), have
brought forth the capability to conduct watershed inventorying and analyses that can
allow watershed stakeholders to become better informed on the current state of the
watershed and predict the future consequences of change.
This chapter provides a basic overview of these technologies and introduces the
concept of a Watershed Natural Resources Inventory (WNRI) as a useful watershed assessment tool for proactive community planning, minimum impact site design, and the overall
management and stewardship of watershed resources. This procedure can also be utilized
for meeting local decision-making needs, for developing protection strategies for watershed lands and source waters, and for the design and implementation of monitoring
programs. Case studies incorporating WNRIs for a large multijurisdictional watershed,
for a small lake watershed involving three communities, and for a series of small wetland
watersheds in a single community setting are discussed. The chapter concludes with an
outline of the steps involved in the implementation of a critical resources analysis, which
represents the optimum combination of top-down (capability/limitations) and bottomup (resource co-occurrence) GIS-aided watershed assessment.
1-56670-562-2/02/$0.00+$1.50
557
558
Introduction
The surface waters of New Hampshire represent a valuable resource contributing to the
states economic base through recreation, tourism, and real estate revenues. Some lakes
and rivers serve as current or potential water supplies. For most residents (as indicated
by boating and fishing registrations and surveys) our waters help to ensure a high quality
of life. New Hampshire currently leads all of the New England states in the rate of new
development and redevelopment (according to 2000 U.S. Census data recently posted on
the Web). The long-term consequences of the resulting pressure and demands on the states
precious water resources remain unknown. Of particular concern is the response of our
waters to increasing nonpoint source pollutant loadings due to watershed development
and land-use activities.
In New Hampshire, as in most New England states, we rely on local decision makers
to act as stewards for our community and watershed resources. These local decision
makers are primarily elected or appointed, or they may be volunteers. They often do not
have, or lack access to, the proper resource information and education on which to base
their decisions. Yet, these local officials are responsible for evaluating development, subdivision, residential, industrial, commercial, and recreational projects, all of which can
have significant impact on a communitys natural resources. They are also responsible for
the implementation of land-use and zoning regulations and the development of the communitys master plan that affects the future land use of their community. Thus, while local
decision making is the key to watershed-based community resource protection, the information and education required to make informed decisions have often been lacking or
severely limited. This lack of information has resulted in a more reactive approach to
watershed resources protection where problems are addressed when they occur or are
perceived to occur or when resources are threatened. To correct this situation, it is imperative that these decision makers are provided with the specific information of the location
and status of critical watershed resources that can then allow for the optimum proactive
protection through comprehensive planning and management.
The natural resources inventory approach was originally developed as a process for
communities to document and locate important resources for developing resource protection plans and for updating town master plans (Auger and McIntyre, 1992; updated
recently incorporating basic GIS analysis and mapping approaches by Stone, 2001; see
also the following section). Our work, initiated concurrently, has specifically focused on
using a water resources sensitive approach that expanded upon this original resources
inventory concept and focused on the GIS analyses that would be useful to use at the
watershed and subwatershed levels (Schloss and Rubin, 1992; Schloss and Mitchell,
1996).
We have developed the Watershed Natural Resources Inventory (WNRI) as a useful
watershed/waterside assessment tool for proactive community planning, minimum
impact site design, and the overall management and stewardship of watershed resources.
This procedure can also be utilized for meeting local decision-making needs, developing
protection strategies for watershed lands and source waters, and the design and implementation of monitoring programs.
Recent developments in computer speed and capacity, access to the Internet, and the
development of more user-friendly technical software have made mapping and spatial
analysis technologies more available. These technologies can facilitate the WNRI process
and allow watershed stakeholders to become better informed on the current state of their
watersheds and predict the future consequences of change.
Chapter II.12:
GIS watershed mapping
559
MAPS and PHOTOS
MONITORING and
(hard copy)
RELATED DATA
DIGITAL IMAGING
INPUT
Map
Image
Database
Digitizing
Processing
Management
Spatial and
Spatial
GIS
Analysis
Attribute Database
Statistical
Analysis
OUTPUT
(onscreen, digital or hardcopy)
MAPS, VISUALIZATIONS, DATA and STATISTICS
Figure II.12.1 Schematic diagram of a geographic information system.
An introduction to GIS and related spatial technologies

GIS
A geographic information system (GIS) can be defined as the hardware, software, and data
that assimilate descriptive information with location referencing to allow for spatial analysis and visualization. Figure II.12.1 displays a schematic diagram of a GIS system. More
simply, GIS can be thought of as a system that allows for digital mapping along with the
capability to perform analyses on the elements of the map as well as related descriptive
information. These graphic map elements include polygons (watershed extent, lakes,
wetlands, land parcels, political boundaries, etc.), lines (roads, trails, pipelines, etc.), and
points (wells, outfalls, sampling stations, etc.). The related data can be further divided
into two categories: descriptive data that contain attribute information about the element
(i.e., lake name, road name, station number) and spatially linked reference data. The
reference data contain characteristics of the map elements (town population, wetland
vegetation, soil characteristic, well yield, station water quality results, etc.). It is the ability
to manipulate both the map elements and reference data that allows for some powerful
analysis to take place (see examples in Chapters II.10, II.14, and II.15).
In addition to the manipulation of map elements and their associated data, a GIS also
allows for spatial manipulation and analysis. These are derived through topology analysis
and include the ability to derive, detect, or perform the following spatial queries:
Connectivity. The intersection or overlap of selected elements allows for the location
of points such as where a road intersects a stream. This can be used to find culvert
locations throughout the watershed or to help select water sampling stations.
Proximity. The closeness of one object to another allows for selecting elements by their
locations relative to other elements. Example uses of this analysis might allow for
the detection of salt storage areas within 500 ft of a water body or hazardous waste
storage within a quarter mile from a school building.
Adjacency. The bordering of one element to another allows for the detection of edges
and transitions. This operation can be used to find land parcels that border a
wetland or to link subwatersheds with the receiving water into which they empty.
560

Table II.12.1 GIS Data Commonly Available
Base Map (Background) Data
USGS topographic maps and components (i.e., 1:24K scale)
Orthophoto (corrected and digitized aerial photography)
Infrastructure
Roads/rail/trails
Water/sewer/gas/oil
Transmission lines
Geology
Surficial
Bedrock
Hydrography
Lakes, ponds, rivers
Streams
Soils (w/attributes)
County surveys
Regional surveys
Wetlands
National Wetlands Inventory (U.S. Fish and Wildlife)
USGS Digital Line Graph data (DLG) (from topo maps)
Derived from soils
Derived from remote sensing
Groundwater resources
Aquifers, wells
Transmissivity/yield/quality
Demographic (Census)
Land cover/land use
Satellite/aerial photo derived land cover
Regional/local planning agency land-use/zoning
Digital elevation models
Watershed delineations
USGS Hydrologic Unit (HU) delineations
Source: USDA Natural Resources Conservation Service.
Buffering allows for the derivation of a boundary of a set width around a point, line,
or polygon element. This operation can provide additional map elements representing setbacks, zones, and regions.
Neighborhood and network analysis allows for the derivation of a linear path through
an analysis that relates a specific area, or point, to those around it. Slope and aspect
are common examples of neighborhood operations, whereas flow and transport
corridor determinations are examples of network analysis that expands upon those
operations.
It is this range of analyses described above as well as the ability to work with and
process a series of map layers (overlay analysis) that separate GIS from other spatial
databases such as computer-aided design (CAD) and computer-aided mapping (CAM)
systems. No other system allows us to combine relational and spatial queries that can
create a new layer (map) from two or more existing layers (Congalton and Green, 1992).
Table II.12.1 lists the typical GIS data that are more readily available. Most GIS maps
utilize existing base map components derived from U.S. Geological Survey (USGS) topographic map quadrangles. Besides the availability of these digitally scanned maps (commonly referred to as digital line graphs, or DLGs) certain elements of these maps (roads,
Chapter II.12:
561
water bodies, wetlands, etc.) may also be available in separate GIS overlays (often referred
to as themes or coverages). Soil maps can be quite useful at the resolution they are generally
mapped at (usually at a 1:24,000 scale with 2- to 5-acre soil units delineated) by the USDA
Natural Resources Conservation Service (NRCS; formerly Soil Conservation Service).
Information contained in their associated soil characteristics data can include how wet
the soils are, slope, depth to groundwater, erodibility, and suitability for septic installation
or homesite development. The U.S. Fish and Wildlife Service has cataloged and digitized
much of the countrys wetlands at a similar scale from aerial photographic interpretation.
Other data may have been generated on a statewide basis from state agencies or cooperative groups.
Although it would appear that the USGS, NRCS, the Census Bureau, U.S. Fish and
Wildlife, or the U.S. Environmental Protection Agency would be your first stops when
gathering available GIS data, it is more important to find out where (or if) you have a
regional or statewide GIS data repository. This can be a critical time and resource saver,
for often these centers offer data at low or no cost. In addition, they may already have
processed the data from sources listed above to be compatible with the map projection
and coordinate system commonly employed for locally produced GIS maps and data. This
is most important, as you need every data layer to register correctly to perform any
accurate mapping or analysis. State GIS repositories may be located at a university, state
library, or agency (state or regional planning, resource management, resource protection,
or transportation agencies are some of the more common locations). In some cases the
data may be distributed among a group of agencies with no centralized clearinghouse.
In addition to the government agencies listed previously, private companies also
provide GIS data for a range of fees. Due to the rapid growth in both the power and
storage capacities of desktop computers and with multiple paths available for the conversion of CAD/CAM data to GIS, it is not uncommon to find larger towns and local
municipalities with their own GIS systems generating their own data on infrastructure
and local resources. In most cases, state or regional planning agencies act to support these
efforts and/or act as GIS service providers for their member towns.
When working with GIS data it is important to understand the limitations of those
data. Coverages are derived from data with varying resolution due to the original spatial
scale employed. A typical USGS 7.5-min topographic quadrangle (1:24,000 scale) is usually
spatially accurate to about 50 ft or so. Temporally, the data may reflect aerial photo
interpretation that can be over a decade or more old. This may mean that the roads or the
forested areas shown are no longer located there. The least accurate layer used limits the
accuracy of any GIS analysis. Typical data sets provide the resolution that allows for
watershed and subwatershed analysis, region or town-wide analysis, and possibly large
zones within these areas. They do not, however, allow for anything close to site-specific
determinations. This is a very important and sometimes abused limitation as GIS systems
allow for production of products that exceed the scale of resolution. Although no direct
conclusions should be based on GIS data for a single site or house lot, if GIS data suggest
a certain resource or condition exists at a site, you may want to be sure to visit that site
to check out the possibility of the occurrence. This may be especially critical if you only
have the resources or time to visit a limited number of sites. Thus, the GIS data may be
employed as a screening tool for situations beyond the datas spatial resolution.
Remote sensing
A second spatial technology, remote sensing, is the use of airborne instrumentation (airplane
or satellite) to collect information by photography or digital sensors that can later be
interpreted to create (after processing, correction, and geo-referencing) additional spatial
562
ORTHOPHOTO
SPOT
(a)
LANDSAT TM
(b)
(c)
Figure II.12.2 Comparisons of resolution from (a) digital orthophotoquads from aerial photography
(1-m pixel size), (b) SPOT satellite panchromatic imagery (10-m pixels), and (c) LANDSAT Thematic
Mapper satellite imagery (30-m pixels).
data layers. Land-cover GIS layers are often derived from satellite remote sensing. These
coverages estimate the generalized land cover (forest, wetlands, water, crops, urban, etc.)
over an area. Older classifications relied on LANDSAT Thematic Mapper (TM) imagery,
which offered 30-m resolution (each picture element or pixel approximately represented
a 30 30-m area on the ground). More recent, SPOT satellite imagery, which offers 20-m
color and 10-m panchromatic (black and white) resolution, has been made available, and
currently, the newest satellites, such as IKONOS, now offer 1- to 5-m resolution, which in
many cases is the equivalent of standard aerial photography. Figure II.12.2 allows for the
comparison of data from this range of resolutions. In all cases, care must be taken when
working with these data. Atmospheric conditions and clouds can limit areas interpretable
in the coverage. Forest canopy cover can hide low-density development, wetlands, and
streams. Steep-facing slopes are sometimes misinterpreted as water or wetland. Most
notably, these scenes represent a moment in time; the date of the available imagery may
not reflect current conditions, and different classification procedures may have been used
on different imagery sets. Thus, care must be taken when comparing the available data.
This is not to imply that the data are not useful. Figure II.12.3 displays the use of
Landsat TM derived land cover for the relatively small Bow Lake, NH, watershed. A GIS
query was run using the satellite-derived, generalized land-cover data for those features
known to be associated with nonpoint source pollution (active agriculture and highly
developed areas) that were 50 ft or closer to a shoreline area, wetland, or lake tributary.
This then allows for the informed selection of water quality sampling sites that can bracket
these presumed hot spots.
For watershed-specific work, other, less commonly available, spatial data coverages
may be important. Aquatic and riparian habitat areas, sampling stations, and their related
water quality data are some examples of information that may be available, digitally or
not, that can assist in inventory and analysis. Resource agencies, universities, recreational,
wildlife, gardening and sporting clubs, angling and hunting guides, and other members
of the watershed community may be important resources that can assist in collecting and
documenting habitat and species occurrence data. Protection agencies, universities, volunteer monitoring groups, and watershed or landowner associations may be able to assist
in providing water quality and related data. In urban and suburban areas, storm drain
system maps can be very important in relating land-use activities to water quality impacts.
GPS
To create additional spatial data or to ground-truth remotely sensed and GIS data, global
positioning system (GPS) technology is used. GPS is a highly accurate locationing method
Chapter II.12:
563
KEY
Land Cover Types
urban
agriculture
water
wetlands
disturbed/
clear cut
Streams
Roads
Figure II.12.3 The use of Landsat Thematic Mapper satellite-derived land cover for the Bow Lake,
NH, watershed. Features known to be associated with nonpoint source pollution (active agriculture,
cleared land, and highly developed areas) 50 ft or closer to a shoreline area, wetland, or lake tributary
are identified by shading. (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
that utilizes a special computerized radio device. This device receives satellite data from the
U.S. Department of Defenses constellation of 24 very high altitude satellites orbiting the
earth, which continuously transmit their positions and associated information. At any given
time and place, if at least four satellites are in view, a GPS receiver can pick up the signals
and use them to determine a position on the earths surface. Initially, the Department of
Defense degraded the public-use GPS satellite signals, on occasion, out of security concerns.
This selective availability occurrence has recently been halted indefinitely. As a result, even
recreational GPS units can now provide a relatively accurate location fix.
As with remote sensing above, atmospheric conditions can greatly affect the accuracy
of GPS. Also, the arrangement of the satellites being accessed in relation to your position
on the ground will affect results (some GPS receivers come with software that will allow
you to plan your time of readings when satellite positions are optimal). With a range of
features available to account for potential errors, the accuracy of GPS receivers can vary
widely. Recreational units are generally accurate to 10 to 30 m. Mapping quality GPS units
commonly employ differential correction in which, after collection, data are compared to
data collected at a nearby base station location at the exact same time to allow for precise
adjustments and higher (sub-meter) accuracy. Additional signal processing options including carrier phase analysis can further improve accuracy. Some mapping GPS receivers and
marine GPS navigational units are designed to receive and process in real time the signal
from a nearby base station, Coast Guard beacon, or commercial satellites. The highest
resolution can be found on survey quality differential GPS receivers and base station
combinations that can provide centimeter or greater accuracies in real time.
The remainder of this discussion highlights examples of how these spatial technologies
have been employed to conduct and implement watershed natural resource inventories.
They have been selected because they represent a variety of project types, occur through
a continuum of small to large watershed projects, and cover a range of community
concerns and resulting initiatives.
564
GIS watershed inventory and analysis of the Squam Lakes watershed

Background
In this first example, we explore a large multijurisdictional watershed where we had
sufficient resources to obtain and/or digitize a wide variety of GIS data layers to explore
what data were most critical and how GIS analysis could provide the needed information
for watershed protection and planning. As part of a model watershed study under the
direction of the NH Office of State Planning, a multiagency task force worked to create a
GIS based resource inventory of the Squam Lakes watershed (Scott et al., 1991). The states
GIS repository, GRANIT (for Geographically Referenced Analysis and Information Transfer), is housed at the University of NH (UNH) but linked to state agencies and regional
planning commissions. Data layers used in this GIS study included bedrock geology,
hydrology (streams, wetlands, lakes, ponds, and aquifers), soils, elevation, land-use zoning, land cover (from aerial photographs and satellite images), and wildlife habitats. This
was in addition to a base map of roads and political boundaries. Also included was over
10 years of water quality data, collected weekly during the ice-free season throughout the
lake, by volunteer monitors of the Squam Lakes Association under the direction of the
NH Lakes Lay Monitoring Program coordinated out of UNH.
The Squam Lakes watershed comprises the land surrounding Squam Lake, Little
Squam Lake, and three smaller tributary lakes located in what is known as the Lakes
region in central New Hampshire. The watershed is truly multijurisdictional because it
lies in three counties with six towns bordering the lake and an additional two towns located
in upland areas of the watershed. The watershed covers 42,418 acres, of which 7847 compass
water area.
Conventional land capability analysis

In the initial study, a conventional top-down GIS analysis of land capability was undertaken to display all of the developable area remaining in the watershed (Figure II.12.4).
The process involved subtracting out all areas in the watershed that were limited to
further development. This included areas of water, wetlands, and the required buffers
around them as per town and state regulations (easily facilitated through GIS proximity,
buffering, and neighborhood analysis), areas restricted due to excessive slopes or poor
soils, protected lands, and already developed areas. The GIS was also used to analyze
information on zoning specific to each town (i.e., land area required for each house lot)
and provide a full buildout scenario that could estimate the maximum number of new
houses allowed by town and by subwatershed, and it even estimated the resulting increase
in population under this development scenario.
The build-out analysis made quite an impressive series of maps. However, it only
took a simple pie chart to translate the bottom line: for the Squam Lakes watershed, about
12% of the watershed was currently developed or protected; about 52% was constrained
or restricted to development; and almost 37% of the watershed was still left to be developed (Figure II.12.5). Although, as a whole, the lake displayed excellent water quality and
was relatively pristine in nature, areas with less desirable water quality conditions
existed within the lake. Thus, the problem was defined: areas of the lake were already
showing signs of water quality degradation, yet current laws and regulations would
allow development within the watershed to expand over three times the area of what
was already developed. What was still needed was a method to locate critical lake areas
and produce additional GIS products to educate and support decision makers and their
communities.
Chapter II.12:
565
Water and Wetlands/Buffers

Remaining Developable
Watershed Lands
Subtracted
from Watershed
Town Boundaries
Zoning (Density Allowed)
Setback Requirements
Developed and Protected Lands
Restricted Slopes and Soils

Watershed Build-out
(Density by Town Zoning)
Figure II.12.4 Land capability analysis of the Squam Lake watershed facilitated through GIS analysis.
Squam Lakes Watershed
Land Capability Build-out Analysis
Developable
37%
Developed or
Protected
12%
Water
18%
Constrained
33%
Figure II.12.5 Pie diagram of results from the Squam Lakes watershed land capability analysis.
Geographic display of data and GIS visualization

The project recommendations included suggestions on taking into account land-based
resources such as productive forest and agricultural soils, and aquifer recharge areas. Still
needed was a method to locate critical lake areas and produce additional GIS products to
assist decision makers concerned with water quality and watershed resource protection.
With that in mind, it was time to go beyond the traditional GIS approaches and push
the envelope by exploring GIS data display and visualization. Displaying the water
566
(a)
(b)
Figure II.12.6 (a) Bathymetric map (depth contour plot) of Squam Lakes, NH, and (b) GIS-derived
3D model and visualization of the lake bottom structure derived from the bathymetric data.
quality data spatially, it became apparent that many of the small coves and embayments
were areas of more degraded water quality. The data suggested that the lake did not react
uniformly to watershed inputs; that it was not just one big reaction vessel or bathtub
as is commonly assumed for many large lake systems. This concept was further enhanced
by taking the bathymetric map (depth contour plot) of the lake and using the GIS to create
a 3-D model of the lake bottom (Figure II.12.6). No experience in topographic readings
was necessary to be able to see how the lake was really made up of multiple basins
connected together and that each of these basins had high sills around it. Many of these
sills reached up to shallow water depth, well above the basin thermocline, effectively
separating the bottom waters of the lake basins for most of the year. Thus, the GIS allowed
for the definition of functionally separated lake basins within the lake system. Of the 18
lake basins defined, 17 had been monitored through the volunteer program, so further
GIS spatial analysis of water quality conditions could be done.
Geographic referencing and spatial analysis

With the basins defined, they could be associated through the GIS with the abutting
subwatersheds. This would allow for analyses of what characteristics of the land around
the basins had an influence on the basins water quality. Although our study team had
the luxury of an extensive GIS database of land cover (right down to the type of tree stand
from aerial photography), we started with some basic GIS analysis using information that
would be more readily available to localities across the state. With some relatively simple
data analyses, areas of the lake that react more critically to nutrient loading were defined
(Schloss and Rubin, 1992). We found good relationships between water clarity for each
basin and the mean depth (derived from the GIS by dividing the surface area of the basin
by the volume of the basin) and between algae levels and a combination of mean depth
and the ratio of the basin surface area to the area of land that drained into it (abutting
subwatersheds). This explained much of the data variation in all basins.
A generalized land-cover analysis was also undertaken to determine the land use
water quality relationships that occurred throughout the watershed. It was found that the
shoreland zone land cover (an arbitrary 250-ft area from the lakeshore created by the GIS)
explained less water quality variation than the total subwatershed land cover. This implies
that although shoreline regulations are important for the Squam Lakes (and most likely
our other pristine lakes) activities throughout the watershed also have major impacts.
Chapter II.12:
567
LEGEND
In-Lake Resources
Loon Habitat
Coldwater Fishery
Salmon Areas
Whitefish Areas
Smelt Fishery
Smelt Brooks
LOON
NESTING
Smelt Reefs
Fish Reefs
SALMON &
WHITEFISH
Holes
TROUT
REEF
SM Bass Reef
Lake Trout Reef
Sampling Site
Public Access
SMELT
BROOK
BASS NEST
700
700
1400 Feet
Figure II.12.7 Section of GIS map displaying the elements collected for the in-lake wildlife resources
inventory of the Squam Lakes, NH. (Color version available at www.gsd.harvard.edu/watercolors.
In-lake resource inventories and integration of the data layers

Through community advisory groups we learned that other aspects of the watershed
besides water quality held equal if not greater importance. To that end, a GIS layer of loon
habitat (provided by volunteers of the NH Loon Preservation Society), bass nesting areas,
cold-water fish reefs and deep holes, and smelt brooks (from NH Fish and Game and
volunteer surveying) was created (Figure II.12.7). The GIS could then reference the various
in-lake and shoreline wildlife extent contained in each of the basins. Now the GIS Watershed Natural Resources Inventory was complete, with information of in-lake water quality
conditions, susceptibility to additional nutrient loading, and significant wildlife resources.
From this information, the GIS was used to perform a bottom-up analysis to locate the
lakes most critical areas. For each basin and adjoining subwatersheds the GIS simply
averaged together all of the criteria scores. Depending on the priorities of any specific
town or stakeholder group, however, a similar process may be employed that uses a
weighting multiplier to favor higher-priority concerns. The resulting integration was best
visualized for stakeholders by draping a color (light or cold for less critical, reddish or
hot for most critical areas) over the 3D plot of the lake basins. Figure II.12.8 is a grayscale
representation of this type of plot.
Useful watershed natural resource inventory GIS products

Whereas we now use many of the GIS visualizations in snazzy animated computer presentations for our statewide and regional education programs, putting the information to
use in implementing local decisions and management plans was a primary goal. The
materials produced for the community associations and decision makers in the watershed
had to be more functional. Many towns and most citizens still do not have easy access to
GIS systems, so a more low-tech set of products was developed. For the town decision
568
LEAST
MOST
Figure II.12.8 GIS visualization of critical lake basins from integrated analysis of the Squam Lakes
Natural Resources Inventory data. Darker shading represents the more critical waters. (Color version
PRINTED TABLE
SQUAM LAKE
WATERSHED
14
12
11 13
9 10
34
2
4
3
33
Subshed=13 Impacts: BASIN 10 (Deep Haven Site)
15
17
16
21 20
22
23
24
7
6 8
32
18
19
25
31
30
26
29
27
28
Basin Area: 962.65 Acres

Basin Volume: 112.4 E+7 cubic ft
Water Quality: Not as Critical (2 of 4); Algae Blooms=RARE
Loon Habitat: 6463.84 ft Score=2 of 4=> Moderate
HARDCOPY MAP
Bass Reef: 75.92 Acres=> High Value

Smelt Brooks: 3=> High Value
Whitefish Reef: 1=> High Value
Trout Reefs: 25.65 Acres=> 2nd Highest Value
Salmon Holes: 2=> High Value
Warm Water Fish Pts: 3 of 4=> High Habitat Value
Cold Water Fish Points: 3 of 4=> High Habitat Value
Figure II.12.9 Diagram of low-tech GIS products created for utilization of Squam Lakes Natural
Resources Inventory data by local decision makers and stakeholders. (Color version available at
makers, a map of the watershed area was provided, delineating the various subwatersheds
and subbasins of the lake labeled by a unique number. These numbers were then referenced to a printed table that contained the water quality and resource information of both
the basins and the abutting subwatersheds (Figure II.12.9). Tabled information could also
be captured to a spreadsheet or a database system, and digital maps could also be provided
to those with GIS display systems. Now, instead of having to decide on the approval of
a project based solely on information provided by the applicant, the decision maker can
look up the subwatershed where the project is being proposed, check on the important
lake resources that may be impacted, weigh the benefits and concerns, and have the
applicant address specifically how he or she will minimize loss or impacts to that resource.
Additionally, with critical areas of the lake and associated subwatersheds catalogued, a
targeted land protection effort using voluntary and (if possible) regulatory approaches
can be implemented.
Chapter II.12:
569
Figure II.12.10 Photograph of Lake Chocorua, NH, with Mount Chocorua in the background. (Color
Results and implementations

Information gained from this study has benefited the lake association, the surrounding
towns, state agencies, and researchers. The lake monitoring group has utilized the GIS
study to reevaluate its monitoring strategy by conducting more frequent sampling at the
more susceptible hot-spot sites. The lake association has also used the information to
call for better enforcement of boat speed rules for the more critical areas defined by the
GIS. One of the towns is reevaluating its zoning and planning laws given the results of
this study. Another town has used the watershed natural resources inventory data in
planning board meetings. Information from the GIS study was used by our state environmental enforcement agency to evaluate its permitting of a large development project in
the watershed. Also, one basin that did not follow the expected outcome of the analysis.
We later learned from investigation that a significant amount of fill used to cover over a
wetlands area at the end of this bay impeded the flow of a tributary that acted to flush
the bay out during the high-flow events that follow the spring snow melt. This was the
only embayed basin with a restricted tributary. It serves as an example of the damage that
can result when wetland areas are filled in.
The Lake Chocorua watershed project

The Squam project afforded us the knowledge of which GIS data layers and analyses
proved most useful to determine various watershed characteristics and related water
quality conditions. For the Chocorua project, where we had limited resources, our first
step was to define the issues that we needed to address and then determine what GIS
analysis approach would be necessary.
Background
The watershed for Chocorua Lake is 13.2 square miles and is well protected except in a
few vulnerable areas. The still-pristine quality of Chocorua Lake and its view of Mount
Chocorua (Figure II.12.10) are widely known as an icon of New Hampshire. The area
attracts many tourists, especially in the autumn when the lake reflects the chiseled mount
570
surrounded by the multicolored leaves of autumn. Compared with most NH lakes, relatively few houses are located in the watershed. Those near the lake have conservation
easements and covenants, which require extensive setbacks for housing and septic systems
and limit shoreland clearing. The lake provides local residents and visitors an area for
swimming, fishing, and low-impact boating.
Well over half of the watershed is uncut forest, the highest-quality watershed. About
95% of this land is protected by conservation easements, which have preserved upland
wooded buffers all around the lake, except on the eastern shore where State Route 16
borders the lake. Several large wetland complexes protect the lake by providing natural
filtration and assimilation. Although the lake is surrounded by protective forest cover in
most of its watershed, areas in the upper watershed are prime for development. Chocorua
Lake is very shallow, with an average depth of 12 ft. As a result, sunlight reaches through
most of the water column. Thus, even low concentrations of nutrients are readily available
to algae and other plant life that occur throughout the lake. Visual surveys of the entire
shoreline suggested that Route 16 was one of the main sources of erosion and nutrients
into the lake; however, before the NH Department of Transportation (NH DOT) would
agree to mitigate the runoff and erosion situations, they wanted to be sure that this was
a major problem for the lake compared to runoff and tributary inputs from the rest of the
watershed.
Lake water quality and diagnostic study

Long-term sampling through participation in the NH Lakes Lay Monitoring program
(mentioned in the Squam example) has shown a continued decrease in lake water quality
with symptoms such as increasing algae levels and aquatic weed growth, and reduced
water transparency. In addition, synoptic sampling near the lakeshore indicated the
highest concentrations of total phosphorous were occurring along the lakes eastern
shore, which borders the highway. In order to be complete and to provide additional
watershed information to allow the Chocorua Lake Conservation Foundation (CLCF)
to develop a land protection strategy, a water and nutrient budget diagnostic study was
initiated.
The GIS Natural Resources Inventory employed for this project was directed toward
compiling the information from which GIS analysis could provide the necessary data to
feed into the water and nutrient budget calculations and the predictive runoff and loading
diagnostic modeling. Of particular importance were the conservation lands, generalized
land cover (derived from remote sensing that was combined with an on the ground
watershed survey to determine development densities), and the extent of wetland complexes. In addition, the road and highway culvert locations were located and digitized
using GPS to facilitate the delineation of subwatersheds that contributed to the road runoff.
The results of the lake watershed diagnostic study confirmed that for its areal extent, as
well as for its relative water contribution, the Route 16 runoff measured through the drainage
culverts contributed the greatest amount of nutrients into the lake (Figure II.12.11); however,
it also disclosed that the northern watershed area contributed the majority of the nutrients
overall. Additionally, it documented the importance of the wetlands in those subwatersheds acting to both reduce the nutrient loading as well as shunt the nutrients until the
end of the growing season. All the pollutants were not necessarily generated and coming
directly off the highway, because there was evidence of sediment deposition occurring
from up-watershed sites bordering the highway, but the culverts were acting to channel
everything directly into the lake. This information provided the NH DOT with the justification it needed to assist in the mitigation of the highway sites. It also directed the CLCF
to concentrate its land protection efforts on the northern watershed lands, especially those
Chapter II.12:
571
Chocorua Lake Watershed

Nutrient Loading Comparison
0.200
Nutrient Loading/Area
(Kg TP/Hectare/yr)
0.180
0.160
CHOCORUA RIVER
0.140
WESTERN DRAINAGE
0.120
RT 16 DRAINAGE
0.100
0.080
0.060
0.040
0.020
0.000
Figure II.12.11 Bar graph depicting relative areal total phosphorus loadings from various subwatersheds of the Lake Chocorua watershed. (Color version available at www.gsd.harvard.edu/
that contained the large wetland complexes. Strategies to be used for this effort include
outright purchases, covenants and easements, and the reevaluation of existing zoning
regulations.
GIS analysis of local wetland buffer options for Deerfield, NH

The examples discussed previously involve a single major watershed focus of concern
and analysis of the component subwatersheds. The following example is concerned with
protecting specific water resources: wetlands, within a single jurisdiction, the town of
Deerfield, NH. It was chosen to highlight additional analysis approaches used to help
implement resources protection (see also Chapter II.14 for an additional GIS study of
protective buffer strips).
Background
A collaborative effort between the Audubon Society of NH, the NH Office of State Planning, UNH Cooperative Extension, and the USDA Natural Resources Conservation Service
produced a guidance document on riparian buffer function that included recommendations for regulatory and nonregulatory buffer widths (Chase et al., 1995). It provides
municipalities with both a scientific rationale and practical actions for protecting and
preserving naturally vegetated upland areas that border surface waters and wetlands.
Through a review of the current scientific literature and recommendations of other states,
and with priority focused on water quality protection, a reasonable minimum buffer
width of 100 ft is recommended in the buffers guide. A larger buffer is recommended for
sensitive wetlands (bogs, fens, white cedar swamps), prime wetlands, endangered or
threatened species protection, or to support wildlife habitat more thoroughly. Ultimately,
local decision makers will need to determine the most appropriate buffers to suit their
needs and the means for establishing them (see also Chapter II.1 for an additional discussion of protective shoreline buffers). In an effort to demonstrate how GIS might be used
to assist in the decision-making process, a pilot project was undertaken and the results
presented to a statewide audience at a GIS workshop for decision makers sponsored by
the NH Office of State Planning and the University of New Hampshire.
572

CHANGES TO ZONING OVERLAYS
Current
Regs. 577
434
2880
Wetland Buffer
Stream Buffer
3840
Proposed
Buffer
1154
Shoreland
Protection
434
0
1000
2000
3000
4000
ACRES IMPACTED
Figure II.12.12 Graph showing pre- and post-implementation effects of incorporating 100-ft buffer
overlay for wetlands and low-order streams in the town of Deerfield, NH. Each bar in set represents
land acreage impacted by a specific regulation.
The town of Deerfield, NH, completed a comprehensive inventory of its natural

resources in the spring of 1991 extensively using GIS (see Appendix D in Auger and
McIntyre, 1992). For investigating the various buffer scenarios it was first necessary to
take an inventory of the water resources of concern. From the GIS base map, surface waters
are already delineated. The GIS soils coverages were used to delineate wetlands areas
(from hydric soils classifications). The inventory of Deerfield disclosed that wetlands
comprise 86% of the towns water resources acreage and many are connected and lie
within stream corridors that run throughout the town.
Existing lands under regulatory restriction

The existing regulatory buffers and setbacks in the town were analyzed using GIS. Two
sets of state laws and regulations are already concerned with maintaining a vegetated
buffer at the shoreline of lakes and streams. The Comprehensive Shoreline Protection Act
requires that a minimum tree basal area must be maintained at greater than 50% within
150 ft from the shore of lakes greater than 10 acres and fourth-order or greater streams
(except those in the NH Rivers Program). State forestry regulations also maintain this
requirement for land within 50 ft from a perennial stream or brook. A setback of 75 ft for
buildings and septic systems bordering wetland areas is also required under town regulations. The GIS display of these overlay zones indicates the existing acreage of these areas
as 434 acres under the Shoreland Protection Act, 577 acres under the forestry regulations,
and 2,880 acres bordering wetlands with town-mandated setback restrictions.
Adopting the 100-ft buffer recommendation and exploring other scenarios

Through the use of GIS, maps were produced that visualized the extent of lands that
would be impacted by the new recommendations. Imposing the 100-ft buffer overlay for
wetlands and streams about doubles the protective acreage around streams and adds
another 1,000 acres that border wetlands (Figure II.12.12). This represents a 40% increase
in the protected areas. This initially might appear to have a large impact and involve many
landowners, but once already protected lands are subtracted out and large land tracts are
considered, the number of landowners impacted is greatly reduced. Thus, using an overlay
of the town tax map, the decision makers are now able investigate the degree to which
different lands might be affected by various regulatory approaches.
Chapter II.12:
573
The buffers document suggests a buffer larger than 100 ft for more critically important
or prime wetlands and for wildlife corridors. Our study explored the use of a 200-ft buffer
in our analysis. An overlay of this buffer was created to visualize the impact and to discern
whether this size chosen was adequate to serve both water quality and wildlife habitat
concerns. The resulting analysis indicated that with the 200-ft buffer some wetlands in
the sample area would be connected to each other, but others would not. If habitat
considerations are a goal, the GIS analysis indicated that other, perhaps nonregulatory,
methods would be needed to establish habitat connections among all of the critically
important wetlands in the town.
Optimizing nonregulatory protection approaches

Nonregulatory approaches to buffer protection were also explored in detail with GIS
analyses. For purposes of wildlife habitat and travel corridor protection and to maximize
the benefits of conservation lands, acquisition of larger buffer areas may be required. To
achieve this level of protection a town may have to rely on land acquisition and/or
conservation easements. Use of the GIS information regarding the wetland and stream
locations, existing and proposed buffer overlays and habitat land-cover information, along
with property or tax map overlays and existing conservation lands, can help decision makers
choose the most cost-effective way of achieving their goals. Creating GIS visualizations of
the extent of impacted areas and the functional connectivity of the lands can provide stakeholders with the perspective needed to gain community acceptance of the proposals.
Critical watershed land and resources analysis

The ultimate synthesis of a GIS Watershed Natural Resources Inventory analysis is to
target those most critical lands and water resources that are at the greatest risk. The risk
may be the total loss or partial degradation that can occur due to development or change
in nearby land-use activities. The identification process utilizes a combination of the topdown and bottom-up approaches already discussed in the examples here but proceeds
through a series of processing steps that integrate the analysis outcomes to yield the results.
We have been piloting a version of this approach in our coastal watershed communities,
where it is predicted that the greatest near-future expansion of our state population is
most likely to occur.
Step 1: Land capability analysis

Initially a conventional top-down land capability analysis is undertaken as previously
described in the Squam Lakes Watershed Study. Lands already developed, permanently
protected lands, and land constrained from development due to local or state regulations
as well as water and wetlands and their associated buffers are located. The land that
remains is cataloged as developable.
Step 2: Prioritize land and water areas of special concern

From the watershed natural resources inventory select and, if warranted, weight the list
of important areas of concern. These may include (but are not restricted to):
Wetlands (designated as prime, highly functional, large, or part of a cluster or
corridor)
Wildlife habitats (in-water, riparian, and upland)
574
Important lands (unfragmented lands, forest, agricultural lands, significant lands)

High-quality waters
Sensitive waters (low assimilation capacity, vulnerable to development)
Endangered or unique plant or animal communities
Important recreation areas
Commercially important plants and animals (aquatic and terrestrial)
Once the selection and weighting is decided a GIS analysis is employed to produce
a co-occurrence data layer. This coverage indicates lands that contain (or in terms of the
water resources, abut) the most types of, or highest-priority (if weighted) resources. Typically, this is displayed on a GIS map as a continuum of shading or coloring. Again, this
is similar to the critical basin process employed for the Squam Lake analysis. This result
is a bottom-up resource analysis product.
Step 3: Determine areas most likely to become developed

By incorporating data on existing infrastructure such as roads, power lines, and water
and sewer service, those areas most likely to become developed next can be predicted.
This is based on the assumption that it is cheaper to build new development where you
are close to existing infrastructure than to cover the expense of expanding that infrastructure out to lands farther away. This also may be considered a bottom-up approach only
because it concerns on the ground infrastructure resources.
Step 4: Process all of the elements together

Combining the results of steps 1 through 3 together creates an analysis of those developable lands (step 1) that have the highest resource value (step 2) and are the most likely to
be developed in the near future (step 3). This may be presented to stakeholders on a map
using various shading or colors for co-occurrences, by employing a series of mylar overlays, or by viewing manipulated GIS layers directly from the computer.
The final step would be to develop a protection strategy (or strategies) to target these
critical lands. Further analyses using land parcel data or different analysis runs may
have to be employed to target the appropriate critical lands for the specific strategy or
approach undertaken.
Lessons learned and future predictions

GIS and related spatial technologies have proven themselves as valuable tools for analysis,
assessment, and communicating to the public. Although high-tech analyses are critical
to successful watershed natural resources inventory projects, the low-tech information
transfer products have a much wider utility in regards to facilitating implementation of
management plans and protective measures. The GIS-based watershed natural resources
inventory data analysis consistently resulted in a proactive approach instead of a reactive
response. It allowed for an informed watershed community to become an empowered
community with fact-based concerns that could more easily result in consensus and
support. An additional important lesson was that all the possible data in the world is not
needed for a useful outcome. The most critical element for success is to understand the
problem to be addressed and then to determine what data will be required to facilitate
informed decision making. As demonstrated in the examples here, as well as in the
symposium roundtable discussions that follow, a key to optimizing the power of GIS
Chapter II.12:
575
spatial analysis is to combine the bottom-up, resource-based data analyses and the topdown land capability analysis to define those most critical lands and watershed resources
to direct management, regulatory, and nonregulatory attention and efforts.
A more complete understanding of landscape-scale interactions within watersheds as
well as their relation to water quality and in-water resources can be provided through
GIS-based spatial analysis. Currently, our resolution is limited for most elements such that
site-specific scaled analyses are not yet within our reach. However, given the rate at which
the technology is improving (more detailed and near real-time remote sensing; more
powerful user friendlier GIS software and related spatial analysis tools), the necessary
resolution to achieve that level of accuracy is just on the horizon.
Literature cited
Auger, P. and McIntyre, J., Natural Resources: An Inventory Guide for New Hampshire Communities,
Upper Valley Land Trust and Univ. of NH Coop. Ext., Durham, NH, 1992.
Chase, V., Deming, L.S., and Latawiec, F., Buffers for Wetlands and Surface Waters: A Guidebook for New
Hampshire Municipalities, Audubon Soc. of NH, Concord, NH, 1995.
Congalton, R.G. and Green, K., The ABCs of GIS: an introduction to geographic information systems,
J. Forestry, 90, 13, 1992.
Schloss, J.A. and Mitchell, F., Promoting watershed based land use decisions in New Hampshire
communities: Geographic information system aided education and analysis, in Proc. Watershed 96, Baltimore, 1996, 830833.
Schloss, J.A. and Rubin, F.A., A bottom-up approach to GIS watershed analysis, in Proc. 1992
GIS/LIS Conf., San Jose, CA, Amer. Soc. Photogrammetry and Remote Sensing (and others), Vol.
2, 672680, 1992.
Scott, D. et al., Squam Lakes watershed plan, NH Office of State Plan., Concord, NH, 1991.
Stone A., Natural Resource Inventories: A Guide for New Hampshire Communities and Conservation Groups,
Upper Valley Land Trust, NH Div. Forest and Lands, USDA Forest Serv. and Univ. of NH
Coop. Ext., Durham, NH, 2001. (A revision and update of Auger and McIntyre, 1992.)
576
Response
Janus planning:
Using computer tools to look backward and forward simultaneously
Geographic information systems, or GIS, has emerged as an extremely powerful tool in
water sensitive planning. The ability to map watershed attributes as varied as soil moisture
(Chapter II.10 by John Felkner and Michael Binford), riparian buffers (Chapter II.14 by
Robert France, John Felkner, Michael Flaxman, and Robert Rempel), or surface permeability and aquifer recharge potential (Chapter II.15 by Amir Mueller, Robert France, and Carl
Steinitz) is an important step in identifying resources for protection from future development. This is the bottom-up approach that Jeffrey Schloss promotes in this chapter.
Unfortunately, GIS has also been used as an excuse to identify areas that are imagined to
be capable of absorbing more development what Schloss criticizes as the top-down
approach.
As Schloss identifies, resource or bottom-up GIS analysis leads to preemptive planning
and protection, which in turn helps toward fostering watershed stewardship. Stewardship
comes about through education, and it is here where GIS serves its most important role
in water sensitive planning. Through providing convenient visualizations of complex and
occult attributes, GIS allows stakeholders to become informed not just about the current
status of their watershed, but also about the status of past conditions as well as the
predicted status of future conditions given different development scenarios (see
Chapter II.15 by Mueller, France, and Steinitz).
The case studies described in this chapter demonstrate a range of GIS applications
such as build-out analyses, lake bathymetry and subwatershed mapping, sensitive species
locations, road effects, wetlands, and riparian buffer options. Schloss presents a logical
framework for undertaking watershed planning through GIS analysis: critical lands analysis land capability analysis prioritization of areas of special concern determination of areas most likely to become developed combination of all elements into a final
planning summary.
Schloss also briefly touches on a serious weakness in GIS technology that offers a
pertinent, cautionary lesson for water sensitive planners. For it is the very elements that
make GIS so effective as a communication and education tool that also seduce many into
using the techniques in a superficial and cavalier fashion. There is a cardinal need to move
beyond the common use of GIS analysis to produce snazzy computer-based presentations,
and instead to push the tool toward new and carefully thought-out directions in watershed
management (e.g., see Chapter II.10 by John Felkner and Michael Binford, and Chapter II.13
by Margot Cantwell). In the end, it is important to remember that GIS is a means toward
another end, never the end itself.
chapter II.13
The effect of spatial location in landwater

interactions: A comparison of two
modeling approaches to support
watershed planning
(Newfoundland, Canada)
Abstract
Two landscape models of landwater interactions were compared for their usefulness in
assisting planners in zoning watersheds and evaluating future development proposals. A
number of lake and water quality models have widespread acceptance by watershed
managers (e.g., Dillon et al., 1986; U.S. EPA, 1988; Riley, 1988). These models are fit to a
particular lake and watershed using landscape, climatic, and basin characteristics as well
as point and nonpoint source pollutant inputs. In many jurisdictions, these models have
been used to predict lake carrying capacity and establish maximum target levels of pollutant input to maintain a desired water quality level. This maximum target level is then
described in planning documents in terms of the maximum amount of development (or
land use) to be permitted in the watershed (e.g., Kings County, Nova Scotia, and Muskoka
Lakes Region, Ontario). One question facing watershed planners implementing the target
level is the effect of spatial location on land-use or development impact does it matter
where in the landscape the development/land use goes? Two approaches were compared
in this study. The first assumed that control of land-use type and quantity was sufficient;
the second approach considered the implications of land-use location. The second
approach was determined to be a useful tool to zone the landscape and evaluate development proposals. The approach builds on the standard lake water quality modeling
approach, allowing managers to predict the amount of pollutant input attributable to a
particular quantity of land use in a specific location within a lake catchment. The model
approaches were evaluated for the Government of Newfoundland and Labrador together
with Environment Canada, using data for Gander Lake and its catchment.
1-56670-562-2/02/$0.00+$1.50
577
578
Introduction
In 1994, the Government of Newfoundland and Labrador faced opposition from a logging
company operating within the Gander Lake catchment. The company challenged the
provinces forestry setback of 150 m (DELA, 1989, 1991) from Gander Lake, stating that
the standard setback was not scientifically defensible. The challenge raised the question
of the appropriate spatial configuration of land-use restrictions within a catchment, and
whether or not water bodies were adequately protected (and scientifically defensible) as
defined by the policy set in the Provincial Environment and Lands Acts. A number of
studies were undertaken to explore this question, including this study, which was to
identify and apply a landscape model of landwater interactions that enabled the spatial
location of development restrictions to be considered.
Gander Lake, an approximately 50,000-m-long 2,000-m-wide 290-m-deep lake is
the drinking water supply for the town of Gander and several smaller towns and villages
in central Newfoundland (48.5 latitude; 55.5 longitude). The lake and Gander River,
which bisects it, are a world-class Atlantic salmon fishing resource. These two resource
functions allowed Gander Lake and River to be studied over the past 20 years by Fisheries
and Oceans Canada, Memorial University, and the Newfoundland and Labrador Department of Environment. The town of Gander has a Canadian Air Force presence and so
Transport Canada records of treatment plant performance and airport and town stormwater runoff quality were documented and available. The catchment represented an excellent opportunity to consider the spatial effects of land use.
The project involved a number of tasks, published in the Watershed Management Plan
for Gander Lake and its catchment (EDM, 1996). Water quality analysis was undertaken
by Jacques Whitford Environment Limited of St. Johns, Newfoundland, and three water
quality models (Dillon, MinLake, and WASP4) were fit to the lake by R. H. Loucks Oceanology of Halifax, Nova Scotia. This work indicated that the lake was oligotrophic, dimictic,
and considered resilient to the pressures of land development. The lake currently meets
or exceeds Canadian drinking water quality guidelines. Phosphorous was identified as
the limiting nutrient and as the focus for water quality management.
Circulation characteristics and water quality results indicated a total annual phosphorous loading for the lake of approximately 11,730 kg/year. Of this total, phosphorous input
from precipitation was estimated at 2800 kg/year. The only point source inputs are from
the sewage treatment plants (none of which discharge directly to the lake). Using treatment
plant records together with input stream water quality results and assuming all cottages
(of which there are only a few) have failing on-site sewage systems, the maximum possible
loading to the lake attributable to sewage was estimated at 130 kg/year. This left approximately 8600 kg/year of phosphorous in the lake attributable to runoff from catchment
land uses. Further modeling indicated that the lake could sustain an approximate 50%
increase in phosphorous input before its trophic state moved from oligotrophic to
mesotrophic.
With a model fit to the lake, the question for the watershed managers was how to
assign land uses within the catchment to sustain the target water quality level. In planning
documents, land uses are assigned by zoning maps that have the following attributes:
land-use type (what use and at what concentration/density?), area (what extent?), and
location (where specifically can development occur?). In making decisions about assigning
zones, managers must consider both how much development (or preservation) should be
allowed and specifically where in the landscape that development/preservation should
be located (see Chapters II.7 and II.12).
Chapter II.13:
The effect of spatial location in landwater interactions
579
The following questions were considered:

1. What is the source of the phosphorous in the catchment runoff and how might it
be distributed spatially?
2. Are standard land-use export values used to calibrate lake water quality models
adequate for policy development, including zoning the landscape?
3. Is a method that includes a spatial component a better approach for predicting the
contribution of phosphorous from the different land uses?
4. What is the likely impact of proposed future land uses on the phosphorous balance
of the lake?
5. What is the most scientifically defensible approach to zoning for future land uses
within the catchment?
Method
Two approaches were considered for managing the Gander Lake catchment:
1. The fact that phosphorous input is primarily described by land-use type and
location is a minor consideration. Watershed policy created using this assumption
would typically set a maximum development level in the policy document (e.g.,
maximum number of acres of development, or a total number of cottages). This is
a common approach to watershed modeling and policy development. For example,
each of the three water quality models fit to Gander Lake used this approach to
establish the land-use input and the carrying capacity of the lake. This approach
was referred to as the land-use model.
2. The fact that phosphorous input is described as much by land-use location in the
catchment as by the actual land-use type is a consideration. Watershed policy
created under this assumption would prescribe both how much development or
land use and in which specific areas of the catchment. Policy might also allow for
increases or decreases in land-use type or extent depending on specific locational
criteria. This approach was referred to as the spatially explicit model.
Phosphorous is commonly the limiting nutrient in watersheds managed for water
clarity (see Chapter II.17). Similar to the Gander Lake catchment, the various land uses in
the watershed are frequently a large contributing source. Of the land-use fraction (after
direct precipitation and point sources are subtracted) soil is typically the primary source,
as phosphorous is a natural soil nutrient and phosphorous tends to attach itself to soil
particles (Dillon et al., 1986; Hickman, 1987; Leopold, 1994; Loucks, 1993). Soil reaches the
water column though erosion processes that occur naturally at relatively low rates and
often are accelerated when soils are exposed, such as during construction for development,
field tilling in agriculture, and silvaculture or road-building practices in forestry (see
Chapter II.16). In addition, phosphorous released from organic matter (e.g., garbage on
streets, tree leaves on yards and in forests, or pet feces), as well as phosphorous added as
a soil fertilizer (e.g., to lawns, gardens, and agriculture fields), also tends to attach itself
to soil particles. Thus, eroding soil itself, as well as dirt washed from streets and other
land surfaces, tends to carry the significant phosphorous load associated with land use.
Therefore, although phosphorous is derived from many sources, its attachment to a soil
particle and its delivery via erosive processes make it possible to model it as part of a soil
erosion process. This is particularly the case in nonurbanized catchments where phosphorous
580
associated with soil erosion is normally the primary source of phosphorous in lake waters
(Leopold, 1994; Dunne and Leopold, 1978).
To compare the different approaches, two different landscape models describing
landwater interactions were fit to the Gander Lake catchment. The first model was a
standard land-use export approach, referred to as the land-use model. Land-use type is
considered by many researchers to be the most sensitive variable in determining how
much pollutant a given land area will contribute to a lake and considered sufficient to
establish lake water quality models (e.g., Dillon et al., 1986; U.S. EPA, 1988). Reference
values for phosphorous and suspended solids contributed from different land uses may
be found in the literature. To evaluate the first approach, land-use export values were
developed for the different land uses found in the catchment. The export values were
derived from similar landscapes in New England and Nova Scotia and calibrated to the
Gander Lake catchment using water quality data and land-use areas from a local subcatchment. The land-use model is defined by Eq. II.13.1.
n
TSS =
(E A )
(II.13.1)
where
TSS = total annual suspended solids input to the lake in kg/yr
n = number of land uses
E = export value of suspended solids for each land use (kg/ha/yr)
A = area of each land use (ha)
To explore the second approach, a landscape model for landwater interactions with
a spatially explicit component was used. The modeling approach taken was similar to
that used by Binford (1989). The approach is a variation on the Universal Soil Loss Equation
(USLE), with a modification to consider sediment delivery to a lake. The USLE is a tool
developed for agriculture to estimate the potential of the landscape to erode. The USLE,
by far the most widely used method for predicting soil loss, was developed by Wischmeier
and Smith (1978). The USLE has been used primarily in agriculture watersheds to predict
soil loss in fields (e.g., Mellerowicz et al., 1994). In recent years, it has been applied to
other land uses, including forestry and urban and suburban development. To date, only
limited application of the USLE to regional-scale planning has been used, although the
approach has merit (e.g., Binford, 1989; Rees, 1996).
The USLE has four locational factors: an R-index describing how erosive the local
rainfall is, (essentially how hard the rainfalls, and thus its likelihood to dislodge soil
particles); a K-index describing the physical nature of the soil; and two factors, L and S,
describing the shape of the terrain (both the slope and the effect of gravity on slope effect,
referred to as the hill slope-length and hill slope-gradient factors). These four factors
represent a sort of thumb print of the catchment. Nothing can be done about these four
factors, and thus they represent an inherent sensitivity or a landscape unit sensitivity.
Two factors are used to consider land-use and property management. The C, or croppingmanagement factor, is a factor that considers the effect of the particular land-use type. The
P, or erosion-control practice factor, allows for a modification of C when certain management
practices are used. The P factor is well developed for agriculture, where different cropping
techniques (e.g., contour farming) have a demonstrated effect on the soil loss. P-factors are
being developed for forestry land uses reflecting different silvaculture practices (e.g., shelter
Chapter II.13:
581
wood and no duff disturbance practices as opposed to clearcuts and forest tilling). P-factors
for development landscape types are less well developed.
The USLE is
A = (RKLS) (CP) (2471)
(II.13.2)
(Inherent sensitivity) (Land use and Management)

where
A = soil loss (kg/ha)
R = rainfall erosive index
K = soil erodibility index
L = hill slope-length factor
S = hill slope-gradient factor
C = cropping-management factor
P = erosion-control practice factor
For the Gander Lake catchment, the USLE was solved in the GIS for each 1-ha grid cell
across the landscape. Obviously, all of the eroded soil does not end up in the water column
of Gander Lake as suspended solids. Total soil loss was translated to a total annual suspended
solids (TSS) loading by multiplying the total soil loss by a delivery ratio (the fraction of the
eroded soil likely to reach the lake) and then multiplied by the percentage likely to remain
in suspension (i.e., the settlable solids fraction was removed). This result was then summed
across the catchment for a total annual TSS contribution (as per Eq. II.13.3). The map algebra
used in the spatially explicit model is illustrated in Figure II.13.1.
n
TSS =
( A D S)
(II.13.3)
where
TSS = total annual suspended solids input to the lake in kg/year
n = number of 1-ha grid cells
A = soil loss (kg/ha/year)
D = delivery ratio
S = % of soil in suspension (kg/ha/year)
Equations II.13.1 and II.13.3 can also be used to estimate the total annual phosphorous
loading to a lake by using published export values for phosphorous in Eq. II.13.1 (as
opposed to suspended solids), and by multiplying A by the fraction of the soil that is
phosphorous in Eq. II.13.3. Total suspended solids (TSS) was chosen for use in the Gander
Lake catchment because there were better local export values (E, in Eq. II.13.1) for TSS
than for phosphorous, and the primary objective was to compare two landscape models
of landwater interaction for their usefulness in preparing plan area policy.
Both modeling approaches were calibrated by fixing the annual soil loss in the catchment and then calibrating for the land-use variables. The annual soil loss in the catchment
was established as follows. First, using the average measured TSS concentration in several
subcatchment input streams together with the total estimated catchment flow, a value of
7,077,000 kg/year of TSS was derived (5.8 mg/l TSS 1.212 E12 l/year = 7.077 E6 kg/year).
10
Watercourses
Elevation Model
Forest Inventory
Watercourses
Elevation Model
Distance from
Watercourses
Slope
Cover / Practice
Factor
Slope - Length Factor
Rainfall Intensity
Factor
Soil K Values
Sediment Delivery
Frictional Co-efficient
Erosion
Water Protection Map
MAP ALGEBRA FOR SPATIALLY EXPLICIT MODEL
Figure II.13.1 GIS map combinations used in the spatially explicit model of landwater interaction (Eq. II.13.3). (From
EDM, Environ. Design and Manage. Ltd., Land-usep plan for the Big Cove First Nation, 2000.) (Color version available
15
15
50
10
582
50
Soil Types
Chapter II.13:
583
Table II.13.1 Expected Annual Soil Loss from the Gander Lake Catchment
Based on Literature Values from Catchments of a Similar Size
Watershed Management Plan for Gander Lake and Its Catchment
Gander Lake catchment area
77,250 ha
298 sq. mi.
U.S. Water Resources Council,
Estimated Soil Loss from Drainage Areas
in the North Atlantic Region
Soil Loss (nonurbanized catchments)
tons/sq. mi./year
kg/year
High value
Average value
Low
1210
250
30
327,321,253
67,628,358
8,115,403
Estimated Soil Loss from Glacial Till Lithologic Type

Soil Loss (nonurbanized catchments)
tons/sq. mi./year
kg/year
High
Medium
Low
300
200
180
81,154,030
54,102,686
48,692,418
Sources: Dunne and Leopold (1978), and Leopold (1994).
Then, from particle size analysis of soil type data [15], the approximate percentage of
settlable solids (the sand and most of the silt, given the residence time in the lake) was
estimated to be about 87%. The TSS of 7,077,000 kg/year was thus assumed to be derived
from a total annual soil loss of approximately 54,500,000 kg/yr (7.077 E6 kg/year 7.69 =
5.44 E7). This annual soil loss value (54,500,000 kg/year) was then compared with literature
values of expected soil loss from similar sized catchments (see Table II.13.1). The resulting
correlation was very good, and the TSS and total soil loss in the catchment values were
adopted and fixed for the catchment.
Both landwater interaction models were set to achieve this resultant level of soil loss
or lake water quality impact. Both were then calibrated by adjusting land-use variables.
Characteristic values for E-factors in Eq. II.13.1 (land-use model), taken from New England
(Borman and Likens, 1979) and Nova Scotia (Vaughan Engineering, 1993), were modified
to reflect measured water quality results in a local subcatchment. C-factors and D-ratios
in Eq. II.13.3 (spatially explicit model) were also modified. No published C-factors were
available for Newfoundland land uses, and so more general literature values (Dissmeyer
and Foster, 1981; Dunne and Leopold, 1978) and values from Acadia National Park (Binford, 1989) were used as a starting place. P-factors were not used in the model (assumed
to be 1, or no modifying practice factor) due to a lack of published values.
Calibration for delivery ratios (D in Equation II.13.3) proved to be a difficult exercise,
moving the project more into the realm of planning than science (i.e., Are we more right
than wrong, as opposed to, can we demonstrate this for certain?). Similar to C-factors,
literature on sediment delivery ratios is very sparse, although there is some ongoing work
to better define them (e.g., Arbour, 1996; Binford, 1995). In addition, models for routing
sediment (and other materials) are being modified for application to urban and developed
land uses. For this study, delivery ratios were developed following the method of Snell
(1985). Working in an Ontario agricultural landscape, Snell assigned a high delivery ratio
584
to areas that were within 100 m of a stream or water body; a medium delivery ratio to
areas of steep slope (greater than 5%) that were set back from a water body by at least a
100-m forested buffer; and, a low delivery ratio to areas that were tributary to a depression,
or had shallow slopes (less than 5%) and were set back from a stream or water body by
at least a 100-m forested buffer. Snell further refined delivery ratios to reflect hydrologic
soil properties in the absence of a buffer zone. This condition (no buffer) existed so rarely
in the Gander Lake catchment that this refinement was not included.
Resultant GIS landscape maps and values describing the current landwater interactions were compared. Then both modeling approaches were used to evaluate proposed
future developments in the catchment and consider the potential effect of the spatial
component.
Results
Figure II.13.2 illustrates the GIS output map that results when the land-use model
(Eq. II.13.1) was applied across the catchment, using land-use areas and land-use export
values for suspended solids delivery. Final export values that achieved calibration are
shown in Table II.13.2, together with the literature values and values from a local subcatchment referred to as the airport ditch from which the calibration values were derived.
Figure II.13.2 reflects the areas of concern if the premise is that pollutant delivery to
the lake is primarily correlated to land-use type. The figure illustrates the expected export
of suspended solids based on the range of land uses currently existing in the catchment.
The map result for phosphorous export produces a similar pattern. Values range from a
high export value associated with developed and cleared landscapes to low export values
associated with forest and bog landscape types. It is interesting to note that the range of
potential contribution is relatively narrow, with a high of 1200 kg/ha/year and a low of
035 kg/ha/year.
Figure II.13.3 illustrates the map that results when the spatially explicit model
(Eq. II.13.3) is applied across the catchment. Values used to calibrate, as well as literature
values from which the calibration values were derived, are shown in Table II.13.3. The
R-index used was 870 as published for Gander (Wall et al., 1983). The K-indices used for
each soil type were derived from local soil surveys (refer to Table II.13.3). The L- and
S-factors were considered together. A slope map for the catchment was generated from
the digital terrain model (DTM) provided by the Newfoundland and Labrador Department of Environment, Water Resources Division. L was defined as the width of the grid
cell in the GIS (in this case 100 m). Using tables in Dunne and Leopold (1978) and
Wischmeier and Smith (1978), the LS (length-slope index) was derived. (A refinement to
this approach where the computer calculates the hill slope-length, as opposed to using
the grid cell width, has since been developed by the author (EDM, 1999, 2000). Final
C-values derived during calibration, as well as literature values used as a starting point,
are also shown in Table II.13.3. The final values used align well with the literature values,
with the following exceptions: (1) the C-value for development was lowered due to the
suburban nature of development in the catchment; and (2) the value for burned areas was
lowered because most of these areas are very old burns with little soil left to erode.
Delivery ratios for Snells high, medium, and low categories were estimated based on
literature values reported by Lowrance et al. (1988) and Cooper et al. (1987), and with
advice from Arbour (1996). A setback value of 200 m was used (as opposed to Snells
100 m), given the more recent literature on the effectiveness of various buffer widths (e.g.,
Lowrance et al., 1988; Cooper et al., 1987; Toth, 1990) and considering specific site conditions in the Gander catchment (a more complete discussion of this is provided in the
Metres
10,000
4.5
27.5
30.7
4.5
5.6
2.6
Unvegetated
Blowdown, Burn and Brush/Regenerating
Recent Clear Cut
Land Cleared for Development
Developed
Gravel Pits
4.1
12.6
7.0
0.7
0.6
0.2
3,193
9,724
5,424
534
438
154
100.00
200.00
400.00
600.00
900.00
1,200.00
Figure II.13.2 Map of the Gander Lake catchment illustrating the contribution of suspended solids to the lake from the different land uses, as predicted
by the land-use model (Eq. II.13.1). Note that the town of Gander and the adjacent Gander airport are both predicted to have a relatively high impact.
(From EDM, Environ. Design and Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ. Ltd., Watershed management plan for Gander
Lake and its catchment, Govt. of Newfoundland and Labrador together with Environ. Canada, St. Johns, Newfoundland, 1996.) (Color version available
Prepared by: EDM Environmental Design and Management Limited, 1996.
Prepared for: Department of Environment, Water Resources Division

Government of Newfoundland and Labrador
Environment Canada
WATERSHED MANAGEMENT PLAN FOR GANDER LAKE AND ITS CATCHMENT
5,000
0
25.6
Forest
Bogs
% of
Catchment
TSS
64.5
10.3
7,945
ok
Bro
Benton
Soulis
Pond
49,838
0.00
Land Use
35.00
% of
Catchment
Area
Area
(ha)
Mile
oo
Br
Suspended
Solids
(kg/ha/yr)
een
Fift
r Lake
Gande
Gander
Jo
Appleton
Chapter II.13:
EXPORT OF SUSPENDED SOLIDS

FROM CATCHMENT IN LANDUSE MODEL
Glenwood
585
586
Table II.13.2 Export Values (E in Eq. II.13.1) Used in the Land-Use Model and Published Export
Values from Which the Calibrated Values Were Derived
Export (E) Values Calibrated for the Gander Lake Catchment Land Uses
Area (ha)
154
49,838
3693
7945
1510
3193
534
5424
438
4521
Calibrated TSS
Total
Export Values for Land Use
(kg/ha/year)
(kg)
Land Use
Gravel pits
Forest
Blowdown
Bogs
Burn
Unvegetated/bedrock
Land cleared for development
Recent clearcut
Development
Brush/regenerating
1200
35
200
0
200
100
600
400
900
200
In the water (approx. 13% of total)

Remaining 87% assumed lost to sediment
Total soil loss (kg/year)
184,800
1,744,330
738,600
0
302,000
319,300
320,400
2,169,600
394,200
904,200
7,077,430
47,364,339
54,441,769
Export (E) Values from the Literature

Land Use
Forested
15% cleared
Agriculture/golf
Residential unserviced
Development
Gravel pits
Recent clearcut
Brush/regenerating
40% development/60% forested
Suspended Solids (kg/ha/year)

New England
Nova Scotia
Airport Ditch
33
380
190
10100 (low)
lowmedium
mediumhigh
500
5001000
lowhigh
lowhigh
326
Sources: New England, Borman and Likens (1979); Nova Scotia, Centre for Water Resources (CWRS), as cited by
Vaughan Eng. (1993); Airport ditch, Gander International Airport, Safety and Technical Services, 19951996
monthly stormwater runoff data.
published document). In addition, low-value areas were split into two categories, Low
Value 1 for areas tributary to a bog, and Low Value 2 for other low-value areas. This was
done because very large bogs cover extensive areas of the catchment and export from
lands draining to these bogs will be less than export from other low-value areas. The final
delivery ratios uses are as shown in Table II.13.3. In Eq. II.13.3, S represents the portion
of the total soil loss that remains in suspension in the water column. The settlable solids
fraction of the total soil loss was estimated at 87%; thus 13% was assumed to remain in
suspension.
Figures II.13.2 and II.13.3 present a very different picture of where in the landscape the
suspended solids (and thus phosphorous) found in the lake are coming from. A number of
comparisons can be made:
Approximate
Suspended
% of
Solids
Catchment
(kg/ha/yr)
TSS
0 - 12.9
1.6
13 - 129.9
13.3
130 - 1,299.9
46.0
1,300 - 12,999.9
26.2
13,000 - 39,000
12.9
Bro
ok
Area
(ha)
38,191
22,609
8,979
679
44
Benton
% of
Catchment
Area
47.7
36.6
14.5
1.1
0.1
Soulis
Pond
Figure II.13.3 Map of the Gander Lake catchment illustrating the contribution of suspended solids to the lake from the different land uses, as predicted
by the spatially explicit model (Eq. II.13.3). Note that the town of Gander and the adjacent Gander airport are both predicted to have a relatively low
impact. (From EDM, Environ. Design and Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ. Ltd., Watershed management plan
for Gander Lake and its catchment, Govt. of Newfoundland and Labrador together with Environ. Canada, St. Johns, Newfoundland, 1996.) (Color version
10,000
Metres
Environment Canada
Soil Loss
(kg/ha/yr)
0 - 99
100 - 999
1,000 - 9,999
10,000 - 99,999
100,000 - 299,999
Mile
o
Bro
es
5,000
een
Fift
r Lake
Gande
Gander
Jo
Appleton
Chapter II.13:
EXPORT OF SUSPENDED SOLIDS FROM CATCHMENT

IN SPATIALLY EXPLICIT MODEL
Glenwood
587
588
Table II.13.3 K-indices, C-factors, and D-ratios (for Eq. II.13.3) Used in the Spatially Explicit Model
as Well as Published Values from Which the Calibrated Values Were Derived
K-Index Approximation
Soil Types
Bn, Bu, Wg
Ga, Gw
Ep
Fb, Gb, Sp, Sw
Py, Tn
Su, Bo, Ho
Peat, muck
K-Index Approximation*
0.34
0.37
0.065
0.27
0.12
0.44
K-Index
Assumed**
0.35
* Soil descriptions and cabability classes described in the 1972 Soil Survey were applied to tables for estimating
the K-index in Dunne and Leopold (for areas on map sheets 2D15 and 2D16). For the 2D14 map sheet (not
covered in the 1972 Soil Survey), more general descriptions and values in the 1993 soil survey were used.
** Descriptions and values for peat and muck are not available; value was assumed from extrapolation of tables.
Sources: Dunne and Leopold (1978); Wells and Heringa (1972).
C-Factor Estimate

Development
Gravel pits
Recent clearcut
Brush/regenerating
Burn
Blowdown
Forested
Unvegetated/exposed bedrock
Bog
Ponds
Acadia National Park*
Other Values
Calibrated for
Gander
0.1
0.5
0.01
0.01
0.01
0.001
0.002
N/A
N/A
0.0380.055***
0.0040.115**
0.010.04***
0.0030.011***
0.001***
N/A
N/A
0.05
0.2
0.2
0.02
0.01
0.007
0.01
0.001
0.002
N/A
N/A
Ontario*
Calibrated for
Gander
80%
50%
20%
80%
50%
20%
10%
* Values used by Binford (1989).

** Dissmeyer and Foster (1981).
*** Dunne and Leopold (1978).
D-Ratio Estimate
High value
Medium value
Low value 1
Low value 2 (bogs)
*Values used by Snell (1985).
Chapter II.13:
589
1. The range of values in each figure is significantly different. In Figure II.13.2, TSS
values range from 0 to 1,200 kg/ha/year. In Figure II.13.3, TSS values range from
0 to 39,000 kg/ha/year. The land-use model approach presents a much more even
distribution of impact than the spatially explicit model approach.
2. The models describe a significantly different distribution of impact. When spatial
implications are considered, the lowest impact areas are reduced in terms of the
total impact they contribute, and the highest impact areas are increased. For example, both models have only a small percentage of the catchment producing high
impact (generating high TSS). In the land-use model (Figure II.13.2), only 0.8% of
the landscape is in the top two categories; corresponding with 1.2% in the top two
categories in the spatially explicit model (Figure II.13.3). Both models also have a
relatively high percentage of the landscape producing low impacts on the lake. In
the land-use model, 78.9% of the landscape produces 100 kg/ha/year or less;
similarly in the spatially explicit model, 84.3% of the catchment produces 100
kg/ha/year or less. The different modeling approaches vary significantly, however,
in terms of the overall impact of these high and low areas. In the land-use model,
the 0.8% of the landscape in the top categories is responsible for only 8.2% of the
total TSS; while in the spatially explicit model, the 1.2% of the landscape in the
top categories produces 39.1% of the total TSS. Similarly, for the low-impact areas,
in the land-use model, the bottom 78.9% produces 30.1% of the impact, where in
the spatially explicit model, the bottom 84.3% is responsible for only 14.9% of the
impact.
3. The spatial location of the land use has a significant modifying effect on the land
use as described in the spatially explicit model. For example, in the land-use model
(Figure II.13.2) the town of Gander and the adjacent Gander airport are predicted
to deliver significant suspended solids to the lake (between 900 and 1,200
kg/ha/year). In the spatially explicit model (Figure II.13.3), the hill slope-length
and hill slope-gradient effect (the town and airport are on very flat ground) and
the delivery ratio (which directs the majority of the runoff through bogs, wetlands,
and streams, to Soulis Pond before it eventually reaches the lake at the very eastern
end) reduce the impact of these land uses to less than 13 kg/ha/year.
To determine the importance of landscape features in the spatially explicit model, the
sensitivity of the various factors in the USLE (A = RKLSCP) was considered. In the model,
R and L are constant throughout the catchment. Variation in soil type (K), while important,
can only explain a change in soil loss of six times, given the range of soil types in the
catchment. P-factors (used to modify C) were not used. This left the LS (variables describing slope or landscape location) and C (cover or land-use type) factors to consider.
Slope sensitivity was investigated by dividing the total number of slope values (n)
into five equal categories according to area (n1, n2, . . . n5), where each represents approximate 200 km2 area. Slope values in the catchment range from 0 to 60%: n1 = 0 to 2%; n2 =
3 to 5%; n3 = 6 to 8%; n4 = 9 to 14%; and n5 = 15 to 60%. Using these slope values, together
with the K-value for Gander Series soil (the most common soil type in the catchment),
and the actual range of C-factors (land-use/cover types) in the catchment, Table II.13.4,
catchment sensitivity to slope was generated. In the table, the USLE for each slope category
was solved. Table II.13.4 and the accompanying chart illustrate how sensitive soil loss in
the catchment is to slope. The table and chart are interesting in that they explain the
enormous variation seen in the spatially explicit model. When soils are completely
exposed, the C-factor (no cover) can describe a 200-times increase in erosion over a forested
590
Table II.13.4 Sensitivity of the Spatially Explicit Model to the Variables Describing Slope (Location)
as Compared with Those Describing Cover (Land Use)
Soil Loss (kg/ha)
USLE Factors
(n1n5)
Slopes
1
4
7
11
21
K*
LS
Forested
0.001
Burn
0.007
870
870
870
870
870
0.34
0.34
0.34
0.34
0.34
0.18
0.61
1.50
2.10
2.70
20
66
163
228
293
137
464
1,140
1,596
2,053
Cover Values
Regenerating
Clearcut
0.01
0.02
195
662
1,629
2,281
2,932
391
1,325
3,258
4,561
5,864
Gravel Pit
0.2
3,910
13,249
32,580
45,612
58,644
* Gander soil type (Orthic Humo-Ferric Podzol); 67% of the catchment is in this soil type.
Actual range of cover values used in the spatially explicit model.
60,000
50,000
40,000
30,000
Soil Loss
kg/ha
20,000
10,000
0. 2
1
0.02
4
0.01
7
Slope
11
0.007
C-Values
21
0.001
landscape. For the majority of the landscape types in the catchment, however, the different
cover types explain a range of soil loss from 1 to 20 times. The slope factor consistently
is responsible for a range of soil loss of 3 to 15 times. Thus, in this catchment, slope
(location) is as important to consider as land use. When steep slopes are combined with
very sensitive land uses, the combined effect is significant, as demonstrated by the peak
in the chart of Table II.13.4.
One concern with the spatially explicit model is the lack to data to calibrate both the
C-factor and, in particular, the D-ratio. In order to further evaluate the approach, a map
result considering only the inherent sensitivity component of Eq. II.13.2 was prepared
(see Figure II.13.4). The inherent sensitivity is a measure of landscape sensitivity irrespective of land use. It asks, which areas are more sensitive to any land-use change, by virtue
5,000
Metres
10,000
EDM Environmental Design and Management Limited,
Environment Canada
Department of Environment, Water Resources Division
1996.
F
n
te
if
ke
a
r L
de
n
Ga
100,000 - 999,999
10,000 - 99,999
0 - 9,999
(kg/ha/yr)
Inherent Sensitivity
ro
1,000,000 - 3,999,999
Mile
Gander
5,030
38,768
28,420
5,032
(ha)
Area
ro
Most Inherently Sensitive
Least Inherently Sensitive
Description
Benton
Souli
Pon
J
o
e
's
Figure II.13.4 Map of the Gander Lake catchment illustrating the inherent sensitivity of the catchment to soil loss. (From EDM, Environ. Design and
Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ. Ltd., Watershed management plan for Gander Lake and its catchment, Govt. of
Newfoundland and Labrador together with Environ. Canada, St. Johns, Newfoundland, 1996.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Prepared by:
Prepared for:
CONSIDERING INHERENT LANDSCAPE SENSITIVITY
Appleton
Chapter II.13:
POTENTIAL SOIL LOSS
Glenwood
591
592
of their physical attributes? All the factors in the equation A = RKLS are well established
in the literature. The map result (Figure II.13.4) illustrates the outcome if the soil is left
exposed across the catchment. The method is consistent with that used by Coote et al.
(1992) in preparing the water erosion risk map of the Maritime Provinces of Canada.
Figure II.13.4 is thus an excellent thumb print of the catchment, illustrating those areas
where soil loss will be minimal and those areas where it may be extreme if subjected to
similar land-use and management practices.
The usefulness of each approach was further considered by evaluating the effect of
proposed future land-use changes in the Gander catchment. The proposed changes are as
shown in Table II.13.5. Of note is a large forest clearcut planned on the southern shore of
the lake, recreation development on the steep slopes between the town of Gander and
Gander Lake, and cottage development on the lakeshore in cottage management areas. In
a land-use only approach, the impact of these changes is identical, irrespective of location.
Table II.13.5 illustrates the potential impact of these land uses using the export values as
calibrated in the first model run.
If, however, the land uses are mapped and considered in the second modeling
approach, a different picture arises. Figure II.13.5 is the mapped result of the spatially
explicit model. C-values for both cottage development and the recreation area were
assigned as development. Former clearcut areas were recategorized as regenerating.
The spatially explicit model predicts the planned large clear to have less impact on the
lake than that predicted by the land-use model, even though the clearcut is immediately
adjacent to the lake. The clearcut is proposed for one of the more resilient areas of the
landscape based on the inherent sensitivity map (Figure II.13.4). On the other hand, the
land clearing for recreation on the sensitive steep slopes between the town and the lake,
as well as the very sensitive lands proposed for cottage development may result in a 46%
increase in sediment loading, even though the area affected is less than 100 ha.
Discussion
Both landscape models of landwater interaction were calibrated to the Gander Lake
catchment. Both approaches describe very different landscape areas as the source of
suspended solids and phosphorous to the lake.
A landscape modeling approach where landwater interactions are described by
assigning export values to land uses (the land-use approach) has been adequate to calibrate
hydrodynamic lake models (e.g., Dillon et al., 1986); however, the same landscape modeling approach may be inadequate for watershed planning policy. This is largely because
a catchment total is all that is required for water quality model calibration. The precise
location of pollutant input is not critical, and assuming a relatively uniform terrain across
the catchment, averaging will tend to even out the site-specific landscape differences.
Planning policy, on the other hand, is a very site-specific task. Applying zoning or
other mechanisms to remove land from development or restrict land use has important
economic consequences, both for individual landowners and for the regional economy. In
addition, policy that is too restrictive can lead to low-development densities, which can
increase other environmental concerns such as landscape fragmentation and development
sprawl. For these reasons, planners must be able to clearly defend the specific restrictions
they apply to watersheds. Evaluation of proposed future land-use changes and development proposals is also a very site-specific task. The question that must be answered by
the planner is whether to permit a specific land use in a specific location (for examples,
see Chapters II.10, II.12, II.14, and II.15). Planners are not able to rely upon averaging
Total
Total
Total
Total
TSS for catchment 1995, kg/year

TSS for catchment 2005, kg/year
increase for catchment, kg/year
% increase
Gravel pits
Forest
Blowdown
Bogs
Burn
Unvegetated/bedrock
Recent clearcut
Development
Brush/regenerating
Recreation and cottage clearing
Land Use
1,200
35
200
0
200
100
600
400
900
200
900
E-Values
kg/ha
154
49,838
3,693
7,945
1,510
3,193
534
5,424
438
4,521
N/A
Area
1995
(ha)
7,077,430
184,800
1,744,330
738,600
0
302,000
319,300
320,400
2,169,600
394,200
904,200
TSS
Total, 1995, kg/year
each land use
154
45,785
3,693
7,945
1,510
3,193
534
3,944
438
9,945
109
Area
2005
(ha)
7,526,475
184,800
1,602,475
738,600
0
302,000
319,300
320,400
1,577,600
394,200
1,989,000
98,100
TSS
Total, 2005, kg/year
each land use
449,045
0
141,855
0
0
0
0
0
592,000
0
1,084,800
98,100
Change
20051995
kg/year
6.34%
0.00
8.13
0.00
0.00
0.00
0.00
0.00
27.29
0.00
119.97
9810000.00
% Change
Total
19952005
Chapter II.13:
Table II.13.5 Impact of Proposed Future Land-Use Changes as Predicted by the Land-Use Model

593
Approximate
Suspended
Solids
(kg/ha/yr)
0 - 12.9
13 - 129.9
130 - 1,299.9
1300 - 12,999.9
13,000 - 39,000
% of
Catchment
TSS
0.8
7.5
26.8
21.5
43.4
Bro
ok
Area
(ha)
36,843
22,622
9,874
929
234
Benton
% of
Catchment
Area
45.5
36.6
16.0
1.5
0.4
Soulis
Pond
Figure II.13.5 Map of the Gander Lake catchment illustrating the contribution of suspended solids to the lake from proposed future land uses, as predicted
by the spatially explicit model (Eq. II.13.3). (From EDM, Environ. Design and Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ.
Ltd., Watershed management plan for Gander Lake and its catchment, Govt. of Newfoundland and Labrador together with Environ. Canada, St. Johns,
Newfoundland, 1996.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
10,000
Metres
Environment Canada
Soil Loss
(kg/ha/yr)
0 - 99
100 - 999
1,000 - 9,999
10,000 - 99,999
100,000 - 499,999
Mile
o
Bro
e's
5,000
een
Fift
r Lake
Gande
Gander
Jo
Appleton
594
PREDICTED EXPORT OF TSS FROM PROPOSED FUTURE

LANDUSES IN SPATIALLY EXPLICIT MODEL
Glenwood
Chapter II.13:
595
across the landscape; they must defend site-specific decisions. Planners are also under
enormous pressure to ensure maximum use of the watershed, both for economic and
environmental reasons.
Insufficient data on C-factors and D-ratios are available in the Gander Lake catchment
to fully embrace the spatially explicit model. Of particular concern is the lack of C-factors
for urban landscape types. The model does offer enormous promise, especially for largely
undeveloped catchments. The model results, which suggest highly variable soil loss across
the landscape supports anecdotal evidence in the construction industry. Construction
managers of large road projects frequently claim that there are vast areas of the landscape
where the erosion and sediment control measures are simply not necessary nothing is
eroding anyway, and then there are other areas where no matter what they do, the earth
just keeps falling away and they exceed their construction permits (e.g., Rushton, 1996).
The results of the spatially explicit model are also supported by data reported by Snell
(1985) and discussion in a report by Wall et al. (1978), both indicating that a very small
portion of the fields in an Ontario agricultural landscape were responsible for the majority
of the stream sediment.
Whereas uncertainty surrounds the C-factors and D-ratios in the spatially explicit
model, much less uncertainty surrounds the RKL- and S-factors. The inherent sensitivity
of the landscape (A = RKLS) may therefore be considered a valid description of the
potential of the landscape to erode. For the Gander Lake catchment, the inherent sensitivity
was reclassified (see Figure II.13.6) and adopted as the zoning plan. The zoning plan
preserves 7.6% of the catchment (red areas) and places severe development restrictions
on an additional 9.1% (blue areas) that are very sensitive to any land-use change. Note
that many of the restricted areas are at a distance from the lake edge a very different
spatial configuration than the buffer setback restriction in previous policy.
If soil is the primary source of phosphorous from the catchment, then the spatially
explicit model of landwater interaction must be considered very seriously as an appropriate model of the spatial distribution of the land-use sources of phosphorous and
suspended solids in lake catchments. This modeling approach in support of planning has
been adopted for Newfoundland and Labrador and to date has been defensible to landuse proponents.
Acknowledgments
This study was funded under the CanadaNewfoundland Agreement Respecting Water
Resource Management. The entire project, including water quality sampling, lake modeling, landwater interaction modeling, and recommendations for management, titled
Watershed Management Plan for Gander Lake and Its Catchment, was published as a
government document by the Government of Newfoundland and Labrador together with
Environment Canada in 1996. Haseen Khan, in the Water Resources Division of the
Government of Newfoundland and Labrador, and Joe Arbour at Environment Canada
were both active and supportive throughout the project. The author acknowledges Dr. Ron
Loucks for his careful lake modeling and critical input to the landwater interaction
approach. A special thanks to all of the team at EDM: Jeff Pinhey for the water balance
and input, advice, and support; Nancy Griffiths for her careful research; and Paula Dykstra, who ran the models in the geographic information system (GIS). A very special thanks
to Professor Michael Binford, now at the University of Florida, who first applied the
spatially explicit approach in Acadia National Park, and who continues to find time and
energy to support former students as they apply his teachings.
ess
k
Broo
t's
r
Mile
Area
(ha)
33,920
24,811
6,435
5,336
% of
Catchment Area
48.1
35.2
9.1
7.6
ZONING PLAN
Restricted Use Zones
ok
Bro
Gander
Jo
e's
Restricted Use
L - Low
M - Medium
H - High
R - Restricted
Bro
ok
Bento
Souli
Pond
Figure II.13.6 Zoning plan for the Gander Lake catchment illustrating areas restricted to development. Note the difference between this map and a standard
lakeshore buffer approach. (From EDM, Environ. Design and Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ. Ltd., Watershed
management plan for Gander Lake and its catchment, Govt. of Newfoundland and Labrador together with Environ. Canada, St. Johns, Newfoundland,
1996.) (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
een
Fift
r Lake
Gande
10,000
Metres
Environment Canada
5,000
Hu
n
Appleton
596
Br
e's
Jo
el
Car
de
Fe
e
ook
Glenwood
Chapter II.13:
597
Literature cited
Arbour, J., Personal communication, Water Resour. Div., Conservation Branch, Environ. Canada,
1996.
Binford, M., Spatially explicit models of landwater interactions for evaluating development proposals, unpublished paper, 1989.
Binford, M., Personal communication, Dept. of Landscape Architect., Grad. Sch. of Design, Harvard
Univ., 1995.
Borman, H.F. and Likens, G.E., Pattern and Process in a Forested Ecosystem, Springer-Verlag, New
York, 1979.
Cooper, J.R. et al., Riparian areas as filters for agricultural sediment, Soil Sci. Soc. of Amer. J., 51, 416,
1987.
Coote, D.R. et al., Water Erosion Risk: Maritime Provinces, Agriculture Canada, Ottawa, ON, 1992.
The Dept. of Environ. and Lands Act (DELA), Newfoundland and Labrador, 1989.
DELA, Newfoundland and Labrador, 1991.
Dillon, P.J. et al., Lakeshore capacity study, trophic status, Res. and Special Proj. Branch, ON Min.
of Environ., Ontario, 1986.
Dissmeyer, G.E. and Foster, G.R., Estimating the cover management factor (C) in the universal soil
loss equation for forest conditions, J. Soil and Water Conserv., 36, 235, 1981.
Dunne, T. and Leopold, L., Water in Environmental Planning, W.H. Freeman and Co., New York, 1978.
EDM, Environ. Design and Manage. Ltd., Land use plan for the Big Cove First Nation, 2000.
EDM, Environ. Design and Manage. Ltd., The Western Common land use plan, 1999.
EDM, Environ. Design and Manage. Ltd., R.H. Loucks Oceanol. Ltd., and Jacques Whitford Environ.
Ltd., Watershed management plan for Gander Lake and its catchment, Govt. of Newfoundland and Labrador together with Environ. Canada, St. Johns, Newfoundland, 1996.
Hickman, R.E., Loads of Suspended Sediment and Nutrients for Local Nonpoint Sources to the Tidal Potomac
River and Estuary, Maryland and Virginia, 197981 Water Years, U.S. Geol. Surv., 1987.
Loucks, R., The First Lake total watershed management project, phase I: neighborhood stewardship,
1993.
Leopold, L., A View of the River, Harvard Univ. Press, Cambridge, MA, 1994.
Lowrance, R., McIntyre, S., and Lance, C., Erosion and deposition in a field/forest system estimated
using Cesium-137 activity, J. Soil and Water Conserv., 43, 195, 1988.
Mellerowicz, K.T. et al., Soil conservation planning at the watershed level using the universal soil
loss equation with GIS and microcomputer technologies: A case study, J. Soil and Water
Conserva., 49, MarchApr., 194, 1994.
Rees, H.W., Personal communication, Land Resource Unit Head, Centre for Land and Biol. Resour.
Res., Res. Branch, Agricul. Canada, 1996.
Riley, M.J., MinLake: External memorandum no. 13, Dynamic Lake Water Quality Simulation Program, St. Anthony Falls Hydraulic Lab., Univ. of MN, Minneapolis, 1988.
Rushton, D., Personal communication, Nova Scotia Dept. of Transport. and Commun., 1996.
Snell, E.A., Regional targeting of potential soil erosion and nonpoint-source sediment loading, J.
Soil and Water Conserv., 40, Nov.Dec., 520, 1985.
Toth, R.E., Hydrologic and riparian systems: The foundation network for landscape planning, Intl.
Conf. on Landscape Planning, Univ. of Hannover, Germany, 1990.
Transport Canada Gander Intl. Airport, Stormwater quality monitoring results, Transport Canada,
Airports Group, 1995.
U.S. EPA, A Hydrodynamic and Water Quality Model, WASP4 (water analysis simulation program),
Environ. Res. Lab., Athens, GA, 1988.
Vaughan Engineering, Ltd., Shubenacadie Lakes planning/pollution control study, Halifax, NS,
1993.
Wall, G.J., Dickinson, W.T., and Greuel, J., Rainfall erosion indices for Canada east of the Rocky
Mountains, Canad. J. Sci. 63, May, 271, 1983.
Wall, G.J., van Vliet, L.J.P., and Dickinson, W.T., Contribution of sediments to the Great Lakes from
agricultural activities in Ontario, Intl. Joint Comm., Windsor, ON, 1978.
598
Wells, R.E. and Heringa, P.K., Soil survey of the Gander-Gambo Area, Newfoundland, Report no.
1: Newfoundland soil survey, Res. Branch, Canada Dept. of Agricul. and Agricultural and
Rural Develop. Act, 1972.
Wischmeier, W.H. and Smith, D.D., Predicting Rainfall Erosion Losses: A Guide to Conservation Planning,
Agriculture Handbook no. 537, U.S. Dept. Agricul., Washington, DC, 1978.
Chapter II.13:
599
Response
Linking land use to landscapes for water quality protection
In this chapter, Margot Cantwell illuminates what has been a somewhat odd dichotomy
in water quality land-use planning. On the one hand, many models have relied on averaged estimates from landscape and land-use classifications in their predictions of development threats to lakes; on the other hand, planning policy is rooted in a strong sitespecific tradition. Cantwells comparative analysis provides a much needed assessment
of the strengths and shortcomings of adopting these differing approaches for water sensitive planning. An important take-away message from this chapter is demonstration of
the utility of incorporating science into the process of land-use decision making for lake
protection. In this respect and similar to John Felkner and Michael Binfords work in
Chapter II.10, and Robert France and Philip Crauls work in Chapter I.7, Cantwells study
demonstrates the need to pay careful attention to soils when engaging in water sensitive
planning and design; in other words, literally ground-truthing projects.
The GIS-based process Cantwell uses is very much in the spirit of the bottom-up
method championed by Jeffrey Schloss in Chapter II.12. Cantwell ably demonstrates the
power of GIS mapping when combined with landscape rate equation models. All too
frequently, GIS analysis is used to produce nothing more than attractive, colorful maps
that do no more than inform about the blatantly obvious. What this particular chapter
demonstrates is the evolution of functional GIS map-modeling. No longer is GIS mapping
regarded as the end in itself; instead is it used as the tool it was always intended for, as
a means to spatially link landscape processes to environmental predictions. This is a very
important lesson and one that Cantwell uses to support the contention raised by Neil
Hutchinson in Chapter II.17. That is the truly important questions involved in water
sensitive planning should be more concerned with where development resides on the land
than perhaps simply how much of it may be there in the first place. In a sense, this is the
parallel argument to that put forward by Larry Coffman in Chapter I.5 in terms of considering the benefits of using spatially explicit low-impact development designs over
averaged percentage imperviousness to effectively manage stormwater runoff.
chapter II.14
Spatial investigation of applying

Ontarios timber management
guidelines: GIS analysis for riparian
areas of concern
Robert France, John S. Felkner, Michael Flaxman, and Robert Rempel
Abstract
Riparian zones are not only ecotones situated between adjacent ecological systems, they
are also the areas on the landscape where conflicting interests about natural resource
management abut. For many regions of the boreal forest in Ontario, the timber industry
is the major employer and generator of revenue, whereas sport fishing is the major
commercial recreational pursuit. In order to ensure protection of the latter, the former is
regulated by precluding logging from occurring within near-shore riparian areas. Despite
site-specific studies on the extent of ecotonal coupling between boreal forests and water
bodies as affected by clearcutting, there is little understanding about what the effects of
establishing such riparian areas of concern might be for the timber industry on a
landscape scale. Despite forestry officials publicly stating fears that as much as 30% of the
terrestrial land area would be included in such buffer strips, our GIS analysis for a region
of northwestern Ontario showed this amount to be closer to 10% of the surface area. This
in turn was estimated by our analysis to correspond to a reduction of about 8% in the
amount of merchantable timber that could be harvested.
In many parts of Ontario merchantable timber occurs in areas adjacent to aquatic environments (shoreline areas). Many of these aquatic
environments, including wetlands, lakes and streams, provide or
can potentially provide valuable habitat for fish.
Timber management guidelines,
Ontario Ministry of Natural Resources (1988)
1-56670-562-2/02/$0.00+$1.50
601
602
Introduction
Natural resource management may often involve tradeoffs made between the economic
viability of resource extraction industries and the ecological integrity of the environment
from which the resources are obtained. Perhaps nowhere is this dichotomy felt more
strongly than in the regulation of timber harvesting near water bodies. This is natural
because there is widespread recognition that the landwater interface is of critical importance for environmentally sensitive planning (Chapters II.1 to II.4). Specifically, after
reviewing the copious scientific information demonstrating the potential for deleterious
effects of clearcutting upon aquatic biota, land-use planners have agreed that there is a
need to establish riparian buffer strips around rivers and lakes in order to protect them
from forestry operations (Brewin and Monita, 1998). This is typically the extent of the
agreement between ecologists, planners, and foresters. Contention arises in trying to
determine the precise size, or width, that is deemed necessary to enable those buffer strips
to perform their job effectively.
Environmentalists argue that when it comes to protective buffer strips, bigger is better,
especially in the absence of absolutely unequivocal scientific data about the possible longterm effects of near-shore logging. Forestry companies counter by questioning the implicit
and often unsupported assumptions that may characterize the environmentalists platform, at the same time as decrying the potential economic losses due to timber reserves
being locked up in such buffer strips. Arguments in both cases are substantially weakened by lack of actual landscape data upon which to engage in an intelligent dialectic.
The purpose of the present study was to provide a preliminary spatial analysis of forest
patterns in northwestern Ontario as an aid toward facilitating more meaningful discussions for future land-use planning decisions in this region of the boreal forest.
At present, an alarming disjunction exists between the scale at which scientific investigations of the effects of riparian forestry have been conducted and the scale at which
industrial forestry operations are being implemented. For example, although studies have
explored the site-specific consequences of riparian deforestation on litter production,
erosion, wind, solar energy, as well as nutrient and food resources (Chapter II.16; France,
1997a; France and Peters, 1995; France et al., 1998; Steedman et al., 1998; Steedman and
France, 2000), the wider consequences of what such effects may be at the landscape scale
are not well known. Forestry officials, bemoaning the, in their minds, overly stringent
guidelines for riparian areas of concern, have provided unsupported statistics to the press
(e.g., The Communications, Energy and Paperworks Union argue that more than 30% of
the land in Northern Ontario is protected when buffer zones around lakes, rivers and
streams are included in the calculation [of protected land] (Mackie, 1998)).
The present research was directed toward one simple goal to examine the validity
of the claim made by the forestry industry that 30% of the land area in northern Ontario
is contained within the riparian areas of concern under the present timber management
guidelines (OMNR, 1988).
Ontario forestry
Ontario contains 1.5% of the worlds forests, a major portion of which is classified as
being boreal and is, therefore, representative of the largest forest type on the planet.
The forestry industry in Ontario contributes a considerable amount of revenue to the
economy and employs thousands. Simultaneously, in certain regions of the boreal forest,
sport fishing represents the second-largest resource industry. Because of a recent history
of contentious (and even litigious) debate between environmental groups and the timber
industry and government resource managers, issues of land-use planning and forest
Chapter II.14:
Spatial investigation of applying Ontarios timber management guidelines
603
and fisheries management have come into the spotlight. In particular, the response of
forests to natural and anthropogenic processes and their sustainable management have
both received much needed attention (Perera et al., 2000; OMNR, 2000).
Special focus has been directed toward understanding the possible interactions between
riparian timber removal and the health of aquatic systems (e.g., France, 1997b; Steedman,
2000; Steedman et al., 2001). Ontarios recommended guidelines (OMNR, 1988) for riparian
buffer strip widths (termed areas of concern) to protect lakes containing commercially
important lake trout are based on the upper values suggested in relation to surface slope
and inferred transport distance of eroded soil (discussed in Chapter II.16). These widths
are 30 m for slopes of 1 to 15%, 50 m for slopes of 16 to 30%, 70 m for slopes of 31 to 45%,
and 90 m for slopes of 46 to 60%.
Study area
The study area is located in northwestern Ontario, 150 km east of the Experimental Lakes
Area and near the OMNRs Coldwater Lakes Field Station (Steedman et al., 2001;
Figure II.14.1). The area is in the Lake of the Woods Plains Ecoregion and is transitional
between the northern boreal and Great Lakes/St. Lawrence forests. Fire disturbance plays
an important role in structuring vegetation communities here. The region is typified by
small headwater lakes situated in catchments characterized by shallow organic and coarsetextured mineral soils with abundant bedrock outcrops and topographic relief generally
not more than about 60 m. Riparian forests are predominantly composed of black spruce
and jack pine, with trembling aspen and paper birch prevalent in backshore areas (France
et al., 1998). The present study area is within a region previously investigated through
GIS analysis for understanding the response of forest vertebrates to landscape-level patterns in forest disturbance (Rempel et al., 1997; Voigt et al., 2000).
Methods
Given existing riparian buffer guidelines, as well as forestry industry claims regarding
the proportion of the total land area (or the total forested area) existing under current
buffer guidelines, the methodology described here had as its goal the physical measurement of the approximate proportion of total land area protected by current buffer guidelines, and the obtaining of the approximate proportion of that area actually wooded, using
current land-cover data in a 315-km2 study area in northern Ontario (see Figure II.14.1).
To achieve this goal, land-cover, topographic, and hydrologic data sets were geo-rectified
and overlaid in a geographic information system (GIS) and a modeling approach was
devised.
The basic structure of the approach was initially conceptualized as follows:
1. Measure the slope of all land bordering water bodies (streams, lakes, and wetlands).
2. Use a GIS buffering function to accurately buffer all water bodies according to
existing riparian timber buffering guidelines to the distance specified by those
guidelines as a function of slope, thereby obtaining the total protected buffered
area for the study site.
3. Examine both the composition (hardwood/softwood) and the merchantability
status (regular production, reserve production, age status, etc.) of timber stands
on those riparian buffer areas to estimate the percentage of merchantable timber
in the total protected buffer areas.
604
Figure II.14.1 Context map showing location of northwestern Ontario and region of study focus.
Once these goals were achieved, further analysis could be performed to derive the
proportions and percentages of softwoods, hardwoods, and merchantable stand types in
relation to various slope and water body edge categories. The final modeling approach
used in this analysis is depicted in Figure II.14.2. Figure II.14.3 displays a larger (595.6-km2)
context region, showing lakes, streams, and wetlands. The smaller (315.0-km2) focus study
area is displayed in Figure II.14.4, and appears to be similar in terms of shoreline density
and land-cover composition to surrounding areas
For the analysis, topographic data, and land features (lakes, streams, wetlands, and their
boundaries) were obtained from the Ontario Digital Topographic Database (DTDB) as part
of the Ontario Basic Mapping (OBM) Program for the Province of Ontario, constructed by
Chapter II.14:
vector
stream
network
elevation
lattice
Digital
Elevation
Model (DEM)
Grid - 25 m
Wetland
Polygons
Vector
Coverage
Hard and
Softwood
Polygon
Coverage
Stream
Network
Grid
Lake
Grid
Wetland
Grid
Hard/
Softwood
Grid
Stream
Buffer
Lake
Buffer
Wetland
Buffer
Slope
values for
Stream
Buffer
Slope
values for
Lake
Buffer
Slope
values for
Wetland
Buffer
Stream
Area of
Concern
Lake
Area of
Concern
Wetland
Area of
Concern
Additional spread
from buffer cells
with slopes > 15%
based on guidelines
Areas of Concern
are combined
with this overlap
priority:
1. lakes
2. streams
3. wetlands
Lake
Polygons
Vector
Coverage
Slope Grid:
Percent of
Rise
Forest
Resource
Inventory
Database
Ontario
Digital
Topographic
Database
Base Data Layers
8
Total
Area of
Concern
605
Final Grid:
forest in
total area
of concern
and slopes
Numbered GIS
Operations:
1. Arc: Topogrid
2. Arc: Reselect
3. Arc: Slope
4. Vector to Grid
5. 1 Pixel Buffer
6. Intersect With Slope
7. Spread Operation
8. Combine Grids
9. Intersect Operation
Extraction of all ha amounts:

ha of hard and softwoods in total
area of concern, individual area of
concern, and based on slope.
Figure II.14.2 Flow chart of the methodology used in the spatial analysis of riparian areas of concern.
the Provincial Mapping Office of the Natural Resources Information Management Branch,
under the Ontario Ministry of Natural Resources (OBM, 1996). Geographic data on forest
cover, including the timber merchantability condition and the hardwood/softwood species composition of individual tree stands, were obtained from the Forest Resource Inventory (FRI) (R. Rempel, unpubl. data).
606
Figure II.14.3 Map of larger (595.6 km2) study area showing landform and water body designations.
Digital data from both sources were available georectified to a Universal Transverse
Mercator (UTM) projection system, and all data were geo-registered in a GIS. This data
allowed the calculation of basic surface-area proportions and percentages including those
of water to land; lakes versus wetlands versus streams versus total area; total landwater
interface amounts; amounts of the total buffer areas versus total study area; and breakdowns for percentages of varying slope amounts on landwater interface areas on the
315-km2 study area, as discussed next.
Both the OMB and the FRI data sources provided topographic, waterbody and timber
stand location and composition data in GIS vector format. The OMB topographic data
were provided in an irregular lattice of mass points format, with point elevation values
obtained at approximately 100-m intervals, and then additional elevation points on the
crowns of hills and bottoms of depressions, as well as along the edges of water bodies.
Because the derivation of topographic slope values was crucial for the model, the first
step in the analysis was the creation of an accurate digital elevation model (DEM). The
DEM was created using the topogrid function in ArcInfo software that is based on the
ANUDEM program developed by Michael Hutchinson (Hutchinson, 1988, 1989, 1993,
1996; Hutchinson and Dowling, 1991). The method uses an iterative finite-difference
interpolation technique which is essentially a discretized, thin-plate spline approach
(Wahba, 1990), where the roughness penalty has been modified to allow the fitted DEM
to follow abrupt changes in terrain, such as streams and ridges. The procedure is specifically designed to consider hydrologic drainage features, which are the primary erosive
Chapter II.14:
607
Figure II.14.4 Map of smaller (315.0 km2) study area showing landform and water body designations, forest types, and riparian areas of concern based on slope.
and shaping force on a landscape. Consequently, vector stream network data from OMB
were used as an input into the DEM creation program. Furthermore, the program assumes
that all unidentified sinks are errors, because sinks are generally rare in natural landscapes
(Goodchild and Mark, 1987). The DEMs were generated with raster grid cells of 25 m,
608
because this was deemed an appropriate level of spatial precision given the lattice elevation point spacing.
Because the analysis depended on the summing of areas of varying slope values all
water body and forest stand vector data were also converted to 25-m raster grid layers
to facilitate comparison with the grid of slope values derived from the DEM. Stream
networks, lakes, and wetland polygons were extracted from the OMB data and converted
to separate raster grids with 25-m cell sizes.
Tree stand polygons of standing merchantable production timber were selected from
the FRI data and their respective hardwood/softwood compositions were considered for
categorization as either predominantly hardwood or predominantly softwood. These stand
polygons, once thus classified, were gridded at the 25-m cell resolution.
Next, it was necessary to determine the approximate slope at the edges of all riparian
bodies, as the OMB buffer guidelines fluctuate depending on slope. To accomplish this, a
one-pixel buffer was created around water bodies in the stream, lake, and wetland grids,
respectively. These buffers were independently combined with the slope grid derived from
the DEM to determine the slope values for each buffer pixel (effectively the slope at the
waters edge). For every pixel with a derived slope value greater than 15%, a further
spread function was performed from those high slope cells according to the specified
guidelines (see above) for each of the three buffer grids, respectively. This final spread
function was performed as a linear distance, according to the guidelines (30 m, 50 m, 70 m,
or 90 m) across the DEM that is, considering topography rather than as the crow flies
distance (although with such small spread distances, no spread was greater than three or
four pixels, and thus the topographic data were unlikely to greatly influence which cells
were included in the spread). This spread provided the appropriate estimated buffer areas
of concern for streams, lakes, and wetlands, respectively. To obtain the total buffer widths
for the study area, the buffers for lakes, streams, and wetlands were derived in separate
steps and then were combined into a single grid.
The final step was the intersecting of the hardwood/softwood grid for the standing
merchantable production timber stands with the total areas of concern grid. The output
from this intersection (Figures II.14.4 and II.14.5) allowed the calculation of the total area
and percentage of merchantable hardwoods and softwoods in official riparian buffer areas,
with specific breakdowns for softwoods and hardwoods.
The accuracy of this analysis depends directly, it should be noted, on the accuracy of
the input data layers. All input data sources the elevation lattice points and the landcover and forest data were obtained by Canadian government environmental management agencies in accordance with established data collection procedures. Nonetheless,
potential error in the input data sources could affect the final analysis. This degree of error
in the final analysis is dependent directly upon the following:
1. The degree of accuracy of the vectorized land-cover data lake, stream, wetland,
and forest stand polygons
2. The degree of accuracy of the elevation mass points
3. The degree of accuracy of the DEM interpolation method
Furthermore, accuracy of final measurements is certainly reduced by the process of
converting the original vector data into 25-m raster cells, and then measuring riparian
buffer widths as spreads in increments of 25-m raster cells, because of the generalization
of surface area that occurs with pixelization. Future research will attempt to overcome
this potential measurement error by deriving buffer widths through uniform vector
spreads.
Chapter II.14:
609
Figure II.14.5 Close-up of study area showing application of riparian areas of concern in relation to
Ontarios Timber Management Guidelines.
Results and discussion

In the larger 595.6-km2 context area (Figure II.14.3), 76.7% of the total surface area is land
and 23.3% is collectively occupied by water bodies (Figure II.14.6). Lakes comprise 17.8%
of the area, with streams and wetlands respectively subsuming 1.8 and 3.7% of the total
610
Land
Wetland
Stream
Lake
Figure II.14.6 Proportion of land and water in the study region of northwestern Ontario.
Potential
Harvestable
Land
Wetland AOC
River AOC
Lake AOC
Figure II.14.7 Proportion of land occurring within the riparian areas of concern for the study region
of northwestern Ontario. AOC = areas of concern.
area. Within the study region, there are a total of 32.3 km of landwater interface: 15.4 km
for lake-riparian contact, 10.5 km for stream-riparian contact, and 6.4 km for wetlandriparian contact. Spatially, for the terrestrial surface area, only 6.7% of the total landscape
is contained within the riparian areas of concern (Figure II.14.7), half this amount being
associated with lake riparian zones. Reciprocally, the amount of land that is potentially
available for timber extraction is 93.3% of the surface area.
Clearly, the forestry industrys contention that 30 per cent of the land in Northern
Ontario is protected when buffer zones around lakes, rivers and streams are included in
the calculation, is untenable for this characteristic subregion study area of northwestern
Ontario. Such a large value occurs only if the proportion of the landscape covered by
water bodies is added into the calculation. For the riparian areas of concern, from 97 to
100% (depending on the type of water body) of the riparian land has surface slopes of
less than 15%, meaning that the minimum riparian buffer strip widths of 30 m were almost
exclusively applied in this analysis.
It is, of course, not really the land that is of most interest to the forestry industry, but
rather that which grows atop it. For the 315.0-km2 area selected for detailed analysis, 26.2%
of the total area is occupied by water bodies, and 11.9% of the total area is located within
the riparian areas of concern. Of the remaining land surface area, 7.3% does not contain
trees of sufficient merchantability, with the majority of this (4.6%) occurring within upland
areas unassociated with riparian zones (Figure II.14.8). For the land that does contain
merchantable trees (both upland and riparian), the majority of these (87%) are softwoods.
Our analysis for this particular region of northwestern Ontario found that only 8.3%
of the total realized merchantable timber reserve was collectively situated within the
designated riparian buffer areas of concern: 6.1% for lakes (comprised of 80% softwoods),
1.4% for streams (comprised of 86% softwoods), and 0.9% for wetlands (comprised of 88%
Chapter II.14:
611
Realized
Upland
Harvestable
Timber
No Trees Upland
No Trees AOC
Wetlands AOC
Streams AOC
Lakes AOC
Figure II.14.8 Proportion of merchantable timber reserves occurring within the riparian areas of
concern in the study region. AOC = areas of concern.
softwoods). Again, this is far less than the 30% reduction in timber harvest (inferred from
an assumed 1:1 correspondence between land-surface area and amount of merchantable
timber) supposed by some industry officials to prevail following continued adoption of
the riparian buffer widths specified in the timber management guidelines. Consequently,
it remains to be proven that in some cases, posited by the OMNR (1988).
The present analysis was designed to provide a first-stage assessment of potential
economic tradeoffs between multiple-use resourcism that might exist in this region of
Ontario. One obvious shortcoming to our present analysis pertains to appraisal of merchantable timber reserves based solely on GIS designations of forest typology classifications with no consideration given to estimations of actual timber yields. We plan next to
undertake a more detailed GIS-economic analysis of how projected timber yields might
be influenced through compliance to the timber management guidelines, as well as to
how alternate scenarios of increased or reduced stringency in prescribing buffer strip
widths might affect financial returns.
Acknowledgments
This study was supported by the Milton Fund and the GSD Deans Fund, both of Harvard
University, and by the Ontario Ministry of Natural Resources. The authors thank Paul
Cote for his helpful suggestions throughout the study.
Literature cited
Brewin, M.K. and Monita, D.M.A., Eds., Forest-fish conference: Land management practices affecting
aquatic ecosystems, Can. For. Serv. Info. Rep. NOR-X-356, 1998.
France, R.L., Land-water linkages: Influences of riparian deforestation on lake thermocline depth
and possible consequences for cold stenotherms, Can. J. Fish. Aquat. Sci., 54, 1299, 1997a.
France, R.L., Stable carbon and nitrogen isotopic evidence for ecotonal coupling between boreal
forests and fishes, Ecol. Freshw. Fish, 6, 78, 1997b.
France, R.L. and Peters, R.H., Predictive model of the effects on lake metabolism of decreased
airborne litterfall through riparian deforestation, Conserv. Biol., 9, 1578, 1995.
France, R., Peters, R., and McCabe, L., Spatial relationships among boreal riparian trees, litterfall
and soil erosion potential with reference to buffer strip management and coldwater fisheries,
Ann. Bot. Fenn., 35, 1, 1998.
612
Goodchild, M.F. and Mark, D.M., The fractal nature of geographic phenomena, Annal. Assoc. Am.
Geograph., 77, 265, 1987.
Hutchinson, M.F., Calculation of hydrologically sound digital elevation models, presented at the
3rd Intl. Symposium on Spatial Data Handling, Columbus, OH, 1988.
Hutchinson, M.F., A new procedure for gridding elevation and stream line data with automatic
removal of spurious pits, J. Hydrol., 106, 211, 1989.
Hutchinson, M.F., Development of a continent-wide DEM with applications to terrain and climate
analysis, in Environmental Modeling with GIS, Oxford Univ. Press, New York, 1993.
Hutchinson, M.F., A locally adaptive approach to the interpolation of digital elevation models,
presented at the 3rd Intl. Conf./Workshop on Integrating GIS and Environ. Modeling, Santa
Fe, NM, 1996.
Hutchinson, M.F. and Dowling, T.I., A continental hydrological assessment of a new grid-based
digital elevation model of Australia, Hydrol. Proce., 5, 45, 1991.
Mackie, R., Loggers also care about the wilderness, says union leader, The Globe and Mail National
News, p. A7B, Dec. 1, 1998.
OBM, Ontario Digital Topographic Database: A Guide for Users, ON Min. of Natur. Resourc., Toronto,
1996.
Ontario Min. of Natur. Resour. (OMNR), Timber management guidelines for the protection of fish
habitat, Toronto, 1988.
OMNR, Forest sustainability beyond 2000 conf., www.forconfor2000.org, 2000.
Perera, A.H., Euler, D.L., and Thompson, I.D., Eds., Ecology of a Managed Terrestrial Landscape. Patterns
and Processes of Forest Landscapes in Ontario, UBC Press, Vancouver, 2000.
Rempel, R.S. et al., Timber-management and natural-disturbance effects on moose habitat: Landscape evaluation, J. Wild. Manage., 6, 517, 1997.
Steedman, R.J., Effects of experimental clearcut logging on water quality in three small boreal forest
lake trout (Salvelinus namaycush) lakes, Can. J. Fish. Aquat. Sci., 35, 190, 2000.
Steedman, R.J. and France, R.L., Origin and transport of aeolian sediment from new clearcuts into
boreal lakes, northwestern Ontario, Canada, Wat. Air Soil Pollut., 122, 139, 2000.
Steedman, R.J. et al., Effects of riparian deforestation on littoral water temperatures in small boreal
forest lakes, Boreal Environ. Res., 3, 161, 1998.
Steedman, R.J., Kushneriuk, R.S., and France, R.L., Initial response to catchment deforestation of
small boreal forest lakes in northwestern Ontario, Canada. Verh. Intl. Verein. Limnol., in press,
2001.
Voigt, D.R. et al., Forest vertebrate responses to landscape-level changes in Ontario, in Perera, A.H.,
Euler, D.L., and Thompson, I.D., Eds., Ecology of a Managed Terrestrial Landscape. Patterns and
Processes of Forest Landscapes in Ontario, UBC Press, Vancouver, 2000.
Wahba, G., Spline Models for Observational Data, Soc. Independ. Appl. Math, Philadelphia, 1990.
Watson, D.F. and Philip, G.M., A refinement of inverse distance weighted interpolation, Geo-Process,
2, 315, 1985.
Chapter II.14:
613
Response
Size matters
As Frank Mitchell outlined in Chapter II.1, shoreline buffers provide a host of beneficial
values to lakes, wildlife, as well as humans, especially if planned on the scale of watersheds
rather than on fragmented locals. The visualized display of spatial data through GIS
analysis offers an extremely powerful tool to decision makers toward informed buffer
strip management in a watershed context, as Jeffrey Schloss points out in Chapter II.12.
Such approaches can be particularly useful when examining previously held beliefs
founded solely on untested assumptions rather than on empirical observations.
The case study described by Robert France, John Felkner, Michael Flaxman, and Robert
Rempel demonstrates the utility of GIS analysis in testing assumed beliefs involved in
contentious environmental management. That the forest resource industry had voiced
their beliefs so strongly and assuredly, and that these in turn had been picked up and
circulated to a wide audience in the national press, provides an interesting lesson in the
needed role of science in the rhetoric of environmental debates.
The scale of timber resource extraction in this region of the boreal forest is truly
immense. Small differences in the size of protective riparian buffer strips could therefore
have immense repercussions on the total amount of forest cleared and on the total amount
of revenue gleamed. This chapter provides a first investigation of some very basic questions often required for effective land-use management and water sensitive planning: What
is the area of water on the landscape? Are rivers, lakes, or wetlands predominant? What
is the extent of landwater interface in the study region? The GIS approach used to answer
these questions in this chapter is really pretty basic in scope. As the GIS analysis found,
however, careful attention was needed toward examining the methods used in order to
obtain as precise data as possible before these questions could be addressed. The lesson
here for planning professionals is an important one even the simplest questions may
be difficult to answer in a defensible manner (especially if part of a contentious debate)
given limitations in the available data.
For water sensitive planning, this study underscores the importance of scaling up sitespecific results to a regional perspective larger than a single watershed. As France et al.
mention, however, it is of course, not really the land that is of most interest to the forestry
industry, but rather that which grows atop it. In other words, what are the economic
implications of these spatial results? Their findings, admittedly just a preliminary examination of the question, suggest that the losses of timber revenues may not be as significant
as imagined by the logging industry. Further, much more precise, research is suggested
by this analysis. As a first cut at the problem, however, this chapter demonstrates a horizon
direction for land-use planners; namely, the intersection of scientific and economic data.
The role of GIS in bridging between these two fields of resource management will become
increasingly important as an avenue of study.
chapter II.15
Aquifer recharge management model:

Evaluating the impacts of urban
development on groundwater resources
(Galilee, Israel)
Amir Mueller, Robert France, and Carl Steinitz
Abstract
Generating different alternatives for regional development plans is insufficient without a
process for evaluating the cumulative impact of each alternative. Therefore, it is necessary
to evaluate the impacts of each plan on the full spectrum of socioeconomic and environmental issues. In this chapter, we develop a model that focuses on groundwater as a
keystone indicator for the cumulative impacts of urban development. The aquifer recharge
management model (ARMM) uses a geographic information system (GIS) of overlays
combined with surface features to evaluate the impacts of different land uses on the aquifer
underlying a region of Galilee, Israel. The model uses imperviousness as a means for
measuring the of impacts of urban development on aquifer recharge areas. The increase in
the total area of impervious surfaces associated with urban development heightens the intensity of floods and increases pollutant concentration from point and nonpoint sources. Polluted
urban runoff degrades the quality of surface water in streams, rivers, and lakes. Furthermore,
it contributes to the pollution of groundwater when it is recharged into the aquifer.
Introduction
One of the challenges in regional and landscape planning is in adequately comparing the
impacts of a wide range of feasible alternatives. Therefore, it is necessary to evaluate the
impacts of each plan on the full spectrum of socioeconomic and environmental issues.
The purpose of this study was to develop a tool that models the impacts of urbanization
on aquifer recharge areas. The ARMM focuses on groundwater as a keystone indicator
for the cumulative impacts of urban development.
The threat to groundwater resources is exacerbated by the increase in the amount of
impervious surfaces associated with urban development (Chapter I.1). Increases in the
intensity of floods and pollution episodes are directly related to increases in the total
amount of impervious surfaces. Because water resources are easily degraded by pollution
from anthropogenic sources, it is important to develop strategies and practices that integrate
land-use decisions with a comprehensive water management plan (Chapters II.7, II.8, II.11,
II.13, and II.17).
1-56670-562-2/02/$0.00+$1.50
615
616
Degradation of groundwater quality

Degradation of groundwater occurs when pollutants are introduced into the aquifer from
diffuse sources such as agricultural and urban runoff, leaching of pollutants from landfills
and industry, and hydrological deficits that alter groundwater flow (Carman et al., 1997;
Berg et al., 1999). In short, anthropogenic impacts on aquifers result in the overall degradation of groundwater quality and the capacity of the aquifer to function effectively
(Goldenberg and Melloul, 1994). Inappropriate surface-water management may contribute
to the accumulation of pollutants in the filtration and saturated zones of an aquifer
(Goldenberg and Melloul, 1994; Kuylenstierna et al., 1997). Reducing or even eliminating
water pollution as well as increasing long-term viable water resources requires altering
current water-use management and land-use policies.
To understand how land-use policies may improve groundwater quality, and perhaps
increase groundwater reserves, it is important to view groundwater as an integral part of
the hydrologic cycle. When water precipitates over land, some of it is lost through evapotranspiration, some flows above ground as runoff in the form of rivers and streams, and
the remainder infiltrates through the surface into the ground. When water reaches an
impermeable layer such as clay or marl, it accumulates in the saturated zone. When the
saturated layer is permeable enough to allow the storage and transport of significant
amounts of water, an aquifer is formed (Gordon, 1984; The Conservation Foundation,
1987). Most aquifer recharge results from precipitation percolating through the ground
surface, seepage from streams, and surface waters.
Existing land uses and land cover as well as the permeability of soils and overlying
geology are important factors in identifying aquifer recharge areas. Recharge in undeveloped areas is critical since it tends to be relatively free of pollutants and can be easily
monitored and controlled. It is important to note that not all land-use activities pose the
same threat to groundwater resources. Threats depend on the inherent hazards associated
with each activity, where the activity takes place, and the number of people it affects
(Gordon, 1984; Jaffe and DiNovo, 1987; Carman et al., 1997; Berg et al., 1999).
Although groundwater is less frequently polluted, once it is polluted, it is much slower
to recover from pollution episodes that would be considered mild for surface water in
lakes and streams. Little dilution or attenuation of pollutants occur in groundwater due
to the slow rate of lateral movement through the saturated zone. Furthermore, pollutants
in the ground may persist for years, and even decades (Jaffe and DiNovo, 1987).
The duration, type, and intensity of pollution will determine the degree of risk to
groundwater quality and quantity. Typically, groundwater pollutants are introduced from
landfill, sewage treatment facilities, and agricultural and industrial runoff (Jaffe and
DiNovo, 1987; Issar, 1993; Goldenberg and Melloul, 1994; Berg et al., 1999). Groundwater
contamination also results from the recharge of untreated urban runoff into the aquifer
(Carman et al., 1997). Potential contaminates come from vehicles, road oiling and salting,
pet wastes, industrial activities, and erosion from construction sites. Pollutants include
phosphates, sulfates, and nitrates from fossil fuels, fertilizers, and detergents. In addition,
bacteria, pathogens, heavy metals, and organic and inorganic compounds are commonly
found in urban runoff (Jones and Clark, 1987; The Conservation Foundation, 1987).
Imperviousness
Urbanization and its subsequent increase in impervious surfaces is affecting an increasing
number of watersheds and aquifers throughout the world (Jones and Clark, 1987). The
increase in impervious surfaces (roofs, streets, parking lots, etc.) has resulted in a significant increase in the amount of rainfall that is now being discharged into streams as surface
Chapter II.15:
Aquifer recharge management model
617
runoff, therefore lost to aquifer recharge. Even in the absence of direct industrial and
municipal runoff, water quality is adversely altered. Suspended sediments in an urban
watershed tend to be an order of magnitude higher than in natural, vegetated watersheds.
Increased sedimentation, especially of fine silts and clays, will increase the amount of
impervious surfaces once it settles in catchments and debris basins, as well as stream flood
plains (Schlosser, 1991; Jones and Clark, 1987).
The health of a watershed, and to a certain extent the aquifer recharge area, is directly
related to the amount of urban development (Chapter I.3). Schueler and Claytor (1997)
have developed a model by which the amount of impervious surfaces in a watershed may
determine the quality of runoff and thus the quality of water recharged into the aquifer.
Imperviousness is a useful tool for measuring the impacts of land development on
watersheds and aquifers. Furthermore, imperviousness is one of a few variables that can
be easily quantified, managed, and even controlled throughout all stages of development
(Chapters I.1, I.3, and I.5). Impervious surfaces collect and accumulate pollution, therefore,
they become a source for pollutants during storms. A direct link exists between the amount
of impervious surfaces in a watershed and levels of pollution (Watershed Protection
Techniques, 1994).
In this model, imperviousness is used as an indicator for the impacts of urban development on aquifer recharge areas. It can be easily inferred from geology, mapped, and
displayed in a series of colored maps that identify potential areas of aquifer recharge.
Furthermore, imperviousness can be used as a measure of environmental health since it
known to contribute to the overall degradation of urbanized watersheds and urban
streams.
Model structure
Before any meaningful evaluation of future alternatives can take place, it is important to
develop a framework that will allow for a comprehensive comparison of the impacts each
alternative has on groundwater resources (Figure II.15.1). For the ARMM to be useful, it
has to be able to evaluate the effectiveness of policy decisions in protecting aquifer recharge
areas. The possible range of alternatives is classified into 10 strategies based on the escape
of tigers. Policy decisions are classified into the 5 levels of reform derived from the
proceedings of a Symposium on the Integrity of Water held in 1975 by the U.S. EPA. The
application of the two models to the ARMM is described next.
In The escape of tigers: An ecological note (Table II.15.1) Haddon (1970) describes
10 strategies that deal with different levels of intervention to prevent the harmful transfer
of energy. These strategies describe different ways to modify the world in order to reduce
or eliminate economic losses associated with the natural interaction of people and the
environment. Strategies 1 to 3 focus on policy decisions; strategies 4 to 7 describe the range
of design and technical solutions; and strategies 8 to 10 are a combination of technical
solutions coupled with monitoring and warning mechanisms. In the ARMM, Haddons
tiger classification is applied to two forms of energy release: increased urban runoff and
the pollution of groundwater.
The Federal Water Pollution Control Act of 1972 was designed to restore, maintain
and protect the integrity on the nations water and water resources. As is the case with
most forms of legislation, the meaning of the word integrity was not made clear in the
language of the act. A symposium held by the EPA in 1975 was meant to clarify the issue.
The symposium focused on two interrelated aspects of the concept of integrity: as a
desirable characteristic of natural ecosystems and as a moral or cultural principle. Later,
Regier and France (1988) classified the various perspectives of the symposium participants
618
Objective:
Protecting Ground Water
Theoretical Framework
Escape of Tigers
Alternatives
Ecosystem
Integrity
Models
Policy Approach
Evaluation Models
Ecological Impacts
Aquifer Recharge Management Model
Impact Models
Policy/Planning
Decisions
Figure II.15.1 Framework for examining alternative impacts on groundwater resources.
into five categories (Table II.15.2). Their classification is based on the degrees of reform
that would achieve integrity and on the level of societal change or technology used as
implied by the acts wording.
The 10 classes of the escape of tigers were coupled with the five categories of ecosystem
integrity to yield the basis for reclassifying the alternatives and their impacts (Table II.15.3).
Chapter II.15:
619
Table II.15.1 Escape from Tigers Management Options

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Prevent the marshaling of the energy.

Reduce the amount of energy marshaled.
Prevent the release of the energy.
Modify the rate or spatial distribution of the release of energy.
Separate in space or time the energy being released.
Separation by interposition of a material barrier.
Modify appropriately the contact surface.
Strengthen the structure.
Detection and evaluation by generating a signal that a response is required.
Return to preevent conditions.
Source: Haddon (1970).
Table II.15.2 Strategies for Managing Ecosystem Integrity

1. Deep reform protection of water as a moral obligation for its inherent worth
2. Partial reform pragmatic, sectoral, step-wise societal change within a specified general direction
3. Incremental advances recognition that integrity is a multidimensional concept that is not valuefree, thus economics may sometimes dictate ecologics
4. Slowing the rate of retreat undertaking inexpensive but visible initiatives to protect an image
of concern
Source: Regier and France (1990).

Many of the existing integrated GIS models require varying degrees of empiricism, functional relationships, and deterministic descriptions of hydraulic processes to accurately
analyze the impacts of development. Furthermore, these models require lager amounts of
data that are not always available without a major investment in site-specific, base-line
studies (Table II.15.4). Simple models that are capable of simulating spatial processes and
identifying the distribution of nonpoint sources of pollution at a watershed scale are scarce
(but see Chapters II.10 and II.13).
In contrast to most GIS models, the ARMM, a raster GIS model, utilizes the grid functions
in ArcView and Arc/Info and requires only three layer of input: a digital elevation model
(DEM), geology, and high-altitude aerial photograph (Figure II.15.2). These data are reclassified to yield the basic building blocks of the model. Vector GIS coverage of topography and
geology is converted into grids. This conversion allows for a more effective handling of the
data by the ArcView grid module. The interactive GIS interface in ArcView allows for a
systematic input of spatial data and for the graphic display of impacts.
Unlike traditional GIS models that require uniform data throughout the study area,
the ARMM can operate on data with high levels of variability taking advantage of the
reclassify option in ArcView. This technique overcomes data limitations, which are based
on high-altitude aerial photography that tends to be high quality in the center and distorted along the edges. Furthermore, the model uses the cell as the smallest unit of surface
attribute. This allows for greater flexibility and accuracy than the traditional point line
and polygon approach of most traditional GIS models.
The DEM provides the base for stream networks and watersheds. Geology is reclassified into an infiltration rate based on the percolation characteristics of each rock type.
The aerial photograph provides the base for a land use/land-cover plan of existing conditions. Each of the land-use/land-cover classifications is given a runoff coefficient based
on empirical data of surface attributes. Empirical data shows that on average only 30%
of precipitation is naturally recharged into the groundwater. In areas where chaparral is
Site
I Cover development with a canopy and divert

water to recharge areas
I Build drains to collect and divert runoff
Source: Adapted from Haddon (1970) and Regier and France (1990).
I Strengthen/fix storm drains

and recharge systems
II Abandon settlements if threshold is reached
I Monitor drains and sewers

II Build ditches and systems as demand
increases
I Connect storm drains to sewage runoff

treatment and injection wells
I Build underground
II Cover roofs/buildings with sod
I Small footprint high-rise

II Build only on impervious geology
I Dispersed pattern of settlement

II Phase construction along with
mitigation measures
I On-site retention and recharge

II Surround site with French drains
to collect runoff
I Regional, large retention basins

with active recharge
I Single-family housing on large lots (>1 acre)

II Retain runoff on-site and discharge slowly
I On-site retention and recharge

II Minimize impervious surfaces
III Collect surface runoff from site
I Reduce the amount of impervious surfaces

II Reduce number of settlements
I Allow runoff onto aquifer recharge areas

II Dispersed pattern, large green spaces
I On-site small-scale active recharge

II Injection wells
Site
I Eliminate impervious surfaces

II Re-vegetate with chaparral
III No new settlement
Escaped Stormwater Region
620
7. Modify appropriately the contact surface incremental advances

I Eliminate grazing areas
I Cover site with impervious material to
I Eliminate imperious surfaces
II Eliminate roads (viaducts)
prevent infiltration of pollutants
II Maintain grazed areas
III Pollution-neutralizing surface
II Pollution-neutralizing surfaces
III Eliminate frosts from aquifer recharge areas
(I.e. wetlands)
8. Strengthen the structure holding the line
I Combined regional sewage stormwater
I On-site sewage/stormwater treatment
I Regional reservoirs recharge aquifer
treatment
II Full containment of pollution
actively
III Closed systems
9. Detection and evaluation by generating a signal that a response is required slowing the rate of retreat
I Monitor streams and wells
I Monitor drains and wells
I Monitor peak floods and pump water to
II Stop/modify activity
II Add clean water from alternate source
reservoirs
III Recharge with desalinized/reclaimed water
II Increase aquifer recharge areas as needed
by overpumping
10. Return to preevent conditions slowing the rate of retreat
I Develop alternative water source
I Pay for cleanup
I Allow current development and recharge
II Desalinization
II Evaluate cost-benefit of aquifer protection
II When threshold reached, fix it
III Water reclamation
IV Abandon aquifer
V Adequate water pricing
4. Modify the rate of energy or spatial distribution of energy release partial reform
I Build off of aquifer recharge areas
I Small buildings
II Build in a dispersed pattern
II Fewer cows, goats, sheep
III Less industry
IV Large open spaces
5. Separate in space or time partial reform
I Phase construction along with
I Locate polluting activities off
mitigation measures
recharge areas
II Alternate between grazing sites
6. Separate by interposition of a material barrier incremental advances
I Fence off aquifer recharge areas
I Treat runoff before recharge
1. Prevent the marshaling of the energy deep reform

I Remove polluting activities
I Collect/treat runoff
from aquifer recharge areas
II Dont build on aquifer recharge areas
III No agriculture, cars, polluting activities
2. Reduce the amount of energy marshaled deep reform
I Reduce the number of polluting activities
I Reduce the amount of polluting activities
II Treat pollution
II Small-scale development in high density
III Treat pollution
3. Prevent the release of the energy deep reform
I Remove pollution activities on aquifer
I On-site wastewater treatment and discharge
recharge areas
II Allow only nonpolluting activities
Pollution Region
Table II.15.3 Classification of Development Alternatives and Their Impacts
Chapter II.15:
621
Table II.15.4 Data Requirements

Aquifer Recharge Management Model
Stormwater Management Model1
1. Digital terrain model
1. Weather data: precipitation, snow melt, wind speed,

snowmelt coefficient, snow distribution, and other
melt parameters
2. Surface quantity: area, imperviousness, slope,
Mannings roughness coefficients, Horton or
GreenAmpt infiltration parameters
3. Subsurface quantity: porosity, field capacity, wilting
point, hydraulic conductivity, watertable elevation,
ET perimeters, coefficients for groundwater outflow
4. Channel/pipe quantity: linkages, shape, slope,
length, Mannings roughness, storage volumes at
manholes and other structures.
5. Storage sedimentation quantity: stage-areavolume-outflow relationships, hydraulic
characteristics of outflows
6. Surface quality: land use, total curb length,
catchbasin volume and initial pollutant
concentration, street sweeping interval, dry days
prior to initial precipitation, friction perimeters for
each land use
7. Dry-weather flow constants variations
8. Particle size distribution, Shields parameter decay
coefficients for channel/pipe quality
9. Storage/treatment: parameters defining pollutant
removal equation, individual treatment options.
10. Storage/treatment cost: parameters for capital and
operation and maintenance costs
2. High-altitude aerial photo
3. Geology layer
Adapted from: EPA SWMM Windows Interface Users Manual.
the main land cover, however, 80% of precipitation is recharged into the aquifer. By
contrast, in areas with dense forest cover, nearly all precipitation is lost due to evapotranspiration. This provided the basic reclassification scheme for the impact of vegetation type
on the aquifer recharge (different vegetation types reduced the infiltration factor of the
underlying substrata by as much as one level; for example, forests on high infiltration
substrates were classified as medium). The model looks at imperviousness as an indicator for loss of water to aquifer recharge, so an increase in the total area of impervious
surfaces indicates a decrease in the total surface area available for water to infiltrate
through the filtration zone.
The model treats all hard surfaces such as buildings and roads as being impervious
unless a policy or technology is implemented that results in bottom-line change of the
surface characteristics. For example, a policy may require all new development to retain
and recharge all runoff on site.
For purposes of simplification, some general assumptions influence the structure of
the model:
All surface runoff flows from areas of no infiltration to areas of maximum infiltration, and all runoff reaches the water table in these areas.
Groundwater protection provides a surrogate for protecting other resources such
as wildlife habitat and stream corridors.
622
Figure II.15.2 GIS-layered approach used in this study.
Case study
As a case study, to test the impacts of conventional development, we reexamined four
alternative land-use development futures that were produced in the Harvard Design
School ISCAR Studio (Steinitz, 1997). The study area is located in the Western Galilee in
northern Israel and is roughly 250 km2, containing seven watersheds. It is home to nearly
a hundred thousand people in several Arab Druze and Jewish settlements. Tefen Industrial
Park is located in the center of the study area and is the main place of employment in the
area. All alternatives assume tripling of the current population at plan buildup without
the addition of new settlements. Furthermore, as population growth is measured as a
factor of area, there is no change in current densities and development practices. Existing
land-use and the resulting imperviousness provide the base against which all alternatives
are compared (Figures II.15.3 and II.15.4).
ISCAR alternatives
1. Ring road (RR)
An expansion of the roads that connect the villages surrounding the center of the
study area in order to decentralize economic growth
Some urban development is diverted to the north and south of the study area
Impacts are concentrated along this road network where the majority of urban
development occurs
Chapter II.15:
Figure II.15.3 Existing land use. (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)

623
624
Figure II.15.4 Existing imperviousness. (Color version available at www.gsd.harvard.edu/watercolors. Password: lentic-lotic.)
Chapter II.15:
625
% Change in Minimum Imperviousness by Alternative
250%
% Change in Area of Minimum Imperviousness
200%
150%
100%
50%
0%
-50%
RR
RRT
UGR
NS
DR
Alternative
1 2 3 4 5 6 7 Mean
PE
SL
PR
Figure II.15.5 Percentage change in maximum imperviousness under different development alternatives: RR = ring road; RRT = ring road through Tefen; UGR = upgrade existing roads; NS =
northsouth road; DR = deep reform; PE = return to preevent; SL = slowing the rate of retreat; and
PR = partial reform.
Overall, this alternative increases the total amount of impervious surfaces by as much
as 63% on average per watershed. The most impacted watersheds are 1 and 5, where the
total area of impervious surfaces more than doubles (Figures II.15.5 and II.15.6).
2. Ring road through Tefen (RRT)
A shorter ring road that cuts through Tefen is expanded to allow development of
the periphery
Most of the development is concentrated south of the center of the study area
Impacts are concentrated along this road network where the majority of urban
development occurs
This alternative has a slightly lower impact, increasing the amount of impervious
surfaces by as much as 62% per watershed. In watershed 1, the total area of impervious
surfaces has increased by as much as 150%; in watersheds 4 and 5 the total impervious
area increases by 100% (Figures II.15.5 and II.15.6).
3. Upgrade existing roads (UGR)
Requires no major infrastructure investment beyond the expansion of the existing
road network to handle the increase in traffic associated with population growth
Most of the impacts are concentrated in the center of the study area around Tefen
and its surrounding settlements
Despite no investment in new infrastructure, this alternative increases the amount of
impervious surfaces by an average of 66% per watershed. Watersheds 4 and 5 increase by
as much as 100%, and watershed increases by more than 200% (Figures II.15.5 and II.15.6).
626

Distribution of Minimum Runoff by Watershed
2500
No. of Cells
2000
1500
1000
500
0
1
Watershed No.
EX
RR
UGR
RRT
NS
SL
PE
PR
DR
Figure II.15.6 Distribution of maximum runoff by watershed for different development alternatives
(identified in Figure II.15.5). (Color version available at www.gsd.harvard.edu/watercolors. Password:
lentic-lotic.)
4. North-south road NS
Requires a major investment in infrastructure by building a high-speed, limited-access
road that connects the two urban areas in the north and south of the study area
Impacts are concentrated mainly in the north and the south and the center remains
relatively undeveloped
Of all the ISCAR alternatives, the northsouth road (Figures II.15.7 and II.15.8) has
the highest impact. Overall it increases the total amount of impervious surfaces by nearly
77% per watershed on average. In watershed 1 it increases by more than 200%, and in
watersheds 4 and 5 it increases by more than 100% (Figures II.15.5 and II.15.6).
Policy alternatives
The next four alternatives were selected from the larger set of possibilities (Table II.15.3)
to test the impacts of policy decisions that favor the protection of aquifer recharge areas.
Natural resource management is not simply about protecting or exploiting resources, it is
also about achieving some dynamic balance between people and natural processes. In this
view, degradation of aquifer recharge areas can be controlled, reduced, and even eliminated by implementing land-use management programs. As the pumping of groundwater
tends to be cheaper than the divergence of surface water, watershed management becomes
a challenging task due to the biophysical, social, organizational, and economic phenomena
associated with water and land uses. The difficulty in watershed management comes from
the diffused use by different segments of a population of a common resource (all ethnic
groups in the region have equal access to the aquifer). Successful implementation of
watershed management depends on appropriate incentives that will resolve the conflicts
between public and private interests that are always associated with water use (Kozub
et al., 1987).
Chapter II.15:
Figure II.15.7 Land-use plan for the northsouth road development alternative. (Color version available at www.gsd.harvard.edu/
627
628
Figure II.15.8 Imperviousness resulting from the northsouth road development alternative. (Color version available at
www.gsd. harvard.edu/watercolors. Password: lentic-lotic.)
Chapter II.15:
629
These alternatives test the ability of the model to deal with the implementation of
different policies. The four policies were developed to mitigate the impacts of development
on aquifer recharge through the implementation of practices that encourage retaining,
treating, and artificially recharging urban runoff into the aquifer. The effectiveness of the
policy is measured by level of reform and compliance.
In the following alternatives, the level of compliance with aquifer recharge management practices varies by alternative and ethnic group. The basic assumption is that it is
easier to enforce compliance in the Jewish sector than in the Druze and Arab sector. This
is due in part to politics and in part to economics. The level of income in the Jewish sector
can absorb the increase in prices due to the use of expensive technology to treat urban
runoff. Mandating the use of the same technology in the Druze and Arab sectors may put
new development out of the reach of local residents. The alternatives also differ in the
level of reform or technology implemented for mitigation purposes.
1. Deep reform (DR)
The plan assumes no population growth since the protection of the aquifer is the
overriding consideration
All existing structures and settlements are retrofitted to treat polluted runoff and
active aquifer recharge
The impact of this plan (Figures II.15.9 and II.15.10) is to improve the aquifer recharge
over existing conditions. In all watersheds, the total amount of impervious surfaces
decreases by 12% per watershed on average (Figures II.15.5 and II.15.6). In watershed 3,
the total area of no impervious surfaces increases by more than 50%.
2. Return to pre-event (PE)
All new development should comply with a policy that requires minimal treatment
of polluted runoff and recharge. The driving force behind this alternative is its
requirement for more stringent compliance to avoid threshold impacts. Half of all
new development must comply with a law that changes the bottom-line characteristics of impervious surfaces to low imperviousness.
No existing structures are retrofitted to treat runoff and provide recharge
The overall impact of this alternative is similar to existing conditions (Figures II.15.5
and II.15.6).
3. Slowing the rate of retreat (SL)
Most new development is not required to comply with the policy
In the Druze and Arab sector, only 20% of new development is required to install
minimal recharge technologies. The Jewish sector has higher compliance requirements with more efficient technologies.
The overall impact is to increase the amount of impervious surfaces by 7% on average
per watershed conditions (Figures II.15.5 and II.15.6).
630
Figure II.15.9 Land-use plan for the deep reform development alternative. (Color version available at www.gsd.harvard.edu/
Chapter II.15:
Figure II.15.10 Imperviousness resulting from the deep reform development alternative. (Color version available at www.gsd.
631
632

4. Partial reform (PR)
Most new development is not required to comply with the policy.
Those that do comply must achieve a high level of recovery of runoff and aquifer
recharge. In the Arab and Druze sectors, 20% of new development is required to
implement technology that would result in a high level of aquifer recharge and
10% is required to employ minimal mitigation. In the Jewish sector, 40% of new
development is required to employ the highest available levels of mitigation and
10% is required to use the medium levels.
The overall impact of this alternative is to increase the amount of impervious surfaces
by as much as 45% on average per watershed (Figures II.15.5 and II.15.6). In watershed
3, the total area of none impervious increases by 50%.
Conclusion
Under existing practices of development and even under some of the less stringent policy
alternatives, the degradation of aquifer recharge areas due to urban development continues. This is evident in the change in the number of cells with maximum runoff (and
therefore, the degree of consequent infiltration).
The ARMM provides planners and decision makers with a simple tool to identify
critical areas of conflict between aquifer recharge and urban development. It also provides
a means to identify the magnitude of the impacts and it informs decision makers of the
most desirable alternative. As is the case with most models, the results are only as good
as the alternatives tested. It is the responsibility of planners and decision makers to ensure
that the widest range of alternatives is tested before meaningful decisions are made.
Literature cited
Berg, R.C., Curry, B.B., and Olshanksy, R., Tools for groundwater protection planning: An example
from McHenry County, IL, Environ. Manage., 23, 321, 1999.
Carman, N., Shamir, U., and Meiron-Pistiner, S., Water-sensitive urban planning: Protecting groundwater, J. Environ. Plann. Manage., 40, 413, 1997.
The Conservation Foundation, Groundwater Protection, Washington, DC, 1987.
Goldenberg, L.C. and Melloul, A.J., Hydrological and chemical management in the rehabilitation of
an aquifer, J. Environ Manage., 42, 247, 1994.
Gordon, W., A Citizens Handbook on Groundwater Protection, Natural Resource Defense Council, New
York, 1984.
Haddon, W., On the escape of tigers: An ecologic note, Technol. Rev. 6, 45, 1970.
Issar, A.S., Recharging and salivation processes in the aquifers in Israel, Environ. Geo., 21, 152, 1993.
Jaffe, M. and DiNovo, F., Local Groundwater Protection, American Planning Assoc. Washington, DC,
1987.
Jones, R.C. and Clark, C.C., Impacts of watershed urbanization on stream insect communities. Water
Resour. Bull., 23, 1047, 1987.
Kozub, J.J., Meyers, N., and DSilva, E., Land and Water Resource Management, Econ. Develop. Inst.
of the World Bank, Washington, DC, 1987.
Kuylenstierna, J.L., Bjorklund, G., and Najlis, P., Future sustainable water use: Challenges and
constraints. J. Soil Water Conserv., 52, 151, 1997.
Mostaghimi, T.U.S.S. and Shanholtz, V.O., Identification of critical nonpoint pollution source areas
using geographic information systems and water quality modeling, Water Res. Bull., 28, 877,
1992.
Chapter II.15:
633
Regier, H.A. and France, R.L., Perspectives on the meaning of ecosystem integrity in 1975, in
Edwards, C.J. and Regier, H.A., Eds., An Ecosystem Approach to the Integrity of the Great Lakes
in Turbulent Times, Great Lakes Fishery Commission, Ann Arbor, MI, 1990.
Schlosser, I.J., Stream fish ecology: A landscape perspective, BioScience, 41, 704, 1991.
Schueler, T. and Claytor, R., Impervious cover as a urban stream indicator and watershed management tool, Watershed Development Effects, 1997.
Steinitz, C., Alternative futures in the Galilee, Israel, www.gsd.harvard.edu/faculty/steinitz/steinitz.html, 1997.
Watershed Protection Techniques, The importance of imperviousness, 3, 100, 1994.
634
Response
Planning by examining alternatives
In this chapter, Amir Mueller, Robert France, and Carl Steinitz present a heuristically
appealing model that differs from many others currently in use due to its overt simplicity.
Model development always involves tradeoffs between realism and generality. The aquifer
recharge management model described here for a particular site in Galilee is easily adaptable to other locations due to its simple reliance on imperviousness as the metric of most
importance. This is the same strategy employed in generation of the open-space and
impervious surfaces model described by Jennifer Zielinski in Chapter I.3. Models such as
these have great utility due to the absence of hidden assumptions and complex mathematics that are sometimes employed in other cases, thereby giving false impressions of
much greater accuracy than really exists. Transparency in the derivation of water sensitive
planning models is necessary if they are to be communicated to, and subsequently
accepted by, the nonprofessional public. This is particularly true when undertaking landuse planning in rural communities where the education base may not be as high as it is
in urban settings.
Another important message in this chapter concerns the temporality involved in water
sensitive planning. The technique used in this study is referred to as alternate futures
scenario planning, a methodology that is becoming increasingly popular. In simplified
terms, it is based on the assessment of current conditions and the use of that data in order
to predict what the future conditions might become given certain assumptions. This
approach is similar in a way to the resource-driven, bottom-up GIS analyses described
by Jeffrey Schloss in Chapter II.12. These studies represent a maturity in the discipline of
land-use planning in their shift in focus from being merely reactive to damage to instead
offering proactive guidelines in order to preempt environmental disturbances before the
latter become manifest. Planning, in its most basic definition, is really all about concern
for the future, and the model developed in this chapter goes far toward illustrating just
how such an approach may be initiated.
L1562-2Ch16Frame Page 635 Thursday, April 4, 2002 9:05 AM
chapter II.16
Factors influencing sediment transport

from logging roads near boreal trout lakes
(Ontario, Canada)
Robert France
Abstract
Ontarios guidelines for protection of aquatic resources from riparian timber harvesting
are based on adoption of buffer strip widths in relation to surface slope from a study
conducted over 40 years ago. That work measured the maximum distance traveled by 36
erosion plumes arising from a single logging road in New Hampshire. Because statistical
reanalysis of that earlier study indicated that surface slope only explained 16% of the
variation in sediment transport, a new study was undertaken in northwestern Ontario of
the factors influencing erosion plumes from logging road washouts. Sixty-eight erosion
plumes were measured along a 4.3-km stretch of logging road in the riparian zones of
three boreal lakes. Road-generated sediment was rarely found to move more than 30 m,
thereby suggesting that the current guidelines, which have a 30-m minimum area of
concern buffer strip width, are for the most part adequate. Sediment transport distance
was found to be unrelated to surface slope but was significantly influenced by both road
gradient and downslope orientation, as well as by the presence of large surface obstructions blocking plume flow. A set of simple design criteria is suggested to minimize or
perhaps even eliminate chances of sediment input to water bodies, a small but still real
possibility under the present guidelines.
Our number one water quality problem in the national forest system
is roads.
M. Dombeck, former U.S. Forest Service Chief
(in Young-Petersen, 2000)
Introduction
Clearcutting has been a prevalent means of timber harvesting since the beginnings of
forestry in Ontario (Burgar, 1983). Today, the magnitude of such logging operations in the
Boreal Forest on the Canadian Shield is immense, exceeded by only a few locations
1-56670-562-2/02/$0.00+$1.50
635
636
elsewhere in the world. Because forestry and sport fishing are the two most important
generators of economic revenue for much of the Canadian Shield, the Ontario Ministry
of Natural Resources (OMNR) has developed series of timber management guidelines to
ensure protection of aquatic habitats during logging of watersheds (OMNR, 1985, 1986,
1988a, 1991, 1995).
Work in western North America has demonstrated that one of the most overt consequences of clearcutting can be substantial increases in the production of sediment (Patric
et al., 1984), much of which originates from the construction of logging roads (Beschta,
1978; Rothwell, 1983; Reid and Dunne, 1984; Bilby, 1985). Although research in northwestern Ontario has investigated the role of litterfall in mitigating interrill and aeolian erosion
following riparian timber harvesting (France, 1997; France et al., 1998; Steedman and
France, 2000), in addition to estimating whole-basin sedimentation rates after forest
removal (Blais et al., 1998), no published studies are available for this region regarding
the extent of sediment yields arising from logging roads on sloped ground.
Lakes, once envisioned as self-contained microcosms (Forbes, 1887), are now considered to be integrated components in a dynamic continuum of landscape processes (e.g.,
Likens, 1984). From this awareness comes recognition that the maintenance of vegetation
near waterbodies can mitigate many of the potential negative effects of [timber] harvesting
. . . [such that] the presence of a vegetated area adjacent to waterbodies acts to buffer the
waterbody from the effects of harvesting (OMNR, 1988b). The present study investigates
landscape influences on sediment transport arising from logging roads on sloped terrain
in northwestern Ontario, and examines the effectiveness of existing shoreline (riparian)
buffer strip reserves (OMNR, 1988a, 1991). Environmental guidelines for forest access
roads crossing water bodies, not considered in this study, are described elsewhere (OMNR,
1995).
Concomitant with the increased scale of timber removal in Ontario has been a substantial decrease in the recommended buffer strip widths from near 200 m around all
water bodies two decades ago, to as little as 30 m at the present time for cold water
lakes, which are known to support trout (reviewed in France et al., 1998). Today, Ontarios
recommended buffer strip widths for trout lakes are based on the upper values suggested
by Trimble and Sartz (1957) in relation to surface slope: 30 m for slopes of 1 to 15%, 50 m
for slopes of 16 to 30%, 70 m for slopes of 31 to 45%, and 90 m for slopes of 46 to 60%.
Trimble and Sartzs recommendations arose from measurements made of the maximum
distance traveled by 36 sediment plumes downslope from small culverts along a single
logging road in New Hampshire. What was not calculated by Trimble and Sartz (1957),
nor mentioned in OMNRs uncritical adoption of these guidelines, however, is that slope
explained only 16% of the variation in sediment transport. Later research on logging road
erosion (e.g., Haupt, 1959) has questioned Trimble and Sartzs conclusions regarding buffer
strip effectiveness. Trimble and Sartz also recognized the limitations of their own data,
acknowledging that distances of sediment deposition may at times be greater than the
maximum estimated from the observations. Therefore, OMNRs (1988a) assurance that
adoption of such guidelines will prevent any input of sediment into water bodies could
be regarded with skepticism.
Given the questionable scientific background for this particular aspect of the present
timber management guidelines, OMNR have themselves admitted that the effectiveness of
the guidelines in mitigating potential negative effects of harvesting has not been determined (OMNR, 1988b), such that more information on the effectiveness of riparian
reserves in northern Ontario is necessary in order to determine if the guidelines are being
followed and are working (Pike and Racy, 1989). This supports a survey of timber
management operations in Canada, wherein Freedman (1982) concluded that while the
usefulness of buffer strips is generally acknowledged, the widths that are required for
Chapter II.16:
Factors influencing sediment transport from logging roads
637
them to be effective are still controversial. The goal of the present study, therefore, is to
obtain information useful for planning the effective widths needed for protective buffer
strips in northwestern Ontario.
Methods
This study was conducted in northwestern Ontario, 45 km northwest of the town of
Atikokan and 150 km southeast of the Experimental Lakes Area, in a region transitional
between the Boreal and Great Lakes/St. Lawrence forest types (France, 1997; France et al.,
1998; Blais et al., 1998; Steedman and France, 2000). The landscape is composed of gently
rolling knolls of Precambrian bedrock interspersed with boggy plains and glacial deposits.
The logging roads surveyed in this study ring three small headwater lakes whose watersheds had been experimentally clearcut three years previously as part of an OMNR
research project (Steedman and France, 2000; Steedman et al., 2000). These lakes support
5 to 7 species of fish, including lake trout (Salvelinus namaychush). Road surfaces are
composed of local sand and gravel and have received little or no motorized traffic since
logging operations ceased in 1996.
During 1999, 4.3 km of logging roads were walked and all observed washout erosion
plumes (predominantly composed of unconsolidated sand) were measured. In addition
to maximum sediment transport distance, plume widths were assessed at several locations
in order to estimate surface-area coverage. Detailed, plume-specific maps were drawn and
are displayed in CD-ROM format (with accompanying photographs) in France (in prep).
Factors that could potentially influence sediment erosion were measured in three basic
locations: upslope of the road, the source road itself, and downslope from the road along
the path of the erosion plume. In all cases, the factors were selected for ease of measurement, suitable for inclusion in a large synoptic regional survey.
Upslope factors included those that might influence minicatchment hydrology: the
overall shape (concentrating, dissipating, flat), average slope determined along 30-m
lengths at the center axis and on each side edge at 45, and extent (low, medium, high)
of small water collection reservoirs; and those with a potential influence on sediment
generation and movement: vegetation density and height (low, medium, high) determined
at six locations along the slope transect lines.
Source road factors included those that could potentially influence the generation,
amount, type, and transportability of sediment: road width, gradient, and orientation to
downslope plume (perpendicular, angled, parallel), whether the road was sunken or
raised, upslope (i.e., cut) embankment height, distance to the nearest erosion washout,
and the particle size (coarse, medium, fine) and surface compaction state (low, medium,
high) of the road.
Downslope factors included those that might be expected to influence the ease of
sediment movement: slope gradient, slope obstruction index (number of branches, boulders, etc. along the plume length, as in Haupt, 1959), estimation of the extent of surface
depressions (difference between straight and surface contoured distances over complete
plume length), forest floor roughness (low, medium, high), vegetation cover and height
(low, medium, high) determined at three locations, shape of general area (concentrating,
dissipating, flat), and plume shape (determined from four cross-width measurements),
number of entry washouts and side rills.
Analyses consisted of examining correlation coefficients for variables treated independently and also in ANOVAs for variables treated interactively. Significance was considered
at p < 0.05. Only those variables with the most demonstrable influence on sediment
transport distance are discussed next. The purpose of analyses undertaken in this study
was to highlight, in a nonpredictive way, the major factors influencing length of erosion
638

30
25
20
15
10
FREQUENCY
5
0
10
15
20
25
30
35
EROSION PLUME LENGTH (m)

30
25
20
15
10
5
0
25
50
75
100
>100
EROSION PLUME AREA (m2)

Figure II.16.1 Frequency distributions of lengths and surface areas of 68 erosion plumes originating
from logging road washouts in northwestern Ontario.
plumes. Other studies concerned with developing predictive empirical models of sediment
movement have used multiple regression techniques (e.g., Haupt, 1959; Packer, 1967; Elliot
et al., 1999).
Results
Of 4.3 km of surveyed roads surrounding the study lakes, 104 individual road drainage
exits were observed, comprising 68 specific erosion plumes. About half of these plumes
were isolated (and therefore independent) from one another, being separated by distances
of over 30 m; about a quarter, however, were situated within 10-m road distance from
each other (and therefore influenced by the same drainage flow). The remaining quarter
were in between these two categories. The average surface area of the plumes was not
great, most covering less than 50 m2 size (Figure II.16.1, bottom panel). Plumes varied
considerably in shape in relation to both source and downslope surface characteristics
(illustrated in France, in prep). In some cases, a single washout spread to multiple lobes,
whereas in other cases, multiple washouts conjoined into a single plume. Other erosion
plumes were intermediate between these extremes; for example, proximal parallel washouts with either considerable or little intermixing of their sediment loads.
Maximum length of erosion plumes varied between 3 and 32 m. Most important, only
2% of all measured erosion plume distances exceeded the smallest buffer strip width
guideline of 30 m, with over half of all plumes being less than 10 m in length (Figure II.16.1,
upper panel).
Chapter II.16:
639
16
12
8
4
0
010%
1020%
2040%
SURFACE SLOPE
16
12
8
4
0
020
2040
>40
# SURFACE OBSTRUCTIONS
Figure II.16.2 Influence of surface slope and number of obstructions on the length of erosion plumes.
Error bars are 2 SD. No significant differences (ANOVA tests; p = 0.05) exist among groupings.
Of all the factors examined for their influence on sediment transport distance, none
displayed any statistical interaction, and only three were independently significant. Surprisingly, neither the landscape characteristics of downslope surface slope nor the number of
surface obstructions played a demonstrable role in affecting plume length (Figure II.16.2).
In contrast, measurements of the sediment road source were found to strongly influence erosion plume length. The orientation of the road was important in that plumes
originating from washouts perpendicular to the aspect of the road were only half the
length of those originating from washouts situated either angled or parallel (illustrated
in France, in prep) to road direction (Figure II.16.3, upper panel). Road gradient also
effected sediment transport, with slopes between 10 to 15% producing plumes about twice
the length of those coming off roads having slopes less than 5% (Figure II.16.3, lower
panel). The worst situations in terms of sediment movement occurred when steep gradient
roads abruptly turned to become contoured with the surface slope. In such cases, a series
of washouts developed as the sediment from road surface did not make the bend and
continued downslope after building up considerable energy (illustrated in France, in prep).
Here, runoff rills were also frequently observed on the road surface, and sometimes erosion
scouring could be seen downslope (photographs in France, in prep). Transported sediment
plumes in these locations were often in the order of 15 to 35 cm deep.
Finally, although smaller surface obstructions were not found to significantly influence
overall sediment transport distance, whether or not the plume abutted against a major
terminal obstruction was important. Plumes that ended in natural or logging terminal
obstructions averaged about 8 m in length compared to about 13 m for those plumes that
did not end abruptly (Figure II.16.4). Natural termini included bedrock walls, boulder
640

20
16
12
8
4
0
Perpend
Angled
Parallel
ROAD ORIENTATION
20
16
12
8
4
0
05%
510%
1015%
ROAD GRADIENT
Figure II.16.3 Influence of road orientation (see text) and gradient on the length of erosion plumes.
Error bars are 2 SD. Significant differences (ANOVA tests; p = 0.05) existed between perpendicular
and both angled and parallel orientated roads (which did not differ from each other), and between
0 to 5% and 10 to 15% road gradients.
16
12
0
Natural
NO TERM
Logging
TERMINUS
Figure II.16.4 Influence of terminal obstructions on length of erosion plumes. Error bars are 2 SD.
Plumes ending in either natural or logging-produced barriers (see text) were significantly (ANOVA
tests, p = 0.05) shorter than plumes with no such terminal obstructions.
Chapter II.16:
641
fields, small water bodies, and mature forest trees, and termini created by logging included
slash piles, skidder ruts, excavated pits, and plowed hills (illustrated in France, in prep).
Discussion
This survey found very few cases in which erosion plumes from logging roads exceeded
30 m in length, the minimum width recommended for protective buffer strips in Ontario.
The straightforward conclusion is that adherence to Ontarios existing guidelines for buffer
strip widths (OMNR, 1988a) should generally be adequate for prevention of most sediment
transport to aquatic systems (at least for this particular study region of northwestern
Ontario). Empirical studies such as the present example have a further use in being able
to identify the most important influences of erosion and to provide a design framework
for reducing sediment transport to levels well below the legal requirements, thereby
satisfying the heightened safety concerns of environmental managers and public interest
groups.
The role of downslope surface gradient in influencing sediment transport is not as
straightforward as Trimble and Sartzs (1957) early work may have suggested. Whereas
some work (e.g., Nieswand et al., 1990) is conceptually based on the assumption of a
strong influence of slope being the primary determinant of erosion plume distance, other
empirical work (e.g., Haupt, 1959) has, like the present study, found no existence of such
a relationship between surface slope and transport distance. In Haupts (1959) study, it
was the downslope surface obstruction index that was the most significant regulator of
sediment flow distance, a finding not duplicated, however, in the present study. Additionally, although other work has highlighted the importance of vegetation density in
influencing sediment transport (e.g., Packer, 1967), this did not seem important in northwestern Ontario. The present work did find that whereas small surface obstructions or
ground cover characteristics were unimportant in influencing sediment transport, large
surface obstructions were significant in their role in ultimately blocking plume movement.
The demonstrable influence of the source road on sediment transport was a finding
similar to both the present study and previous research. As was the case observed for
northwestern Ontario, both Packer (1967) and Bilby et al. (1989) considered road gradient
to be extremely important in determining the maximum distance of sediment movement.
Such a result is largely a function of sediment-laden runoff moving rapidly down road
surfaces at high velocity, resulting in gullies being formed, which in turn produces more
sediment movement, which then adds to further road cutting and sediment transport, etc.
For the present study site, when such steep slopes occurred in relation to roads that were
oriented parallel to the surface gradient, the surface area of potentially erodible sediment
was much greater than that for contoured roads situated perpendicular to the surface
gradient; i.e., perpendicular road widths were generally only about 6 m compared with
unobstructed runout distances of well over 20 m for downslope-oriented roads.
The present results can be important for developing management strategies for limiting sediment erosion from logging roads in northwestern Ontario, because all the factors
found to be significant determinants of plume length can be easily designed to maximize
their effectiveness in reducing the possibility of extended sediment transport. In other
words, the following brief list of simple design criteria should be examined by professional
foresters to mitigate erosion damage to aquatic systems:
1. Avoid, whenever possible, running roads downslope toward the water.
2. If this cannot be achieved, limit where downslope roads occur to only those
locations having gentle surface gradients.
642

3. If downslope roads of steep gradient are absolutely unavoidable, then reduce the
likelihood of extensive runoff plumes developing by:
a. Building a gradual rather than sharp turn of the road where it meets the
contoured grade so as to limit serious erosion washouts
b. Grading the contoured road surface down on the upslope (cut) edge to form a
raised downslope (fill) lip over which sediment will have more difficult time
moving
c. Bisecting the downslope road surface by numerous cross-ditches in order to
prevent the buildup of high runoff energy by channeling water flow away
d. Constructing buttresses of logging slash and other debris (e.g., boulders) or
create deliberate excavation pits at the road bends to thwart the straight-on
runout movement of sediment further downslope
Experimental testing and monitoring of these simple, common-sense measures for

roads on slopes could provide information needed for improved protection for Ontarios
aquatic resources under the existing guidelines for buffer strip widths (further best management practices applicable to roads crossing streams are detailed in OMNR, 1995). These
measures are not new, having been previously suggested for mitigating road erosion in
other regions: shallow grade (Packer, 1967) and insloped (Elliot and Tysdal, 1999) roads,
cross-ditching (Elliot et al., 1999), and brush barriers (Swift, 1986). Casual surveys of
logging operations by a variety of companies over a wide area of northwestern Ontario
suggest, however, that more careful attention paid toward implementing these measures
might be beneficial toward more effectively limiting sediment movement.
Sediment plume lengths measured in this study in northwestern Ontario are generally
shorter than those for other areas characterized by steeper terrain and greater rainfall
(tabulated in Elliot et al., 1999). Nevertheless, in the absence of more careful attention paid
to logging road design and plume barrier construction, there still remains the possibility
that some sediment might reach water bodies in this region. Therefore, another management option to guarantee no inputs of sediment to Ontarios water bodies is to simply
increase the smallest recommended buffer strip width from 30 m to some, as yet to be
determined, width. In such a situation, detailed attention to site-specific design of logging
roads would not be of as much paramount importance as under the present guidelines
of probable, but not definitive, effectiveness.
Given that surface slope was found to be relatively unimportant in influencing sediment transport distance, does it then make sense to maintain buffer strip guidelines that
are largely based on this factor? Yes, because although small-scale surface undulations
(and also obstructions) may not play a major role in washout plume length, the overall
surface slope will affect road gradient and orientation, two factors of marked importance
for determining the likelihood of sediment transport. Therefore, notwithstanding the
implicit managerial/regulatory problems of dealing with a sliding scale of buffer strip
width recommendations (rather than having a single width value irrespective of the local
landscape), it makes sense to use slope as a reflection of sediment transportability due to
its controlling influence on effective road design.
Finally, the issue of what elements should actually comprise an effective buffer strip
has not been addressed in this study (but see France et al., 1998) (i.e., Do buffer strips
really have to be forested to work?). The present results suggest that the role of buffer
strips as sediment filters (only one of the many important protective roles played by
riparian forests) may actually not need trees to effectively trap sediment. In such cases,
buffer strips, or perhaps more appropriately areas of concern (OMNR, 1988a), can
perform this function without an intact forest being present; however, it is important to
recognize that trees, although perhaps unimportant in terms of reducing road erosion,
Chapter II.16:
643
may influence other forms of erosion and consequent sediment transport (Steedman and
France, 2000; France, 1997).
Acknowledgments
This work was supported by grants from the Ontario Ministry of Natural Resources and
the Milton Fund and Design School Deans Fund of Harvard University. R. Steedman is
thanked for providing research facilities at the Coldwater Lakes Camp and many discussions about forestry operations, as is F. de Kock for generous field assistance.
Literature cited
Beschta, R.L., Long-term patterns of sediment production following road construction and logging
in the Oregon Coast Range, Water Resour. Res., 14, 1011, 1978.
Bilby, R.E., Contributions of road surface sediment to a western Washington stream, Forest Sci., 31,
827, 1985.
Bilby, R.E., Sullivan, K., and Duncan, S.H., The generation and fate of road-surface sediment in
forested watersheds in southwestern Washington, Forest Sci., 35, 453, 1989.
Blais, J.M. et al., Climatic changes in northwestern Ontario have had a greater effect on erosion and
sediment accumulation than logging and fire: Evidence from 210Pb chronology in lake
sediments, Biogeochem., 43, 235, 1998.
Burgar, R.J., Forest land-use evolution in Ontarios Upper Great Lakes Basin, in Flader, S., Ed., The
Great Lakes Forest: An Environmental and Social History, Univ. of MI Press, 1983.
Elliot, W.J. and Tysdal, L.M., Understanding and reducing erosion from insloping roads, J. Forestry,
97, 30, 1999.
Elliot, W.J., Hall, D.E., and Graves, S.R., Predicting sedimentation from forest roads, J. Forestry, 97,
23, 1999.
Forbes, S.A., The lake as a microcosm, Bull. IL Natur. Hist. Surv., 15, 537, 1887.
France, R.L., Potential for soil erosion from decreased litterfall due to riparian clearcutting: Implications
for boreal forestry and warm- and cool-water fisheries, J. Soil Water Conserv., 52, 452, 1997.
France, R.L., Peters, R., and McCabe, L., Spatial relationships among boreal riparian trees, litterfall
and soil erosion potential with reference to buffer strip management and coldwater fisheries,
Annales Botanni Fennici, 35, 1, 1998.
Freedman, B., An Overview of the Environmental Impacts of Forestry, with Particular Reference to the
Atlantic Provinces, I.R.E.S. Research Dalhousie Univ., Halifax, NS, 1982.
Haupt, H.F., Road and slope characteristics affecting sediment movement from logging roads, J.
Forestry, 57, 329, 1959.
Likens, G.E., Beyond the shoreline: A watershed-ecosystem approach, Internationale Vereinigung fur
Theoretisch and Angewandte Limnologie Verhandlungen, 22, 1, 1984.
Nieswand, G.H. et al., Buffer strips to protect water supply reservoirs: A model and recommendations, Water Resour. Bull., 26, 959, 1990.
Ontario Ministry of Natural Resources (OMNR), Class Environmental Assessment for Timber Management on Crown Lands in Ontario, Toronto, 1985.
OMNR, Timber Management Planning Manual for Crown Lands in Ontario, Toronto, 1986.
OMNR, Timber Management Guidelines for the Protection of Fish Habitat, Toronto, 1988a.
OMNR, Prevention, mitigation and remedy of potential negative effects. Statement of evidence panel
X: Harvest. OMNR, Thunder Bay, ON, Environ. Impact Assess. II, 512, 1988b.
OMNR, Code of Practice for Timber Management Operations in Riparian Areas, Thunder Bay, ON, 1991.
OMNR, Environmental Guidelines for Access Roads and Water Crossings, Toronto, 1995.
Packer, P.E., Criteria for designing and locating logging roads to control sediment, Forest Sci., 13, 2, 1967.
Patric, J.H., Evans, J.O., and Helvey, J.D., Summary of sediment yield data from forested land in the
United States, J. Forestry, 76, 101, 1984.
Pike, M. and Racy, G.D., Why Have Riparian Reserves in Timber Management? A Literature Review,
OMNR, Thunder Bay, ON, 1989.
644
Reid, L.M. and Dunne, T., Sediment production from forest road surfaces, Water Resour. Res., 20,
1753, 1984.
Rothwell, R.L., Erosion and sediment control at road-stream crossings, The Forestry Chron., 25, 62,
1983.
Steedman, R.J. and France, R.L., Aeolian transport of sediment from new clearcuts into boreal lakes,
northwestern Ontario, Canada, Water, Air and Soil Pollution, 122, 139, 2000.
Steedman, R., Kushneriuk, R., and France, R., 2000, Initial response to catchment deforestation of
small boreal forest lakes in northwestern Ontario, Canada, Internationale Vereinigung fur
Theoretisch and Angewandte Limnologie Verhandlungen, in press.
Swift, L.W., Filter strip widths for forest roads in the southern Appalachia, Southern J. Appl. Forestry,
10, 27, 1986.
Trimble, G.R. and Sartz, R.S., How far from a stream should a logging road be located? J. Forestry,
53, 339, 1957.
Young-Petersen, T., Ripping roads for watershed health, Orion Afield, 4, 23, 2000.
Chapter II.16:
645
Response
Empirically testing planning assumptions
Shoreline buffers are a recognized and established best management practice to protect
water bodies from sediment generated as a result of riparian development. As Frank
Mitchell outlines in Chapter II.1, such shoreline buffers can also play a variety of other
important roles such as providing wildlife corridors and offering human amenities, the
latter a topic elaborated on by Charles Flink in Chapter II.3. Issues of how wide buffer
strips need to be in terms of providing their protective function is an issue touched upon
by Jeffrey Schloss in Chapter II.12, and by Robert France, John Felkner, Michael Flaxman,
and Robert Rempel in Chapter II.14.
Particularly, it is the erosion caused by the construction and continued operation of
roads that is regarded as being the major source of sediment and concordant nutrient
pollution to lakes and rivers from land use. In Chapter II.12, Jeffrey Schloss explores the
implications of one such road on the water quality of a lake in New Hampshire through
use of GIS analysis. In the present chapter, Robert France examines the implications of
road construction and sediment transport at the much smaller scale of a few individual
sites.
As France discusses, the width recommendations for Ontarios buffer strip guidelines,
based as they are on a single, possibly flawed, study in New Hampshire over four decades
ago, needed to be critically reexamined. The lesson here for water sensitive planners is an
important one. Assumptions can, through time, become easily codified into planning
recommendations that may be unrealistic. It is always important, therefore, to critically
reexamine such entrenched assumptions with fresh, modern eyes and methods.
In this case, France determined that the established buffer strip width recommendations, notwithstanding their dubious origin, nevertheless did seem adequate in terms of
preventing sediment transport from logging roads to receiving waters. The approach taken
in this chapter, namely that of an empirical survey of numerous sites, is an investigative
tool that is well adapted to objectively testing assumptions of water sensitive planning.
It is easy to imagine such an approach being adapted to a wide range of investigations,
for it transcends idiosyncrasies of a specific site and enables broader conclusions to be
drawn about land management in general. Further, as France demonstrates in this chapter,
it is possible, and indeed seems natural, to move from empirical scientific analyses to
design recommendations, as, for example, with respect to the construction of logging
roads.
One of the most interesting and provocative results from this study is the finding that
the role of buffer strips as sediment filters may occur independent of whether or not
standing trees are actually present. The finding that surface obstructions such as piles of
logging slash can effectively trap sediment, and therefore reduce the potential for aquatic
damage from road erosion, is of major interest to timber companies already fearing the
economic implications of what they imagine to be overly protective buffer strips surrounding water bodies in this region (see Robert France, John Felkner, Michael Flaxman, and
Robert Rempel in Chapter II.14).
L1562-2Ch17Frame Page 647 Monday, April 8, 2002 10:08 AM
chapter II.17
Limnology, plumbing and planning:

Evaluation of nutrient-based limits to
shoreline development in Precambrian
Shield watersheds
Neil J. Hutchinson
Abstract
The concept of using water quality as a planning tool for recreational lakes has been
in active practice in Ontario and parts of the United States for approximately 25 years. In
practice, assumptions regarding anthropogenic loadings of phosphorus to a watershed
(generally septic systems servicing shoreline development) are linked to estimates of
natural phosphorus loading. The resultant model estimates total phosphorus concentration and the response of trophic status indicators such as water clarity and dissolved
oxygen in specific lakes. Linking the model to a water quality objective allows planners
to set capacities for anthropogenic phosphorus loads, and hence shoreline development
such as cottages, resorts, or permanent homes. This chapter presents an example of how
the concept can be applied in practice, based on the application of the authors experience
to a test watershed in south-central Ontario. Practical examples are given to show the
development and calibration of accurate trophic status models, the use of monitoring data
to set ecologically valid water quality objectives and their translation into shoreline development capacities, and to show the strengths and weaknesses of the approach.
The availability of a scientifically based water quality model has overemphasized
water quality as a planning tool and generated unrealistic expectations of a single-capacity
determinant among the public. Recent advances in our understanding of the geochemistry
of domestic septic systems indicates that less phosphorus may be mobile than was previously assumed. In addition, as alternative septic technologies for phosphorus abatement
are developed a refocusing of capacity determinants will be required. A combination of
land-use regulations and a scientifically based management program is recommended as
an alternative to a single, phosphorus-focused approach. These could address stresses to
the ecology of the riparian and littoral zones and acknowledge the importance of social
determinants such as noise, crowding, powerboats, and the wilderness aesthetic. This
would promote a diversity of planning approaches, shift the existing focus away from
plumbing and septic systems, and provide a more holistic management program which
protected more components of the lake system.
1-56670-562-2/02/$0.00+$1.50
647
648
Figure II.17.1 Southern Ontario tertiary watershed divisions showing Muskoka River watershed
draining west to Georgian Bay.
Introduction
Lake-based recreation and tourism is a fundamental cornerstone of the economy for many
small rural communities situated on the Precambrian Shield in Ontario. The economic
stimulus provided by tourism has diversified the economy of these communities, which
were historically dependent on declining revenues from forestry or mining. Tourism on
the southern portions of the Shield draws from the large and prosperous demographic of
southern Ontario, in particular the Golden Horseshoe of 7 million people, which extends
from east of Toronto to Niagara Falls, but also draws from the Great Lakes states of
Michigan, Ohio, and New York.
The Precambrian (or Canadian) Shield is made up of Precambrian gneiss and granite
bedrock that lies near or on the surface throughout most of the northern areas of the
province. Landscapes on the southern Shield are dominated by bedrock ridges and knolls,
thin soil cover, and pine and hardwood forests. Lakes are also a dominant presence on
the landscape. The Province of Ontario has over 260,000 lakes greater than 1 ha in size
(Cox, 1978) and most of these are situated on the Shield. Thin soils and acidic bedrock
mean that most of these lakes are nutrient-poor and possess high water clarity. Clear
waters, rocky shorelines, and forest cover produce exceptional recreational attributes and,
as a result, southern portions of the Shield have served as a focus for shoreline recreation
in the form of cottages and resorts.
The Muskoka River Watershed was chosen as a test case to illustrate the development,
validation, and application of a trophic status model to govern recreational shoreline
development. The watershed covers approximately 5737 km2 and is located on the southernmost fringes of the Precambrian Shield, approximately 180 km north of Toronto
(Figure II.17.1). Four large (>5,000-ha) lakes Lake Muskoka, Lake Rosseau, Lake Joseph,
and Lake of Bays form the recreational focus of the region. These were opened up to
summer recreation with the arrival of railways in the 1880s and the formation of steamship
service on larger lakes shortly thereafter. Approximately 150 smaller lakes also support
shoreline recreation. Three major towns are located in the watershed, each with a population exceeding 8,000 permanent residents. Eight smaller communities of 500 to 2,000
people also exist. Some of these are serviced with sewage treatment plants discharging to
the watershed, while others are serviced by septic systems.
Development on these lakes consists of approximately 18,000 shoreline residences,
4,300 resort units, and 1,400 trailer sites. Approximately 4,800 ha of urbanized areas are
Chapter II.17:
Limnology, plumbing and planning
649
located adjacent to shorelines and 2,700 kg/year of phosphorus are added from point
source STPs in urban centers. The Lake Rosseau watershed also includes approximately
4,800 ha of agricultural land use. There are approximately 4,400 vacant shoreline lots across
the watershed, which represents a substantial resource base of future development potential. Approximately 1,400 back lots (i.e., set back from the shoreline) exist, and about one
third of these are vacant.
The Province of Ontario and various municipal governments in recreational areas
have maintained water quality programs since the late 1970s, to manage recreational
growth in recognition of the important economic link of tourism to water quality. These
programs generally consist of four elements:
1. Policies to maintain water quality through limits to shoreline development
2. Predictive models linking shoreline development to water quality
3. Lake-specific policies, including development objectives based on water quality
and,
4. Monitoring programs to track changes in water quality in lakes
This chapter describes a process for developing a water quality model, validating its
predictions of water quality, and setting development objectives on the basis of water
quality. Issues of water sensitive land-use planning with respect to rural lakes are also
covered in Chapters II.1, II.12, II.13, II.14, and II.16.
The author has made use of water quality and land-use data for a set of lakes situated
within the Muskoka River watershed. The concepts and observations herein are those of
the author alone and are intended to guide technical practitioners of water quality planning
in recreational lakes in Ontario and elsewhere. They are not presented as specific recommendations for water quality planning for lakes in the Muskoka River Watershed but the
Muskoka watershed is used as an example of how these models can be applied throughout
the Precambrian Shield in Ontario and elsewhere.
The history and origins of lakeshore capacity planning in Ontario

The first lakeshore capacity planning initiatives in Ontario grew out of the efforts to control
eutrophication of Lake Erie in the 1970s. Lake Erie was a large and visible example of the
threats posed by enrichment of surface waters with the algal nutrient phosphorus and of
the success of remedial programs that were centered on managing the lakes phosphorus
budget. In the same era, the eutrophication of inland lakes was also documented in
response to inputs of partially treated sewage effluent. Among these was Gravenhurst
Bay on Lake Muskoka, which suffered a history of algal blooms until tertiary sewage
treatment was implemented in 1972 (Michalski et al., 1975) and enhanced treatment and
relocation of the outfall were implemented in May 1994.
The primary water quality concern in Ontarios cottage country is also nutrient enrichment. Excessive phosphorus input promotes the growth of algae, causing a loss of water
clarity. The lake user sees this as greener water of less aesthetic appeal or as surface
blooms of nuisance algal growth. Algae settle to the bottom of the lake, where their
decomposition consumes oxygen, reducing the amount of cold, oxygen-rich habitat available for sensitive aquatic life such as lake trout (Salvelinus namaycush) and triggering
remineralization of sediment-bound phosphorus. Residential or cottage development on
a shoreline may increase the input of phosphorus to a lake. Domestic septic systems may
be a significant component of the loading, but clearing of the shoreline, fertilizer application, and increased erosion are also important.
650
The Lake Erie experience prompted an examination of phosphorus sources to inland

lakes, in the understanding that unchecked recreational development and phosphorus
loadings from septic systems might impair recreational water quality. The DillonRigler
model (Dillon and Rigler, 1975) was the first model to specifically address the relationship
between potential eutrophication of Ontarios Precambrian Shield lakes and the density
of residential development on their shorelines. Its rapid recognition and acceptance by
the international scientific community led to the development of Ontarios Lakeshore
Capacity Study (1976 to 1980) in the belief that substantial predictive relationships might
be developed for other responses of lakes to shoreline development. The Lakeshore Capacity Study was coordinated by the Ontario Ministry of Municipal Affairs and Housing and
published in 1986 (Downing, 1986). It produced predictive models for land use, fisheries
exploitation, wildlife habitat, microbiology, and water quality and integration of all components into one capacity model. Although several of these models were very useful, the
water quality model was the only one adopted for routine use by management agencies
in Ontario. It was a minor variant of the Dillon Rigler model (Dillon et al., 1986).
The Ontario Lakeshore Capacity Model was based on the nutrient or trophic status
of a lake. It provided an accurate and quantitative linkage between the amount of shoreline
development, phosphorus and chlorophyll a concentrations in a lake, and resultant
water clarity. Both the DillonRigler and Ontario models were developed and calibrated
for small Precambrian Shield lakes in the Muskoka and Haliburton recreational areas of
Ontario. Subsequent work refined the models and added significant new areas of understanding, most notably a component linking shoreline development and oxygenated lake
trout habitat (Molot et al., 1992) and testing of model predictions on the original study
lakes (Dillon et al., 1994) and an independent set of study lakes (Hutchinson et al., 1991).
Lakeshore capacity planning in Ontario

Lake-based recreation and tourism are fundamental to the economics of municipalities in
the southern Precambrian Shield region of Ontario, and so water quality is also linked
with tourism. Shoreline residents buy property with an expectation of stability in the water
quality of their lake and of stability in their social environment. Under Ontarios Planning
Act, municipalities have the legislative jurisdiction to monitor and regulate land use.
Ontarios Provincial Policy Statement also requires that municipal planning policies recognize and protect water quality.
As part of this responsibility, some municipalities undertook programs to implement
variants of the provinces Lakeshore Capacity Model in the late 1970s and early 1980s.
In areas of the province that were not organized into municipalities, the provinces Ministry of the Environment implemented lakeshore capacity planning based on the provinces
model. The amounts of shoreline development allowed by these approaches are termed
lakeshore capacities or development capacity. Management of lakeshore capacity is
intended to protect water quality directly. Shoreline development capacities, however,
whether based on trophic status or other means, also help to maintain the social stability
desired by shoreline residents by assigning a finite limit to the amount of shoreline
development.
The provincial and municipal models have enjoyed widespread popular support since
their implementation. Adaptations to personal computers were made in the late 1980s and
early 1990s, but no thorough technical reviews or evaluations have been undertaken.
Periodic reviews and reworking of these models are required, however, to incorporate
recent advances in scientific understanding.
Chapter II.17:
651
Trophic status models and shoreline development policy

One of the significant breakthroughs achieved by the DillonRigler model and its variants
was their perceived ease of use and the accuracy of their predictions of water quality.
Although the model is conceptually simple, its application can be complex. The model is
supported by the results of approximately 25 years of detailed measurements on calibrated
watersheds in south-central Ontario, and these calibrations must be validated periodically.
The model was originally developed and calibrated on headwater lakes but, in practice,
it is used in a watershed context (Dillon and Rigler, 1975; Dillon et al., 1986). Watershed
modeling is conceptually straightforward, and the calculations eased by the use of personal computers, but implementation of policy in a watershed context, often between
municipal government boundaries, is very complex. The model must be supported by
monitoring programs to ensure the validity of its predictions. The management endpoints,
both for water quality and development capacity, must be substantiated and defensible.
Finally, implementation of a water quality program must be done by policy that is clear,
fair to all resource users, and defensible. Therefore, the conceptual simplicity of the model
may not be carried through into its application. This is reflected in four requirements for
managing shoreline development by trophic status:
1. An accurate and defensible model based on sound inputs and data sources and
defensible assumptions (The model must be able to distinguish natural sources of
phosphorus [which are not manageable] from anthropogenic sources of phosphorus [which are the intent of the management program])
2. A process of model calibration and operation to validate predictions against measured water quality
3. Water quality end points, expressed as both water quality and allowable limits of
development (as either anthropogenic phosphorus load or number of lots or development units)
4. An implementation strategy, formal planning instruments, and a process to guide
implementation and ensure fair allocation of capacity
Technical basis of the models

The water quality models used in Ontario are all mass balance models that predict the
trophic status of lakes at steady state with their phosphorus and hydrologic loadings. In
summary phosphorus loading from the atmosphere and from the watershed as a function
of soils and geology is linked with the lakes water load. Phosphorus loss from the water
column is modeled as a settling velocity, specific to whether the hypolimnion is oxic or anoxic.
The result is a prediction of natural or background phosphorus concentration. Human
influence is added in the form of phosphorus loading from septic systems servicing shoreline
residences, as a function of a per-capita phosphorus contribution of 800 gm/year, a count of
the number of shoreline residences and estimates of their usage (as capita years per year,
Dillon et al., 1986). A schematic of model operation is given in Figure II.17.2.
Although the mass balance principles of phosphorus loading and expression are
generic, the resultant predictions of phosphorus and chlorophyll a concentration, Secchi
depth transparency and hypolimnetic oxygen in the Ontario models were all derived from
regression relationships specific to the MuskokaHaliburton region of central Ontario
(Hutchinson et al., 1991; Dillon et al., 1994). Phosphorus export coefficients for the natural
landscape will also change as a function of wetlands, soil types, geology, and land use
652
Figure II.17.2 Schematic of water quality models used in Ontario. Lake trout habitat is not modeled
by municipalities, as fish habitat management is a federal government responsibility in Ontario.
(Dillon and Molot, 1997). Finally, the settling velocity for loss of phosphorus to lake
sediments used in the models is an average of local estimates made from calibrated
watersheds and lakes with oxic and anoxic hypolimnia (Dillon et al., 1986). In-lake phosphorus retention may vary regionally and from lake to lake. The use of these models must
always, therefore, be preceded by calibration to specific local conditions.
Three major assumptions underlie the use of the Ontario Trophic Status model:
1. The first is that 100% of the phosphorus loaded to a shoreline septic system will
ultimately be expressed as increased trophic status (Dillon et al., 1986, 1994). This
assumption has only been tested indirectly as a function of the fit of predicted with
measured phosphorus in study lakes (Hutchinson et al., 1991; Dillon et al., 1994).
Recent investigations of septic system geochemistry (Robertson et al., 1998) and
the mechanisms of phosphorus mineralization in soils (Isenbeck-Schroter et al.,
1993; Jenkins et al., 1971) suggest that this assumption is debatable where soils are
present between a septic system and a water body and that 100% phosphorus
export is, in fact, unlikely.
2. The second assumption is that all anthropogenic phosphorus sources within 300 m
of the lakeshore, or any inflowing tributary, must be included in the lakes phosphorus
budget (Dillon and Rigler, 1975; Dillon et al., 1986). Although this assumption is based
on the need to place boundaries on the inclusion of phosphorus sources, the distance
of 300 m is arbitrary and has neither been substantiated nor tested.
3. Finally, the lakes used to calibrate the Ontario model are small, headwater lakes (Dillon
et al., 1986). Although straightforward adaptation of the model to an entire watershed
is recommended (Dillon and Rigler, 1975; Dillon et al., 1986), published validations have
focussed only on headwater lakes (Dillon et al., 1994) or on simple watersheds with a
limited number of lakes (Hutchinson et al., 1991). The additional complexity inherent
in expanding the original concept and development allocation schemes to large watersheds must be addressed as part of any policy development process.
These assumptions form the basis of challenges to the concept of determining lakeshore capacity through trophic status (Michalski, 1994) and guided much of the technical
recasting of trophic status models reported herein.
In summary, trophic status models have been developed that can accurately predict
the responses of lakes to natural and anthropogenic phosphorus sources. The models are
regionally specific and must be validated against local conditions before they can be used
Chapter II.17:
653
with confidence. In addition, assumptions guiding anthropogenic loading estimates and

watershed implementation are open to some debate on both empirical and mechanistic
grounds and should be addressed before beginning any management exercise.
Model validation
The accuracy of predictions made by a trophic status model must be confirmed against
measurements of water quality in the subject lakes before the models can be used with
confidence in policy setting. This requires:
1. Establishment of a water quality monitoring program to determine existing levels
of phosphorus in lakes for comparison against present-day model predictions
2. Maintenance of the monitoring program for the long term to determine any trends
in water quality
3. Calibration of natural phosphorus loadings and basic model operation in undeveloped lakes, with no human phosphorus sources
4. Calibration of the model on developed lakes to determine if assumptions on
anthropogenic phosphorus sources are valid
5. A process to resolve inaccuracies in model predictions and to update the model
on the basis of monitoring results
Model validation must also meet the requirements of the planning process it supports.
The intent of a lake management program is to achieve stable and predictable water
quality. This must occur in concert with the requirements for a stable and consistent
planning and policy environment. Water quality programs should set stable targets for a
minimum of 20 years time to resolve new steady states in water quality in a monitoring
program and to provide a stable economic and planning environment, without jeopardizing the resource through over-allocation of development.
Monitoring programs
The monitoring program supporting a lakeshore capacity policy must strike a balance
between practical implementation, accuracy, and expense. Ontarios water quality models
were developed on the basis of an intensive, long-term research program on a small
number of study lakes which was undertaken by the provincial Ministry of the Environment in the Muskoka and Haliburton regions of Ontario. The program was based on
dedicated laboratory procedures and analytical staff, long-term personnel, and routine
scientific review. In contrast, a municipal program may have to be implemented with
limited financial support, summer or term staff, a variety of commercial or government
laboratory analyses, lack of in-house expertise, and staff who manage water quality only
as one aspect of their career. None of these, on its own, jeopardizes the integrity of a water
quality program, but all represent the potential for error and the need for stable policy
support. The water quality program must, therefore, be supported by a cost-effective
monitoring program that can be maintained for the long-term.
The water quality model is based on predictions of total phosphorus, and so the best
comparisons of model accuracy are obtained by measuring total phosphorus directly.
Published water quality relationships for MuskokaHaliburton lakes (Clark and Hutchinson, 1992) suggest that long-term trends in water quality can be determined by making
one phosphorus measurement each year at the time of spring overturn, when the lake is
completely mixed from top to bottom, thus reducing program costs. Accordingly, an
effective water quality monitoring program may consist of:
654

1. An annual spring overturn measurement of total phosphorus for comparison with
water quality predictions made by the model
2. Biweekly measurements of Secchi depth during the summer to track long-term
changes in water clarity, the recreational attribute that forms the basis of the water
quality program
3. An annual measurement of the dissolved oxygen profiles made at the end of
summer, when oxygen stress is most likely, to determine the oxygen status of lakes
for model input, to establish the suitability of aquatic habitat, and to track longterm changes
The results of the monitoring program can be used to calibrate the water quality model
and to identify those lakes for which the model does not produce accurate estimates of
trophic status.
Model calibration
The primary requirement for a water quality model, once accuracy has been demonstrated,
is that it be based on a solid mechanistic understanding of watershed and lake dynamics.
A purely empirical approach, in which understanding and technical substantiation are
ignored and the model fit is the only rationale for model acceptance, may not be
technically defensible. The coefficients and assumptions that make up the predicted water
quality must be clear and documented, as they will form the basis of challenges to the
model. Clear technical rationale is also required so that all model users can understand
the model and improve it in the future. Wherever possible, the water quality model should
be substantiated by reference to the primary scientific literature.
The Ontario Lakeshore Capacity Study calibrated its trophic status model (Dillon et
al., 1986) on research lakes that had no shoreline development. This was done in order to
quantify basic lake processes in the absence of the additional uncertainty inherent in
assumptions of human phosphorus loading from shoreline septic systems. While this is
mechanistically sound and represents the ideal approach, it may not be practical for
municipalities or other jurisdictions. Water quality programs are focussed on those lakes
where shoreline development is present, because they are the lakes that require management activities such as development capacities. Uninhabited lakes are not generally monitored as part of a water quality program, posing difficulties in calibrating basic model
elements such as settling velocity, natural watershed loads, or in-lake retention. These
factors must be well quantified before considerations of the uncertainties in quantifying
human phosphorus loadings.
Where the number of uninhabited lakes in the calibration set is not considered adequate for model development, then the calibration exercise should include lakes in which
only limited shoreline development is present. A cutoff where less than 10% of the total
potential load is added from shoreline development may suffice in these cases. Measurements of water quality may be available for larger numbers of these sparsely inhabited
lakes to assist with model calibration.
Calibrating natural phosphorus sources. Consideration of two factors is recommended to improve model fit for natural phosphorus loading.
Incorporation of wetlands to describe natural phosphorus export. Recent research on
south-central Ontario watersheds showed that phosphorus export from wetlands (plus
atmospheric deposition) determined natural phosphorus loading to a lake. This was stated
as the following estimate of phosphorus load (from Dillon and Molot, 1997):
Chapter II.17:
655
Figure II.17.3 Influence of dissolved organic carbon (DOC) on average long-term total phosphorus
concentrations in Precambrian Shield lakes.
Figure II.17.4 Relationship of average long-term measured phosphorus in Precambrian Shield lakes
with wetland area in catchment.
kg TP/year = catchment area (km2) * (3.05 + (0.54 * % wetland))
This relationship is driven by the export of phosphorus with dissolved organic carbon
from wetlands in the catchments of the lakes. This is shown for lakes of the Muskoka
River watershed in Figure II.17.3. Total phosphorus concentrations were significantly
related (p < 0.000001, r2 = 0.39) to the amount of wetland in the catchments of lakes in the
Muskoka River watershed (Figure II.17.4.). Natural phosphorus loading from all catchments containing wetland can therefore be estimated from wetland area.
Phosphorus retention in shallow lakes. Both the DillonRigler and Lakeshore Capacity
Study models were developed and calibrated for use in lakes that are deep enough to
656
stratify. The models are not intended for use in shallow lakes, in which phosphorus does
not settle to hypolimnetic waters. The major requirement for modeling a shallow lake is
to devise an accurate estimate of in-lake retention of phosphorus. The models estimate
retention as a function of the areal water load (m3 of runoff/year per m2 of lake surface)
and the settling velocity of phosphorus, as modified by the hypolimnetic oxygen status
(Dillon et al., 1986). For shallow lakes, this method overestimates retention as it does not
allow for wind-driven resuspension of phosphorus.
Several methods of estimating retention in shallow lakes can be attempted, most of
which involve modification of the apparent settling velocity. In the end, shallow lakes are
best calibrated individually. In-lake phosphorus retention can be modified to achieve a
best fit between measured and predicted estimates of total phosphorus concentration.
Calibration of the many shallow lakes present in the watershed, most of which have no
shoreline development, provides the final improvement to the model and, together with
the other modifications, provides the confidence in model accuracy to allow its use for
setting development capacities.
Atmospheric phosphorus loading. Trophic status models must include a phosphorus contribution from the atmosphere directly to the lake surface. Atmospheric loading includes contributions from precipitation and from dry loading (or dust). Values such as 20.7 mg/m2/year for
the Precambrian Shield in south-central Ontario can be obtained from published monitoring
studies, such as those of the Ontario Ministry of the Environment (Dillon et al., 1992).
Hydrologic loading. Trophic status models require input of hydrologic loading for
each lake. The models are long-term steady-state models and so are cast to use average
annual depth of runoff. These values can be obtained from published figures (Canada
Dept. of Fisheries and Environ., 1978). They can then be coded into the model in a lookup table and determined for each lake on the basis of latitude and longitude of the lake,
as entered by the user. They should be reviewed periodically, however, in light of the
potential changes in runoff stimulated by climate change.
Calibrating anthropogenic phosphorus sources. Consideration of wetlands and shallow lakes improves the fit of the model so that measured and modeled phosphorus
concentrations do not vary in undeveloped lakes. Review of model results for developed
lakes showed that the measured phosphorus concentrations were lower than the concentrations predicted by the model but higher than those predicted for a case where no
development was present. From this, it can be concluded that some portion of the potential
anthropogenic phosphorus load is being expressed in the lake but that the model must
be recalibrated to account for the portion of the development load that is not expressed.
Several modifications are required to reduce the large positive bias (overprediction of
measured phosphorus) observed in developed lakes.
Incorporation of phosphorus retention by soils. The water quality model for lakes in
the Muskoka River watershed incorporates a substantial departure from other Ontario
models in its geochemical assumptions regarding phosphorus movement from septic
systems to lakes. Both the original Dillon and Rigler (1975) and Ontario models (Dillon
et al., 1986) assumed that all septic system phosphorus generated within 300 m of the
shoreline would ultimately migrate to the lake. This assumption may be considered
reasonable as a conservative approach but has never been tested directly.
Since the publication of the original models, direct monitoring studies and mechanistic
understanding of soil and phosphate interactions have provided evidence that conflicts
with the original assumptions. Mechanistic evidence (Stumm and Morgan, 1970; Jenkins
Chapter II.17:
657
Figure II.17.5 Quaternary Geology of Muskoka, Ontario (from Barnett et al., 1991). Pink areas (1)
denote igneous and metamorphic bedrock, which is exposed or covered with thin drift. Blue areas
(24) are glaciolacustrine deposits of clay and silt. Remaining colors are glaciolacustrine deposits of
outwash and till, sand and gravel (18, 23, 25) or organic peat and muck (32).
et al., 1971; Isenbeck-Schroter et al., 1993) and direct observations made in septic systems
(Willman et al., 1981; Zanini et al., 1997; Robertson et al., 1998) all show strong adsorption
of phosphate on charged soil surfaces and mineralization of phosphate with Fe and Al in
soil. The mineralization reactions, in particular, appear to be favored in acidic and mineralrich groundwater in Precambrian Shield settings (Robertson et al., 1998), such that over
90% of septic phosphorus may be immobilized. The mineralization reactions appear to be
permanent (Isenbeck-Schroter et al., 1993), and direct observations suggest that most septic
phosphorus may be stable within 0.5 m of the tile drains in a septic field on the Precambrian Shield (Robertson et al., 1998).
The mechanistic and geochemical evidence is supported, in part, by trophic status
modeling. Dillon et al. (1994) reported that only 26% of the potential loading of phosphorus
from septic systems around Harp Lake, Muskoka, could be accounted for as measured
phosphorus in the lake. The authors attributed the variance between measured and modeled estimates of phosphorus to retention of septic phosphorus in thick tills in the catchment of Harp Lake. Although the Muskoka watershed is frequently characterized as an
area of thin to no soils over bedrock, this description is in no way universal. The central
corridor of the watershed (in which Harp Lake is located) occupies a glacial outwash plain
of alluvial sands and gravels, and many catchments contain substantial soil deposits.
Western and southwestern Muskoka represent the more typical topography of thin soils
and granite ridges and outcrops (Figure II.17.5). Even in these areas, however, tile fields
are often, by necessity, built on imported fill and so some attenuation is possible.
Revisions to trophic status models should use the findings of these recent studies to
improve the positive bias (i.e., overprediction of measured phosphorus) in the model by
accounting for a 74% retention of septic phosphorus for those lakes with suitable soils in
their catchments (Dillon et al., 1994). The positive bias is apparent in model results for all
developed lakes, but most pronounced in heavily developed lakes. All of the study lakes
658
can be located upon a map of surficial geology (i.e., Southern Ontario Engineering Geology
Terrain Study; Mollard, 1980). Those lakes situated within the same types of thick till as
are found around Harp Lake, and those lakes situated within outwash or alluvial plains
can also be assumed to retain 74% of their septic phosphorus in soils without expression
in the lake. The model should therefore be coded to reduce the septic phosphorus contribution by 74% for those lakes. Soil classifications (from Mollard, 1980) for which 74%
retention can be assumed are:
MG/R = ground moraine over bedrock
LD
= glaciolacustrine delta
GO
= outwash plain
Adoption of soil retention provides substantial improvement to the fit between measured and modeled estimates of phosphorus concentration in recreational lakes. The
estimate of 74% may still be considered conservative, as Robertson et al. (1998) and Wood
(1993) both reported retentions well in excess of 90% for septic systems located in the
Precambrian Shield in general and at Harp Lake, specifically.
Distance of development from shorelines. Much of the recreational shoreline development in Ontario is located in one tier around a lakeshore, with a minimum setback of
15 to 30 m between the septic system and the waters edge. As development intensifies,
however, and prime shoreline space is occupied, development may occur in a second or
third tier back from the lakeshore, with water access from communal docking or bathing
facilities. Although it is reasonable to assume that second- and third-tier development will
have some influence on water quality, it is unlikely that it will have the same influence
as development located on the shoreline. The water quality model, therefore, must find a
means of accommodating distance of development from a lake while still including some
contribution of phosphorus to the lake.
The DillonRigler and Ontario models both adopted a convention of assuming that
100% of the septic system phosphorus generated within 300 m of a lake would ultimately
migrate to the lake. Although it is a useful figure to define the limits of a modeling exercise,
it is very difficult to defend technically, given the knowledge of phosphorus geochemistry
described above. It also leads to counterintuitive interpretations, in which a septic system
located 299 m from the shore has a 100% impact, while one located 301 m back has no
impact. The absence of substantiation or a mechanistic basis for the assumption, and the
complexity involved in site-specific derivation for a watershed modeling exercise, make
it very difficult, however, to propose a valid alternative approach.
Setback of development can be accommodated by modifying the assumption of phosphorus contribution with distance; for example, by coding the model so that:
1. Development within 100 m of the shoreline provides 100% contribution of septic
phosphorus (as modified by soil thickness and type, see above)
2. Development between 100 and 200 m has its phosphorus contribution reduced by
one third
3. Development between 200 and 300 m has its phosphorus contribution reduced by
two thirds
4. Development beyond 300 m has no phosphorus contribution
Although this approach still suffers from arbitrary distinctions it does accommodate the
concept of distance of development from the shoreline and phosphorus attenuation by soils.
Chapter II.17:
659
Table II.17.1 Usage of Shoreline Residences

in the District Municipality of Muskoka
Zone 1: Outlying area
Zone 2: Close to major highway
Zone 3: Close to major urban center
Resort unit usage
Trailer and camp sites
0.82
1.23
2.09
1.23
0.41
All values are given in capita years per year for

each residential type.
Revised per-capita phosphorus contribution. The original DillonRigler and Ontario

models used a figure of 800 gm/C/year as an estimate of per-capita contributions of
phosphorus to septic systems from human waste and household cleaning products. This
figure was originally derived, in part, from measurements of total phosphorus in septic
tanks (Dillon et al., 1986) made between 1965 and 1980. Phosphorus concentrations ranged
from 5 to 21.8 mg/L, with a mean of 13.2 mg/L. More recent research conducted by the
Ontario Ministry of the Environment. (Gartner Lee Limited, 2002, in preparation) found
a range of 4.3 to 13.3 mg/L of total phosphorus in septic tanks serving shoreline residences
and an overall average of 8.2 mg/L. The more recent values reflect the limitation of
phosphate in laundry detergent in the early 1970s and represent 62% of the phosphorus
concentrations used in Ontarios Lakeshore Capacity Study. Strict application of the
reduced concentration produces an estimate of 500 gm/C/year (i.e., 800 * 0.62). No figure
will be completely accurate, however, and so 100 gm/year was added to the measured
value in order to maintain a protective and conservative approach to estimating phosphorus loadings. A water quality model should therefore consider a per-capita phosphorus
contribution of 600 gm/year to the septic system.
Shoreline development also adds phosphorus to a lake from the conversion of a
forested landscape by clearing and lawn planting, and hardening of soils. Previous Ontario
models did not include a contribution from these sources. A model should account for
this clearing by including an estimate of 2000 m2 for the average size of the developed
portion of each shoreline lot and an increased export coefficient of 17 mg/m2/year ( = 34
gm/lot/year) from those areas.
Validation of cottage usage figures. The final requirement for estimating human
phosphorus loadings from shoreline development is to obtain estimates of the number
of days that residences are used in a year. Seasonally occupied residences will be
occupied for fewer days per year than permanent homes, and different usage figures
apply to resort units or trailer sites. In the 1970s, the Ontario Lakeshore Capacity Study
estimated seasonal and permanent usage of shoreline residences figures as 0.89 and 2.55
capita years/year, respectively, on the basis of a cottage survey conducted in the
MuskokaHaliburton region of Ontario (Downing, 1986). Twenty years later, the District
Municipality of Muskoka (1995) undertook a similar study and determined that the overall
lot usage had not changed substantially (Table II.17.1). Usage figures did not support the
commonly held perception that large numbers of cottages were converting from seasonal
to permanent use to accommodate retirees or telecommuters. Only two lakes had high
proportions of permanent residents, and these were close to a major urban center (the
town of Huntsville). Regionally specific lot usage surveys should accompany development of lake trophic status models to ensure accuracy.
660
Lake Vernon
61 lakes
Lake Rosseau
39 lakes
Lake Joseph
32 lakes
Trading Bay
43 lakes
Dwight Bay
37 lakes
Mary Lake
32 lakes
Lake of Bays
25 lakes
N Muskoka River
22 lakes
S. Muskoka River
31 lakes
Lake Muskoka
32 lakes
Muskoka River Watershed

12 sub watersheds
376 modelled lakes
124 monitored lakes
Moon River
43 lakes
Georgian Bay
Figure II.17.6 Schematic of Muskoka River Watershed.
Other phosphorus sources. The model should also include anthropogenic sources of
phosphorus from other activities in the watershed:
1. Loadings from sewage treatment plants which can be obtained from operating
records.
2. Areas of agricultural land use can be determined from air photo interpretation. A
coefficient of 45 mg/m2/year can be used to describe export from agricultural land
use (Winter and Duthie, 2000).
3. Urban runoff of 45 mg/m2/year for those portions of urban areas within 300 m of
a lake or a river. These loadings are not reduced with setback distance, as urban
areas are hardened and offer little opportunity for attenuation.
Final model results

The Muskoka River drains a watershed with a total area of 5,737 km2. The watershed
model is built around 12 subwatersheds of the Muskoka River, as defined by the major
river or lake at the downstream end of each (Figure II.17.6). The hydrologic and nutrient
output from each subwatershed enters the next subwatershed downstream, from the
western fringes of Algonquin Park at the furthest extent upstream, to Georgian Bay of
Lake Huron. The total area of surface water included in the model is 56,000 ha, and the
size of modeled lake segments ranges from 3 to 6,200 ha.
The model is both large and complex. It provides explicit characteristics and model
output for 376 individual lakes, rivers, or portions of each. The model contains nearly
12,500 individual calculations or data points to calculate total phosphorus concentrations,
and approximately 15,000 additional calculations to estimate and track development
capacities.
Model accuracy
The final version of the Muskoka Watershed water quality model produced excellent
estimates of total phosphorus concentrations in 123 lakes for which monitoring data were
available. Mean negative bias in predictions (i.e., underestimate of measured water quality)
occurred in 3 lakes and was within 1% of measured values (Table II.17.2). The median
Chapter II.17:
661
Table II.17.2 Summary Statistics Showing the Percentage Agreement between

Modeled and Measured Estimates of Water Quality in the Muskoka Watershed Model
Average error
Median error
No. lakes
No. > 40% error
All Lakes
+Error
Error
D.I. < 1.21

+Error
Error
D.I. 1.11
+Error
Error
14.74
8.08
120
9
5.89
2.81
47
1
2.83
2.34
32
0
0.84
0.51
3
0
0.28
0.28
1
0
0.28
0.28
1
0
overestimate was 8.1% in 120 lakes. The average overestimate of 14.7% included 9 lakes
for which the bias exceeded 40%. The criterion of 40% was chosen as the mean coefficient
of variation in measured phosphorus concentration for all lakes. Model errors in excess
of 40% were considered unacceptable as they were outside the range of natural variance
in water quality. These errors only occurred as a result of positive bias, however, as an
indication that not all of the potential human phosphorus load modeled was actually
expressed in the lake.
Model accuracy was further expressed by comparison of model error with the proportion of the total phosphorus load contributed by human sources. These sources were
described using the development index (DI), which was the ratio of total phosphorus
load to potential anthropogenic phosphorus load for each lake. A DI of 1.0 represents a
lake with no anthropogenic loading, a DI of 1.5 denotes a 50% increase, 2.0 a doubling,
and so on.
For 32 undeveloped lakes (DI < 1.11) the average overprediction was <3%. Positive
bias increased with development intensity such that median and average error were 2.8
and 5.9%, respectively, for 47 moderately developed lakes (DI < 1.21, Table II.17.2). A
scatterplot comparison of measured and modeled phosphorus concentrations
(Figure II.17.7) shows a) the good correspondence between the two and (b) the tendency
for the model to overpredict phosphorus concentration. The positive bias persisted, even
after accounting for attenuation of septic phosphorus (see Model Validation section
above).
The absolute difference between measured and modeled phosphorus was <1 g/L
for 75 lakes (Figure II.17.8), and error for 77% of the lakes was within 2 g/L. All of these
errors represent positive bias, however, such that the model has a strong tendency to
overestimate phosphorus concentrations in developed lakes.
The positive model bias was related to the estimate of phosphorus loads from shoreline
development. Median and average positive bias increased with the intensity of development
(DI, Table II.17.2), but negative model bias did not change. Positive bias in excess of 10% was
confined to developed lakes. The trend to overprediction appeared on undeveloped and
sparsely developed lakes but was less than 5% and likely reflected model variance.
The positive bias was reduced by accounting for the retention of phosphorus in soils
(see Model Validation). Average and median errors were 9.7 and 5.7%, respectively, for
the 66 lakes in which 74% phosphorus retention was assumed on account of the soil
characteristics. Error only exceeded 40% on two of these lakes. For lakes where retentive
soil mantles were not assumed in the model, average and median error increased to 21.2
and 13.0%, respectively, and 8 lakes had positive bias above 40%. This suggests that some
retention may be occurring around all lakes and not just those with thick soil mantles in
their catchments.
In spite of the positive bias in model results, it is clear that shoreline development has
influenced the phosphorus concentrations in lakes to some extent. Phosphorus concentrations have increased with development, but by far less than predicted on the basis of an
662

Accuracy of Muskoka River Watershed Model
Modelled TP in g/L
25
20
15
10
5
0
0
10
15
20
25
Measured Total TP in g/L

Figure II.17.7 Accuracy of water quality predictions made with final version of the Muskoka watershed water quality model. Thick line shows perfect fit (1:1), and narrow lines enclose accuracies of
40%.
Absolute Error of Model Prediction for 123 Lakes

80
No. of Lakes
70
60
50
40
30
20
10
0
Variance in g/L TP
Figure II.17.8 Absolute difference (in g/L) between modeled and measured estimates of total
phosphorus concentrations in Precambrian Shield lakes.
assumption that 100% of all septic phosphorus is mobile. This observation was supported
by the modeling exercise, by the detailed observations of phosphate mobility in septic
systems summarized in Robertson et al. (1998) and by geochemical descriptions of phosphate behavior (Isenbeck-Schroter et al., 1993; Stumm and Morgan, 1970). The setting of
water quality objectives and development limits must account for these understandings
in phosphate behavior and be aware that significant overestimates of the impacts of
shoreline development may hinder the implementation and defense of water qualitybased development limits.
Chapter II.17:
663
Setting water quality objectives and development limits

The intent of a water quality program is to use the monitoring and modeling exercises to
support water quality-based shoreline development capacities. Some review of approaches
to setting phosphorus objectives is therefore warranted.
Surface-water management in Ontario

The Ontario Ministry of the Environment (MOE) manages environmental quality primarily through two pieces of provincial legislation: The Environmental Protection Act and
The Ontario Water Resources Act. Policies and procedures for management of surface
water quality that arise from this legislation are elaborated in implementation documents
such as Water Management: Policies, Guidelines, Provincial Water Quality Objectives of the
Ministry of Environment and Energy (OMEE, 1994). The goal of surface water management
in Ontario is to ensure that the surface waters of the province are of a quality which is satisfactory
for aquatic life and recreation (OMEE, 1994).
Ontario established Provincial Water Quality Objectives (PWQOs) in the 1970s in order
to meet this goal. The first objectives were mostly adopted from other agencies, such as
The International Joint Commission, but were later developed in Ontario (OMEE, 1992).
Provincial Water Quality Objectives (PWQOs) are numerical and narrative
ambient surface water quality criteria. They are applicable to all waters of the
province (e.g., lakes, rivers and streams) except in those areas influenced by
MOEE approved point source discharges. In specific instances where groundwater is discharged to surface waters, PWQOs may also be applied to the
groundwater. PWQOs represent a desirable level of water quality that the
MOEE strives to maintain in the surface waters of the province. In accordance
with the goals and policies in Water Management (OMEE, 1994), PWQOs
are set at a level of water quality which is protective of all forms of aquatic
life and all aspects of the aquatic life cycle during indefinite exposure to the
water. The Objectives for protection of recreational water uses are based on
public health and aesthetic considerations (OMEE, 1994).
Two policies are used to interpret the water management goal and application of the
PWQOs to specific water bodies (OMEE, 1994).
Policy 1: In areas which have water quality better than the Provincial
Water Quality Objectives, water quality shall be maintained at or above
the Objectives. Although some lowering of water quality is permissible in
these areas, degradation below the Provincial Water Quality Objectives
will not be allowed, ensuring continuing protection of aquatic communities
and recreational uses.
Policy 2: Water quality which presently does not meet the Provincial Water
Quality Objectives shall not be further degraded and all practical measures
shall be taken to upgrade the water quality to the Objectives.
Municipal responsibilities for water quality in recreational lakes

The planning system in Ontario is established in the Planning Act, which provides the
legislative jurisdiction for municipalities to monitor and regulate land use subject to policy
664
provided in the 1997 Provincial Policy Statement. The Planning Act requires that in reviewing any planning application, a municipality must have regard to provincial policy. The
provincial policy respecting water quality states:
The quality and quantity of groundwater and surface water and the
function of sensitive groundwater recharge/discharge areas, aquifers and
headwaters will be protected or enhanced.
In practice, municipal policy for water quality protection is interpreted against Ontario
PWQOs, and against Policies 1 and 2 for protection of surface water quality, as outlined
above. Municipalities are allowed to be flexible in their specific interpretations but must
meet the intent of provincial policy. Municipalities can make use of trophic status models
to develop shoreline development objectives, which can then be lodged in official plan
documents.
Where a development objective exists, policy may then direct that new development
shall not result in the predicted nutrient level exceeding the objective for that particular
lake. Where a lake is at capacity, development or lot creation may only be permitted in
limited circumstance, and in particular where:
1. There is an existing vacant lot of record.
2. The redevelopment of the property would not result in an increase of phosphorus
loading to the water body.
3. The septic system and leaching bed can be set back from the waterbody by 300 m.
4. The septic system and leaching bed can be placed in another watershed that is not
at capacity
More general policy should also encourage the maintenance of shoreline vegetation,
the restoration and preservation of the waterfront shoreline where it has been artificially
altered, and aesthetic factors such as setback, footprint or coverage of built structures. Site
Plan Control is a planning instrument which can be used to implement site-specific
development details.
Ontarios Planning Act requires that municipal official plans be reviewed from time
to time to ensure that they continue to provide appropriate policy direction. Such review
should include the need to update water quality objectives and lake models to ensure
that the most recent scientific understanding of nutrient loading is properly incorporated.
Ontarios existing PWQO for total phosphorus

The existing PWQO for total phosphorus was developed in the late 1970s (OMEE, 1979).
It drew on the trophic status classification scheme of Dillon and Rigler (1975) to protect
against aesthetic deterioration and nuisance concentrations of algae in lakes, and excessive
plant growth in rivers and streams. The rationale (OMEE, 1979) acknowledges that elemental phosphorus can be toxic but that it is rare in nature and so toxicity is rarely of
concern. (In fact, there is only one documented case of elemental phosphorus poisoning
an aquatic (marine) system in Canada.) Instead, the purpose of the objective was to protect
the aquatic ecosystem from nontoxic forms of phosphorus: phosphorus must be controlled,
however, to prevent any undesirable changes in the aquatic ecosystem due to increased algal growth
(OMEE, 1979).
Chapter II.17:
665
The 1979 PWQO was given the status of guideline to reflect the uncertainty about the effects
of phosphorus and to acknowledge the difference between managing toxic and nontoxic pollutants.
Current scientific evidence is insufficient to develop a firm objective at
this time. Accordingly, the following phosphorus concentrations should be
considered as general guidelines which should be supplemented by sitespecific studies:
To avoid nuisance concentrations of algae in lakes, average total phosphorus
concentrations for the ice-free period should not exceed 20 g/L;
A high level of protection against aesthetic deterioration will be provided
by a total phosphorus concentration for the ice-free period of 10 g/L or
less. This should apply to all lakes naturally below this value;
Excessive plant growth in rivers and streams should be eliminated at a
total phosphorus concentration below 30 g/L.
Total phosphorus and the PWQO development process

There are several shortcomings with Ontarios existing PWQO for total phosphorus and
the province has been reviewing its approach to phosphorus management. The approach
is derived from that first proposed in Hutchinson et al. (1991), and elements of it are
summarized in this section. These can be considered in cases where a municipality wishes
to derive its own phosphorus objectives to assist with managing shoreline development.
Phosphorus as a pollutant
Development of a water quality objective for total phosphorus is distinctly different from
that for toxic substances. Most aquatic pollutants are directly toxic to some target tissue,
such as the fish gill, even if some of them are required nutrients at trace amounts, i.e.,
copper or zinc. As a result, the health of aquatic organisms, and hence the ecosystem,
declines rapidly at concentrations slightly above ambient levels (Figure II.17.9). Phosphorus, on the other hand, is a major nutrient. Concentrations can increase substantially with no
direct toxic effects. In fact, the first response of the aquatic system is increased productivity
and biomass. Beyond a certain point, however, indirect detrimental effects become apparent, which ultimately decrease system health.
The first detrimental responses of a lake to enrichment (i.e., water clarity, algal blooms)
are aesthetic and of concern mostly to humans. Assessment of aesthetic impacts is highly
subjective; perceived changes in water clarity are based largely on what one is used to
(Heiskary and Walker, 1988). The development of a phosphorus objective must therefore
acknowledge an element of subjectivity in dealing with human concerns. The objectivedevelopment process may also consider that aesthetic impacts begin where a change in
water clarity is first noticeable to the human eye, or where the mean water clarity first
exceeds natural variation. Unfortunately, human perception of water clarity has not been
established. Existing guidelines are based on trophic status classification schemes. They do
not consider other water clarity influences such as inorganic turbidity or dissolved organic
carbon or how lake users perceive changes in water clarity.
Finally, trophic status indicators such as water clarity, chlorophyll a, or dissolved
oxygen cannot be managed directly, but only through management of phosphorus. In
addition, there may be delays of up to decades between the addition of phosphorus sources
666
Figure II.17.9 Generalized responses of an ecosystem to toxic and nontoxic pollutants.
to a watershed (i.e., septic systems), its movement from the source to surface water
(Robertson, 1998), and its expression as a change in trophic status. As a result, phosphorus
management in Ontario requires the extensive use of models relating shoreline development to the trophic status of the receiving water. Phosphorus management may therefore
be considered as a process of predicting the predictor. Previous sections of this chapter
have emphasized the importance of validating or accounting for these assumptions.
The shortcomings of a numeric phosphorus objective

Ontarios existing two-tiered numeric objectives for total phosphorus obscure fundamental
differences between lake types and their nutrient status in the absence of human impact.
Most Precambrian Shield lakes are characterized by excellent water quality as represented
by low concentrations of total phosphorus. For example, the average of the mean annual
measured concentration for 123 lakes in the Muskoka watershed was 9.3 4.2 g/L and 62%
of the lakes averaged 10 g/L or less (Figure II.17.10). Within any set of lakes, however, there
is still a large diversity of water clarity, controlled by both total phosphorus concentrations
and dissolved organic carbon.
The provincial phosphorus objective allows lakes that currently contain less than 10 or
20 g/L of phosphorus to increase to a maximum of 10 or 20 g/L. The logical outcome of
this two-tiered objective is that, over time, all recreational waters would converge on one or
the other of the water quality objectives. This would produce a cluster of lakes slightly below
10 g/L and another slightly below 20 g/L, decreasing the existing diversity in water quality
in lakes and, with it, the diversity of their associated aquatic communities.
The second shortcoming is that, over time, some lakes would sustain unacceptable
changes in water quality while others would be unimpacted, producing both ecological and
economic asymmetries as the resource was developed. A lake with a natural phosphorus
concentration of 4 g/L is a fundamentally different lake than one that exists at 9 g/L. Both
lakes, however, would be allowed to increase to 10 g/L under the existing PWQO. One lake
would experience no perceptible change (9 to 10 g/L) and be overprotected, but the other
(4 to 10 g/L) would be underprotected and change dramatically. In both cases, human
perceptions of aesthetics are ignored in the objective. Allocation of phosphorus loadings
between these two lakes would be unfair as well. The higher-phosphorus lake could
Chapter II.17:
667
Distribution of Total Phosphorus in 123 Precambrian

Shield Lakes
40
Number of Lakes
35
30
25
20
15
10
5
0
4
10
12
14
16
Total P in g/L
18
20
>22
Figure II.17.10 Distribution of total phosphorus measurements in 123 lakes in the Muskoka River
watershed, 19901998.
sustain a greater change than the low-phosphorus lake but would be restrained to a much
lower load.
In summary, the existing two-tiered numeric objectives overprotect some lakes
and do not protect others adequately. Allocation of phosphorus loadings is unnecessarily restricted in some lakes and overly generous in others. Neither biotic nor
aesthetic attributes are adequately protected. Over time the diversity of trophic status
presently represented in Ontario will decrease as lakes converge on one of two numeric
objectives.
Environmental baselines and measured water quality

An emerging concern in environmental assessment is the need for a standard baseline for
comparison against environmental change. Inland lakes respond quickly to point source
phosphorus inputs. Detection of change is much more difficult, however, for nonpoint
sources such as leachate from domestic septic systems.
The incremental nature of shoreline development (no lake is ever developed all at
once) results in a slow and gradual increase in trophic status. The high degree of seasonal
and annual variance in phosphorus levels in lakes (Clark and Hutchinson, 1992) means
that changes may not be detectable without an intensive monitoring program, based on
many samples and a precise and replicable analytical method.
Human observers may not notice a slow increase in trophic status over a generation.
Environmental change that occurs over one generation becomes the status quo for the
next. Over a long period, therefore, any assessment baseline based on measurements of
total phosphorus will increase.
Any phosphorus objective that relies exclusively on measured water quality will
therefore suffer from:
Detection problems due to natural variance and analytical problems
The lag time between addition of phosphorus to a watershed and its expression
in a lake
Failure to detect incremental changes in water quality
Human perceptual conditioning, which reduces the apparent change in water
quality over time
668
As a result, an increasing assessment baseline and incremental increases in nutrient

loading may slowly degrade water quality past any objective. Impacts will accumulate
by virtue of delay in their expression, repetition over time and space, extension of the
impact boundary by downstream transport, or by triggering indirect changes in the
system, such as anoxic sediment release. Nonpoint source phosphorus pollution, particularly from septic systems serving shoreline development, is thus an excellent example
of a pollutant that may produce cumulative impacts to the aquatic environment.
An ecologically sound approach to objective setting

The emergence and validation of mass balance phosphorus models for lakes offers an
opportunity to correct some of the disadvantages of water quality measurements and
conventional assessment techniques. In the past, these models were used to establish the
amounts of development that would maintain trophic status within the numeric objectives.
They were used after the fact, to implement a water quality objective. The recommended
approach would be to use the water quality model itself to set lake-specific phosphorus
objectives.
The merits of a modeled assessment baseline

The basis of the revised approach is increased reliance on water quality modeling in the
objective setting process. Recent advances in trophic status models allow us to calculate
the predevelopment phosphorus concentrations of inland lakes (Hutchinson et al., 1991).
This is done by:
1. Modeling the total phosphorus budget for the lake
2. Comparing the predicted concentration to a reliable water quality measurement
to validate the modeled result
3. Subtracting that portion of the budget attributable to human activities
The main advantage of the modeling approach is establishment of a constant assessment baseline. A modeled predevelopment baseline is based on an undeveloped watershed and so will not change over time. This serves as the starting point for all future
assessments. Every generation of water quality managers will therefore have the same
starting point for their decisions, instead of a steadily increasing baseline of phosphorus
measurements.
The new approach therefore proposes phosphorus objectives based on modeled predevelopment phosphorus concentrations. This will provide water quality managers with:
1. A constant assessment baseline
2. A buffer against incremental loss of water quality
3. A buffer against variable water quality measurements
The predevelopment phosphorus concentration should not be interpreted as the
objective itself. Pristine or predevelopment phosphorus levels have not existed in
Ontarios Precambrian Shield lakes for over a century, and their attainment is not costeffective on heavily developed lakes. The modeled predevelopment concentration only
serves as the starting point for the objective and a reference point for future changes.
The degree of lake development can also guide the selection of a modeled baseline for
use in objective derivation. The recreational settlement of lakes in Ontario began in the late
1800s and was very well advanced when trophic status models emerged 100 years later. There
was no way to confirm how much of the phosphorus from a century of development was
Chapter II.17:
669
already being expressed in the lakes, and new development was added every year. Measured
phosphorus concentrations were therefore higher than the natural baseline but potentially
lower than their final, steady-state levels. The natural variability in phosphorus concentrations, and the lag time before the expression of septic-derived phosphorus (as discussed
above), prevents the use of measured values as a planning baseline. Modeled, predevelopment baselines of phosphorus concentrations, validated against local lakes which are not yet
developed, are therefore recommended as the starting point for objective development.
A model-based objective has two additional advantages. First, the modeled response
of the watershed to future changes is instantaneous. It applies new development directly
against capacity, without the intervening decades it takes for phosphorus to move to a
lake and be expressed as a measured change in water quality (this approach, however,
also requires assumptions on the ultimate mobility of phosphorus, which may not be valid
(see Calibrating Anthropogenic Phosphorus Sources)). Second, the trophic status model
is based on entire watersheds and so allows explicit consideration of downstream phosphorus transport in the assessment.
One disadvantage of the model-based baseline, however, lies with the inevitable
changes in scientific understanding of lakes and watersheds. Any baseline derived from
a water quality model is therefore subject to change as improved understanding or refinement produces changes in export coefficients, atmospheric deposition, or quantification
of in-lake dynamics. For this reason, a modeling exercise must begin with locally validated
coefficients and calibration of the water quality model. Model improvements should also
be implemented within a defined schedule of Official Plan review, so that scientific understanding is incorporated at a schedule consistent with measured responses of lakes and
the planning process.
The merits of a proportional increase

The second component of the objective is a proportional increase from the modeled
predevelopment baseline. The proportional increase accommodates regional variation in
natural or background water quality through the use of one numeric objective for all
lakes. It is, in fact, a broader, yet simpler, application of the regionally specific, multitiered
objectives proposed in other jurisdictions as a means of accommodating regional variation
in background water quality (e.g., Heiskary and Walker, 1988).
One consideration is to adopt an allowable phosphorus increase of 50% above the
modeled predevelopment level from anthropogenic phosphorus sources (Hutchinson et al.
1991). This approach is being considered by the Province of Ontario. Under this proposal,
a lake modeled to a predevelopment phosphorus concentration of 4 g/L would be
allowed to increase to 6 g/L from shoreline development or other human activities.
Predevelopment concentrations of 6, 10, or 12 g/L would increase to 9, 15, or 18 g/L,
respectively. A cap at 20 g/L would still be maintained to protect against nuisance algal
blooms.
There are numerous advantages to this approach.
1. Each waterbody would have its own water quality objective, but this could be
described with one number (i.e., predevelopment plus 50%).
2. Development capacity and ultimate phosphorus status would be proportional to
a lakes original trophic status.
3. As a result, each lake would maintain its original trophic status classification. A
4-g/L lake would be developed to 6 g/L and therefore maintain its distinction as
oligotrophic. A 9-g/L lake would be developed to 13.5 g/L, would maintain its
mesotrophic status, and development would not be unnecessarily constrained to
10 g/L.
670

4. The existing diversity of trophic status in a region such as Ontario would therefore
be maintained over the long term, instead of the ultimate outcome of a set of lakes
at 10 g/L and another at 20 g/L.
Alternative water quality endpoints and objectives

Several other potential water quality objectives can also be considered in order to establish
reasonable and defensible recommendations for development limits in recreational lakes.
There is no one obvious best water quality objective. The uncertainty in objective setting
for phosphorus and its status as a nontoxic pollutant (see Phosphorus as a Pollutant) mean
that there will be an element of subjectivity in any figure that is ultimately used.
Background + 40%. One refinement could consider the predevelopment background
phosphorus concentration plus an increase corresponding to the variance in phosphorus
concentrations detected by a monitoring program. The average of the mean annual measured
concentration for 123 lakes in the Muskoka River watershed was 9.3 4.2 g/L. Interannual
variance in the mean total phosphorus concentration can be substantial, however, and ranged
from 10 to 125% in individual lakes. The mean coefficient of variation for annual phosphorus
measurements was 40% or 4 g/L. A routine water quality monitoring program is, therefore, capable of determining the mean annual phosphorus concentrations in lakes to within
40% of the true value in any given year. More intensive, research-level programs, however,
can detect changes of 20% between years (Hutchinson et al., 1991; Clark and Hutchinson,
1992), but this level of precision is not always attainable in routine programs.
This variance of 40% represents the ability of a routine monitoring program to detect
changes and the natural variance in runoff or biotic factors from one year to the next. Any
changes of 40% or less may not, therefore, be detectable over the long term, would be
perceived as routine variance in water quality by lake users, and so may represent an
objective based on the ability to detect change.
Background + 4 g/L. A standard increase of 4 g/L of phosphorus from background may also be considered and modeled as an objective. The figure of 4 g/L represents the natural variance in phosphorus concentration determined by a routine monitoring program but is expressed as an absolute limit instead of a proportional change. This
approach is not recommended, however, because the proportional increase (40 or 50%)
provides a fairer allocation of development to a diversity of lakes. An absolute increase
of 4-g/L allows a doubling of phosphorus from development in nutrient-poor waters
(i.e., 48 g/L) but only a 25% increase for mesotrophic waters (i.e., 1620 g/L).
Existing + 10%. Reviews of the model accuracy showed that much of the septic
phosphorus loading predicted by the model was not expressed as increases in the measured phosphorus concentration in lakes. A final approach would acknowledge this lack
of response but still provide a margin of error for future changes. In this scenario, the existing
measured phosphorus concentration would be used as a baseline and development allocated
to a modeled increase of 10% higher than the measured mean. The small increase would
acknowledge the uncertainty in the assumptions of phosphorus migration from septic fields
so that development would proceed cautiously. Changes in the measured response would
be reviewed against monitoring results every 10 years. If lakes began to respond to shoreline
development, then the objectives could be reviewed against the nature and degree of response
and development limits revised accordingly. Although this approach represents an adaptable
response to uncertainty in model assumptions and maintains a proportional increase, it
does not provide the desired attribute of a stable baseline over time.
Chapter II.17:
671
Background + 50%. The Province of Ontario has considered a revised water quality
objective for total phosphorus in surface waters which is based on a 50% increase in
anthropogenic loadings above the modelled natural background. Although the figure of
50% can be debated, it does reflect the merits of a proportional increase and a modelled
baseline, as discussed previously. Municipalities must consider Provincial Policy, and so
an objective of Background + 50% is used here to illustrate the implications of a potential
Provincial Water Quality Objective.
Filters and water quality objectives

A water quality objective is not the only determinant of development capacity for lakes. Other
physical factors can be considered as filters, additional constraints to development that will
modify any numeric objective developed for water quality. They apply equally to any water
quality objective. Review of water quality objectives against other development filters
helps to determine which aspects of a lake are most limiting to development and to place
the water quality objective in the context of other capacity determinants.
Perimeter filter. The first such filter is shoreline perimeter, or the availability of
waterfront lots based on physical constraints to development. The amount of lakeshore
is limiting for any lake; some Ontario municipalities, for example, require a minimum lot
frontage of 200 ft for new lots. Other municipalities may adopt larger or smaller lot sizes
as a response to narrow embayments or other biophysical limits. Sensitive wetland areas,
steep topography, or lack of soil may impose additional physical constraints that exclude
portions of a lakeshore from development or require larger lot frontages.
In many cases, particularly lakes at the low end of the watershed, perimeter may be
a more restrictive development limit than water quality. Total shoreline perimeter for each
lake in the model can be determined using a geographic information system (GIS). The
perimeter is then divided into 200-ft lots to provide an estimate of maximum shoreline
available for development. This presents an overestimate, however, as steep shorelines or
other physical constraints to development may further reduce the number of developable
lots. These must be assessed on a lake-by-lake basis.
Crown land filter. A second physical filter possibly reducing development potential
is consideration of Crown land. Many lakes in Ontario are surrounded by Crown
land, publicly owned land managed by the province in the name of Her Majesty, the
Queen. These lands are not developed at present and cannot be subdivided by private
interests unless their status is revised by the province. This has been an uncommon
occurrence in southern Ontario and is unlikely to occur in the immediate planning horizon.
Modeling of all lakes in a watershed to the water quality objective of Background + 50%
(including Crown land lakes) thus overestimates their ultimate phosphorus loading. In a
watershed-based water quality model, the phosphorus loading from these lakes is accumulated in the anthropogenic load for downstream lakes, thus restricting future development. Because Crown lands are unlikely to be developed (and any consideration will
involve extensive consultation between the Crown and municipal governments), the
potential future load from shoreline development on these lakes can be removed from
the modeling exercise as an additional limit or filter on development.
Vacant lot filter. Many lakes contain vacant lots on their shorelines. These lots have
been legally created but have not yet been developed. Owners of these lots retain the legal
right to build on them at any time in the future and so their potential phosphorus load
must be subtracted from the future development capacity to account for their ultimate
development. The 123 monitored lakes that are used in this exercise contain 4,400 vacant
672
shoreline lots in addition to over 18,000 developed shoreline lots. These vacant lots represent
a limit to future development of new lots and should be removed from the future capacity
of all lakes in the model as an additional filter.
Lake trout habitat filter. The final filter which can be considered is based on the
presence of lake trout (Salvelinus namaycush) in lakes. The requirement of the species for
cold, hypolimnetic waters and high-oxygen tensions makes it particularly vulnerable to
phosphorus loading and associated oxygen demand (MacLean et al., 1990). Management
of lake trout and their habitat remains a responsibility of the federal and provincial
governments in Ontario and so lake-specific habitat requirements are not managed by
municipalities. Nevertheless, any program must consider the reduced phosphorus contribution imposed by development restrictions on lake trout lakes to acknowledge their
potential contribution to downstream capacity.
In practice, some development of these lakes may be allowed, subject to review by
provincial agencies. The Ministry of Natural Resources (MNR) and the Ministry of the
Environment (MOE) model those lakes that support lake trout populations. In application
the MNR will allow development on lake trout lakes up to a limit based on hypolimnetic
oxygen status. The MOE and the MNR determine this limit through a separate modeling
exercise. When the MNR has identified a problem of depleted oxygen in a lake that
supports lake trout, no further development is assumed for that lake.
Watershed-based development limits

Water quality models are set up to account for phosphorus movement from one lake to
the next lake downstream, and hence throughout the watershed (Dillon and Rigler, 1975;
Dillon et al., 1986). The need for such watershed-based planning has long been
encouraged by provincial management agencies, but specific guidance has been limited.
Watershed-based trophic status models include all hydrologic and phosphorus sources
in a watershed and so meet the intent of watershed-based planning. The models add
the total phosphorus load from one lake to the load for the next lake downstream, after
accounting for in-lake retention. As a result, export of development-derived phosphorus
from upstream lakes makes up part of the development loading for downstream lakes
and must be included as part of the contribution from shoreline development. High
levels of shoreline development on upstream lakes will therefore limit the development
potential of downstream lakes. Sensitive downstream lakes will also limit development
upstream.
The greatest problem with implementing development limits on a watershed basis is
determining the extent of upstream influence of a lake that has reached capacity. Strict
interpretation of watershed-based planning involves extending development limits to all
lakes upstream of any lake that has reached its development capacity. This is logically
consistent, as some of the phosphorus that enters the upper end of a watershed will
ultimately reach the lower end of the watershed.
Although it is logically consistent to account for downstream transport of phosphorus
in setting development capacities it is difficult to develop a spatial limit governing how
far upstream of a capacity lake any limit on development should extend. The implications
are considerable. If a lake in the lower watershed had no additional development capacity,
then strict interpretation of a watershed-based planning policy would see future development restricted on all upstream lakes to prevent downstream transport of phosphorus.
Chapter II.17:
673
In many cases, significant urban development and shoreline development potential may
be located in the upper watershed. The trophic status of downstream lakes could, therefore,
in the strictest sense, prevent any further development upstream.
In the practical sense, the watershed manager must determine the balance between
protecting recreational water quality through watershed-based planning and the strictest
approach, which would limit all development upstream of a capacity lake. Watershedbased planning remains an attractive concept, but quantitative advice on its implementation limits its utility.
Implications of watershed-based phosphorus limits

One approach to watershed-based planning is to first assess its implications:
1. Will the standards of water quality protection be reduced if development is not
limited upstream of a lake that has reached capacity?
2. Does the added protection of a watershed-based approach translate into measurable or predictable improvements in water quality?
The implications can be assessed by running the watershed model in two scenarios:
1. In the first, future development is added to every lake in the watershed up to the
limits prescribed by the Background + 50% or any other objective. These total
future loads to upstream lakes are added to the downstream loading before downstream capacities are set. This approach represents true watershed-based planning.
2. In the second scenario, all lakes are developed to their individual limits, based on
the difference between present day development and development to Background
+ 50% without accounting for the additional load from future upstream development. The phosphorus from this future development is then added to the downstream lakes, which are already at their own, independent capacity limits. This
approach represents the implications of ignoring watershed-based planning.
The results of this exercise show that the implications of allocating development
capacity independently of downstream transport are not significant for most lakes.
Figure II.17.11 shows little overall deviation between watershed-based and independent
allocations of development to the limits of the Background + 50% objective for lakes in
the Muskoka River watershed. Lakes where the two development outcomes do not differ
are shown as a straight 1:1 relationship. Departure from the 1:1 relationship, where the
figure shows points above the 1:1 line, represent the degree to which the final water quality
exceeds or overshoots the background + 50% objective because of upstream loading.
The maximum deviation is 4.4 g/L, in which a lake that should be at 11.1 g/L ends up
at 15.5 g/L. (Figure II.17.11 shows one point with a very high deviation. This lake has
already exceeded its objective as a result of upstream agricultural inputs.)
The ultimate, post-development phosphorus concentrations remained within 20% of the
Background + 50% objective in 93% of the lakes modeled (Figure II.17.12). Water quality
in 12 of the 376 modeled lakes could overshoot the objective by 40% or more as a result of
downstream phosphorus transport. The water quality in these individual lakes would still
be better, however, than that allowed if all were developed to Ontarios present-day Provincial
Water Quality Objectives of 10 or 20 g/L. The implications of exceeding the revised objective
are therefore minor, and less than those of adhering to the present objective.
674

Consideration of Downstream Export in Objective Setting
Objective after Downstream

Export in g/L
40
35
30
25
20
15
10
5
0
0
10
20
30
40
Independent Objective in g/L
Figure II.17.11 Implications of independent versus watershed-based development allocation.
Exceedance of Objective by Downstream Transport

Cumulative
250
225
200
175
150
125
100
75
50
25
0
100
90
80
70
60
50
40
30
20
10
0
0
10
20
30
40
50
Percentage of Lakes
Number of Lakes
Percentage
>50
Percentage Over Objective (BG + 50%)
Figure II.17.12 Number of lakes in which water quality exceeds objective if downstream transport
of phosphorus is not considered.
Although it is logically consistent to account for downstream transport of phosphorus

in setting development capacities, it is difficult to develop a spatial limit governing how far
upstream of a capacity lake any limit should extend. It is therefore clear that watershed-based
planning should be interpreted and implemented in a reasonable fashion. The model already
accounts for downstream transport of anthropogenic phosphorus from all existing development. The scenarios tested here reveal that the implications of not implementing a development freeze upstream of capacity lakes are minor and not necessary to ensure that water
quality is protected. It is therefore evident that a watershed-based orientation of the trophic
status model provides for sustainable levels of water quality. We note, however, that this
conclusion is based on the relatively low loads associated with shoreline cottage development.
Higher loadings, for example from point source, agricultural, or urban inputs, may have a
greater impact on downstream water quality.
Chapter II.17:
675
Setting development objectives

The objective-setting approach must protect water quality as well as assess the implications
of model assumptions to policy for shoreline development. Development and calibration
of the model for the Muskoka River watershed produced accurate estimates of water
quality of the subject lakes, after accounting for reduced phosphorus loading to septic
systems and the role of soils in attenuating phosphorus migration from shoreline septic
systems to lakes. An objective of Background + 50% was tested to correspond to provincial initiatives and filters were developed to compare water quality and other determinants of development capacity. The resultant development capacities can then be summarized for each lake, each subwatershed, and the entire watershed. The capacities can
then be compared with existing development density to provide insight into the degree
of social stability and public expectations of policy.
Development objectives are best expressed as a phosphorus load in policy, so that the
model can be used to compare resort, point source, or other development loadings against
objectives. For ease of interpretation at the implementation stage that load can be converted to Seasonal Residential development using the occupancy figures for individual
lakes (see Validation of cottage usage figures), or to any other type of development.
For this exercise, water quality objectives were stated as the total allowable anthrpogenic phosphorus load in kg. Objectives were set for each of the 376 lakes, bays, or rivers
in the Muskoka River watershed as an index of potential development load, and for the
123 monitored lakes as an index of the loading of those lakes that are currently managed.
Although some jurisdictions may wish to optimize the allocation of future development to each lake, optimization was not attempted in this exercise. Certain popular lakes
may well benefit from the adoption of stricter upstream controls on development, in order
to maintain development opportunities in the popular lakes. Optimization would involve
reducing the development capacity upstream of highly desired lakes, in order to maximize
development of preferred locations. There is a near-infinite number of optimization strategies inherent in a large watershed, and optimization would involve a variety of stakeholders with different interests. Optimization could be considered on a case-by-case basis
where there is a need to reallocate development opportunities.
Reconciliation of model accuracy and assumptions with objectives

Although the Muskoka River watershed model produced a very good correspondence
between measured water quality and modeled estimates on average, there were many
lakes in which a large discrepancy between measured and modeled phosphorus concentrations remained in the final model. The objective of a 50% increase in phosphorus
against background was used here to illustrate a potential starting point for setting
development limits. For this exercise, the 50% increase was modified, based on the
agreement of the modeled estimate of phosphorus concentrations with phosphorus
measurements for individual lakes, to accommodate the observed (versus the theoretical) expression of phosphorus in lakes. Where water quality measurements do not exist,
the model should be assumed to be accurate.
The logic of reconciling objective development with model accuracy can be summarized as follows.
Criterion #1. If the measured phosphorus concentration exceeds the modeled Background + 50% objective and the modeled total phosphorus exceeds Background
+ 50%,
Then no additional development is allocated (accurate model, lake at capacity).
676

Criterion #2. If the measured phosphorus concentration is 80% or more of the modeled total phosphorus concentration and the modeled total phosphorus exceeds
Background + 50%,
Then no additional development is allocated (accurate model, lake at capacity); agreement of 80% or better is considered an acceptable indicator of model accuracy.
Criterion #3. If the measured phosphorus concentration is between 40% and 80%
of the modeled total phosphorus concentration,
Then additional development is allocated up to the Background + 50% objective,
but the objective is modified (increased) to account for the discrepancy between
measurements and modeled estimates (inaccurate model, lake at capacity).
A conservative (protective) approach assumes that 80% of the modeled phosphorus
would ultimately be expressed. The objective is therefore modified to Background
+ (1/0.8 * 50%) = Background + 62.5%.
Criterion #4. If the measured phosphorus concentration is less than 40% of the
modeled total phosphorus concentration,
but the objective is modified (increased) to account for the discrepancy between
measurements and modeled estimates (inaccurate model, lake at capacity).
A conservative (protective) approach assumes that 40% of the modeled phosphorus
would ultimately be expressed. The objective is therefore modified to Background
+ (1/0.4 * 50%) = Background + 125%.
Criterion #5. If no water quality measurements exist,
under the assumption that the model is accurate.
Resultant phosphorus objectives for the Muskoka River watershed example

The total number of shoreline lots on the 123 measured lakes was approximately 14,000
in 1999. This corresponds to an anthropogenic phosphorus load of 9,204 kg. The Background + 50% objective allows an additional 3,403 kg of phosphorus from shoreline
development, after consideration of the perimeter and Crown land filters. This is reduced
to 3,010 kg when the vacant lot filter is added. Application of the filter restricting development on lake trout lakes reduces the allowable load to 2,638 kg. Development of all 123
lakes to the water quality objective of Background + 50% will thus allow a 37% increase
in phosphorus loading from existing levels, or 29% when lake trout and vacant lots are
considered.
Including all of the 376 modelled lakes in the exercise produces a corresponding
increase in phosphorus loading. There are approximately 18,000 lots on these 376 lakes,
corresponding to an anthropogenic phosphorus load of approximately 11,700 kg. The
Background + 50% objective allows an additional 34,143 kg of phosphorus from shoreline development, or 9,451 kg after consideration of the perimeter and Crown land filters.
This is reduced to 8,860 kg when the vacant lot filter is added. Application of the filter
restricting development on lake trout lakes reduces the allowable load to 8,071 kg. Development of all 376 lakes to the water quality objective of Background + 50% will thus
allow a 289% increase in phosphorus loading from existing levels, or 68% when all filters
(perimeter, Crown land, vacant lots, and lake trout) are considered. The implications of
each of the filters, particularly the perimeter and Crown land filters, are shown in Figure
II.17.13. It is clear that water quality alone is not the most sensitive determinant of
development capacity, and that physical restrictions are more limiting.
Chapter II.17:
677
Capacity Determinants in 376 Lakes

Existing Load
Allowable Increase
Percentage Increase
400
350
300
250
200
150
100
50
0
Water Quality
Objective
Crown Land
Filter
Perimeter
Filter
Vacant Lot
Filter
Lake Trout
Filter
Figure II.17.13 Comparison of development constraints (filters) in 525 lakes in the Muskoka River
watershed.
Value of nutrient-based water quality objectives

Phosphorus is the nutrient limiting the growth of algae in the nutrient-poor lakes of the
Precambrian Shield. When the phosphorus load to a lake increases because of anthropogenic sources and water quality declines, the recreational value of a lake will be
diminished. In many municipalities on the southern Precambrian Shield in Ontario,
lake-based recreation and tourism are the foundations of the local economy. A mechanism that allows local decision makers to define and understand the carrying capacity
(whether based on water quality or otherwise) of the lakes within a municipality will
ensure that further development does not unduly stress the natural resources upon
which the area depends.
The intent of water quality-based development policies is to protect water quality
from eutrophication induced by overdevelopment. It is therefore surprising that water
quality is not always the strongest limitation on development capacity in lakes. The
Muskoka River example was tested with a Background + 50% water quality objective
and filters that limited development based on
1.
2.
3.
4.
Physical limits of available shoreline (perimeter)

The presence of undevelopable Crown Lands
Vacant lots that were already committed to development
The presence of lake trout in lakes
Water quality alone did not represent the most significant restriction on shoreline
development potential.
The exercise of modeling and monitoring an entire watershed is complex and costly.
Water quality-based development limits are a worthy exercise in cases where the effort
produces a substantial improvement in water quality protection; for example, if large point
sources or urban areas are present. Biophysical and regulatory concerns may constrain
development capacity far more, however, than water quality. Consideration of Crown
Land and the physical shoreline limits may also reduce development capacity below that
allowable under a very conservative phosphorus water quality objective of Background
+ 50%. Lake trout habitat may be the most conservative filter and result in the lowest
estimate of development capacity.
678
Simple consideration of physical development constraints may therefore, in a whole

watershed analysis, provide sufficient protection of water quality in all but the most
sensitive lake trout lakes. In all cases, however, the manager must review the large-scale
findings at a finer resolution, as there will always be individual lakes to which the physical
constraint does not apply and water quality is the most sensitive determinant. This conclusion applies to shoreline development in the form of recreational or residential development built on individual 200-ft lots. It must be reconsidered, however, for more intensive
forms of phosphorus loading such as urban runoff, sewage treatment plants, agriculture,
and high-density resort development.
Conclusions
Incorporation of the most up-to-date science on phosphorus loadings and dynamics in
Precambrian Shield lakes produced very accurate estimates of phosphorus concentrations
in the lakes of the Muskoka River watershed. The major conclusion was that previous
modeling exercises, which assumed that 100% of the phosphorus in septic systems within
300 m of a lakeshore was mobile, could not be substantiated on an empirical or a mechanistic basis. Adoption of soil-based attenuation produced a substantial improvement in
predictive capability and defensibility of the planning exercise. Although the lakes showed
some response to shoreline development, the degree of response was much less than that
originally predicted.
In the final analysis it is clear that water quality is not the most sensitive determinant
of shoreline development and that adoption of simple standards such as a 200-ft minimum
frontage on shoreline lots may achieve a high level of water quality protection without
the need to rely on a whole watershed model of phosphorus dynamics and lake-specific
water quality objectives. Some form of water quality assessment is essential, but it may
not have to take the form of a complex predictive model a sensitivity analysis may
suffice. Water quality was also well protected in most cases without the need to freeze
all development upstream of a lake which had reached capacity.
Social pressures and user conflicts are becoming increasingly important in cottage
country. These may be partly managed if lake area (recreational space) is used to help
determine cottage density, independently of water quality. It is also clear that properly
managed near-shore development, in which minimum lot sizes are coupled with:
enhanced septic system setback, ensured naturalization of shorelines, protection of significant natural areas, wetlands, and scenic vistas, modern septic systems, mandatory septic
inspection and a strong stewardship program, will likely be as successful in maintaining
water quality as will development limits. Expansion of resorts and cluster or subdivisiontype cottage development provide the potential to use package sewage plants that are
capable of achieving very high effluent quality and that can be monitored to confirm their
inputs, thus reducing the need to rely on assumptions regarding the effectiveness of septic
systems. Finally, the development of lake-specific management plans by lake users or
residents in cooperation with provincial or municipal authorities may bring all users
together to draft a common vision of a lakes future and actions to achieve it based on
site-specific local concerns and consensus.
In summary, lake management planning extends beyond consideration of plumbing.
There will always be some phosphorus enrichment as lakes are developed, whether from
land clearing, septic systems, urbanization, or expanded point sources. Recreational water
quality must therefore remain a component of lake management. Development capacities
based only on phosphorus, however, are costly, complex, and vulnerable to challenge.
Chapter II.17:
679
Managers are encouraged to consider and implement a broader spectrum of management

approaches, in addition to phosphorus-based development capacities.
Literature cited
Barnett, P.J., Cowan, W.R., and Henry, A.P., Quaternary Geology of Ontario, Southern Sheet; Ontario
Geological Survey Map 2556, 1991.
Canada Department of Fisheries and Environment, Hydrological Atlas of Canada, Surveys and Mapping Branch, Dept. of Energy, Mines, and Resources, Ottawa, ON, 1978.
Clark, B.J. and Hutchinson, N.J., Measuring the trophic status of lakes: Sampling protocols, Ont.
Min. Envir. Tech. Report., 1992.
Cox, E.T., Counts and Measurements of Ontario Lakes, Fisheries Branch, Ontario Ministry of Natural
Resources, 1978.
Dillon, P.J and Molot, L.A., Effect of landscape form on export of dissolved organic carbon, iron,
and phosphorus from forested stream catchments, Water Resour. Res., 33, 2591, 1997.
Dillon, P.J. and Rigler, F.H., A simple method for predicting the capacity of a lake for development
based on lake trophic status, J. Fish. Res. Board. Can., 32, 1519, 1975.
Dillon, P.J., Nicholls, K.H., Scheider, W.A., Yan, N.D., and Jeffries, D.S., Lakeshore capacity study, trophic
status, Research and Special Projects Branch, Ontario Ministry of Municipal Affairs, 1986.
Dillon, P.J., Reid, R.A., and Evans, H.E., The relative magnitude of phosphorus sources for small,
oligotrophic lakes in Ontario, Canada, Verh. Internat. Verein. Limnol., 1992.
Dillon, P.J., Scheider, W.A., Reid, R.A., and Jeffries, D.S., Lakeshore capacity study: Part 1 Test of
effects of shoreline development on the trophic status of lakes, Lake and Reserv. Manage., 8,
121, 1994.
District of Muskoka Planning and Economic Development Department, Water Quality Program,
Lot Usage Study, 1995, unpublished.
Downing, J., Lakeshore capacity study committee report, Research and Special Projects Branch,
Ontario Ministry of Municipal Affairs, 1986.
Gartner Lee Ltd., Alternative Technologies for Removal of Phosphorus from Domestic Septic Effluent,
prepared for Science and Technology Branch, Ontario Ministry of the Environ. Dorset Environmental Science Centre, 2002 (in preparation).
Heiskary, S.A. and Walker, W.W., Developing phosphorus criteria for Minnesota lakes, Lake Reserv.
Manage., 4, 1, 1988.
Hutchinson, N.J., Dillon, P.J. and Neary, B.P., Validation and use of Ontarios trophic status model
in establishing development guidelines, Lake and Reservoir Manage., 7, 13, 1991.
Isenbeck-Schroter, M., Doring, U., Moller, A., Schroter, J., and Matthe, G., Experimental approach
and simulation of the retention processes limiting orthophosphate transport in groundwater,
J. Contam. Hydrol., 14, 143, 1993.
Jenkins, D., Ferguson, J.F., and Menar, A.B., Chemical processes for phosphate removal, Water Res.,
5, 369, 1971.
MacLean, N.G., Gunn, J.M., Hicks, F.J., Ihssen, P.E., Malhiot, M., Mosindy, T.E., and Wilson, W.,
Environmental and genetic factors affecting the physiology and ecology of lake trout, in Lake
Trout Synthesis Physiology and Ecology Working Group, Ontario Ministry of Natural Resources, Toronto, 1990.
Michalski, M., Shoreline development, phosphorus loadings and lake trout a scientific and bureaucratic boondoggle, Conf. Proc. Wastewater Nutrient Removal Technol. and Onsite Manage.
Dist., Univ. of Waterloo Centre for Groundwater Research, June 6, 1994.
Michalski, M., Nicholls, K.H., and Johnson, M.G., Phosphorus removal and water quality improvements in Gravenhurst Bay, Ontario, Verh. Internat. Verein. Limnol., 19, 1871, 1975.
Mollard, D.G., Southern Ontario Engineering Geology Terrain Study. Database Map, Muskoka Area.
Parry Sound and Muskoka District, Ontario Ministry of Natural Resources. Ontario Geological Survey Open File Report 5323, 1980.
Molot, L.A., Dillon, P.J., Clark, B.J., and Neary, B.P., Predicting end-of-summer oxygen profiles in
stratified lakes, Can. J. Fish. Aquat. Sci., 49, 2363, 1992.
680
OMEE, Rationale for the Establishment of Ontarios Provincial Water Quality Objectives, Ontario Ministry
of the Environ., 1979.
OMEE, Ontarios Water Quality Objective Development Process, Ontario Ministry of the Environ., 1992.
OMEE, Water Management: Policies, Guidelines, Provincial Water Quality Objectives of the Ministry of
Environ. and Energy, Ontario Ministry of Environ. and Energy, 1994.
Robertson, W.D., Schiff, S.L., and Ptacek, C.J., Review of phosphate mobility and persistence in 10
septic system plumes, Ground Water, 36, 1000, 1998.
Stumm, W. and Morgan, J.J., Aquatic Chemistry. An Introduction Emphasizing Chemical Equilibria
in Natural Waters, Wiley Interscience, New York, 1970.
Winter, J.G. and Duthie, H.C., Export coefficient modelling to asses phosphorus loading in an urban
watershed, J. Am. Wat. Res. Assoc., 36, 10531061, 2000.
Wood, J.S.A., Migration of septic system contaminated groundwater to a lake in a Precambrian
Shield setting: a case study. M.Sc. thesis, Dept. of Earth Sciences, University of Waterloo,
Ontario, 1983.
Chapter II.17:
681
Response
Landlake linkages and land-use limits
This chapter by Neil Hutchinson represents what may very well be the start of an important paradigm shift in water sensitive planning for rural lakes. It is the contention of the
authors that planning based on setting water quality goals may have been overemphasized
in the past. Such approaches have attempted to implement development carrying capacities based on the use of water monitoring data such as total phosphorus models in relation
to the trophic status of lakes. Hutchinson argues that the tens of thousands of calculations that some of these models require may not really be the most accurate barometer of
shoreline development pressures. This occurs due to the inattention devoted to soil processes in nutrient transport and lake eutrophication (see Chapter II.13 by Margot Cantwell
for another discussion of the role of soils in watershed planning).
Instead, this chapter informs us of the need for continual reexamination and reappraisal of existing models of landlake linkages. Due to the important economic link
between tourism and water quality, it is also critical to address such social determinants
as noise, crowding, powerboats, beach use, etc. In other words, diversified planning should
shift its focus from simply septic system plumbing to more comprehensive land-use
pattern analyses. For example, Hutchinson advances the idea that adoption of a minimum
standard for shoreline lot frontage may achieve a high level of water quality protection
without the need to rely exclusively on watershed phosphorus modeling.
In certain situations then, it may be unnecessary to halt all development if a program
of regulating lot size restrictions and locations could be invoked to keep lakes oligotrophic.
This represents a stage in moving beyond water quality monitoring toward setting development objectives through build-out analyses and futures alternatives such as described
in Chapter II.15 by Amir Mueller, Robert France, and Carl Steinitz.
Social and political issues in managing

riparian buffers and corridors
Participants: Kathy Freemark, Stephen Hurst, Frank Mitchell, Arnold
Valdez, Francois de Kock, Leslie Zucker, and Robert France
Past
Prevalence of a fragmented rather than integrated approach, which fails to treat
landscapes as mosaics; e.g., wetland protection separated from riparian buffer management
Disciplinary instead of inter-/transdisciplinary focus of study
Failure to include social components and understanding of processes
Confounding by legal constraints and variability among jurisdictions with respect
to the management of riparian corridors
Traditional engineering solutions
Regulatory approaches subject to political compromises that can result in a minimum amount of riparian protection
Limited abilities to link across spatial scales; i.e., to ratchet up the implications of
site-specific disturbances and interventions to the entire riverine corridor/network
Present
Development of experiential learning to engender an environmental ethic among
landowners
Beginning emphasis on voluntary participation, community-based planning, collaborative approaches, and partnering incentives
Understanding of, and integration among, ecological processes/patterns across
spatio-temporal scales, including technical innovations and socioeconomic implications
Cumulative impact assessment in relation to integrated land-use planning and
management in concordance with the carrying capacity or limits to growth
Emerging recognition of indigenous knowledge and its inclusion in modern landuse planning decisions
Fledgling support for the need for a cultural shift toward humility in dealing with
rivers and their sustaining landscapes
Early moves from a reactive to a proactive mode in protecting water resources within
a framework of open-space planning
Future
Evolution toward regionally appropriate land-use management
Social equity; i.e., elevation of the rights of the many in relation to the rights of
the individual, or in other words, a move toward elements of land-use policy and
governance as practiced in Canada
1-56670-562-2/02/$0.00+$1.50
683
684

Exploration of establishing a network of learning regions as currently being
developed in Europe, facilitated by use of the Internet for bioregional ecological
educational
Acknowledgment of the full range of human characteristics in conscious evolution
(i.e., intellectual, spiritual, emotional, physical) involved in respect for the land
Encompassing such difficult-to-quantify aspects of land respect in decision making
in addition to factual knowledge
Recognition and embracing of landscape architecture, land art, and environmental
ritual in fostering river stewardship
Greater federal government-led aggressiveness in enforcing environmental regulation as imagined to be practiced in Canada
Equitable distribution of resources on a global scale and acceptance of the working
principles of sustainability
Multiple objectives in watershed

management through use of GIS analysis
Participants: Jeffrey Schloss, Paul Zandbergen, Markley Bavinger, and
Robert France
Past
Lack of shared understanding about ecological strengths and development threats
in a watershed context by both the public and regulatory agencies
Absence of long-term vision; specifically the failure to comprehend the implications of alterations in development and development guidelines
Piecemeal assessment of the cumulative effects of imperviousness
Impression of drowning in a cacophony of Internet information with no clear
understanding of what may or may not be important to the interests of any specific
watershed group
Ignorance of what has been previously accomplished in other locations so impression is left that the wheel must be invented anew every time for each watershed
Only limited involvement of locals in land-use planning
Limited use of GIS in only a top-down management approach; i.e., identification
of locations for future development including consideration of zoning, carrying
capacity, and build-out analyses
Present
Recognition and circumvention of the gaps in knowledge through use of GIS
modeling
Education of communities to think in an entire watershed context through enabling
visual comprehension of everyones place in the landscape
Establishment of voluntary controls to development and acknowledgment of the
importance of land conservation
Movement toward use of GIS in a bottom-up management approach; i.e., identification of resources to be protected from any future development including wetlands, groundwater recharge areas, habitat for threatened species, fisheries, etc.
Fusion of the two GIS approaches through iterative feedback to provide the different
interest groups with a logical strategy for spatial analysis; i.e., first, where are the
critical areas that need to be protected?; second, where can new development easily
occur due to available infrastructure?; third, where are the intersection points?
Increased communication facilitation and community building by taking the message directly to the town council
Understanding of the importance of proactive planning; i.e., need to act now to
preserve special features of the watershed
Embracing the guiding principles of low-impact development
1-56670-562-2/02/$0.00+$1.50
685
686
Future
Further transparency and access to information for the public to foster local empowerment in decision making
Small decisions on the local scale have to be informed from a watershed perspective
that must link immediate site-specific knowledge to long-term watershed inventories and future projections
Increased ecological sophistication of the public fostered through computer technologies
Moves toward technical standardization and national referencing of appropriate
GIS approaches
Development of clear visions of the future through alternate futures land-use
planning
Strengthen governmental management and enforcement
Need to jolt people into thinking and planning on larger scales instead of being
lost in the idiosyncrasies of individual sites
Postscript: Implementing water sensitive

planning and design
The chapters selected for this volume represent the best models about how to engage in
water sensitive planning and design as put forward by the most advanced thinkers and
practitioners in the field. For many readers, though, a pressing and fair question may very
well be How can we get there from here?
To this end, a companion volume is currently produced that addresses such concerns.
Titled Practical Support for Watershed Management: Environmental Communication, Demonstration Projects, and Education Outreach, this work is designed to help provide a foundation
from which to move toward effecting positive change in watersheds. Topics covered
include ecosystem health, economic sustainability, innovative diffusion learning, dispute
resolution, ecological integrity, environmental indices, alternate futures modeling, project
scope generation, historical perspectives, combining art and science, buffer gardens, public
art installations, designbuild training workshops, education CD-ROMs, Internet watershed atlases, setting limits to growth, community participation, laboratory training, design
workshops, production of educational publications, exploring evolutionary infrastructure,
GIS decision-maker training, watershed street celebrations, and technology transfer
through extension school outreach.
Understanding the wealth of opportunities that exist for promoting water sensitive
planning and design in watersheds through adopting these nontraditional approaches
will mean that the examples and precedents covered in the present book have a good
likelihood of becoming realized in other locations by other participants.
1-56670-562-2/02/$0.00+$1.50
687
L1562Index Frame Page 689 Friday, April 5, 2002 1:58 PM
Index
A
Abbotsford wetland park, British Columbia,
193194
amenities afforded by, 198201
design for, 196197
educational opportunities afforded by, 200
Fishtrap Creek site for, 194196
habitat provided by, 201202
management of, 201
public use of, 201202
vegetation and planting for, 197198
Aquifer recharge management, 615617
model of, 615
aquifer protection alternatives, 626, 629632
conventional alternatives, 622628
data required, 619, 621, 622
data variability, 619
digital elevation model, 619, 621622
Areas of concern, Ontario, Canada
Ontario Great Lakes Remedial Action Plan,
449450
timber management guidelines, GIS analysis,
602611
Arizona, see Sweetwater Wetlands, Tucson (AZ)
Art Creek, 83
Atlanta (GA), see Peachtree Creek watershed
Australia, see Homebush Bay area (Sydney, Australia)
water garden/pond, 142148

in suburban watersheds, 4964, see also Low-impact
development (LID); Open space design
in urban watersheds, 1127, see also Low-impact
development (LID); Urban watershed(s)
long-term study, University of Virginia, 310
swales for, design testing, 128131
Boston area (MA)
pond and water supply restoration in, 407429, see
also Fresh Pond Reservation
wetland ponds restoration in, 215216
Chandler Pond, 317338, see also Chandler Pond
restoration
funds needed for, 229
Halls Pond, 216223
Hills Pond, 223228
professional consultation in, 229
public educational process in, 231
soils testing in, 228229
vegetation management plan required in, 231
wetland expertise required for, 231
British Columbia, Canada, see Abbotsford wetland
park, British Columbia
Buckman Heights (Portland, OR), watershed
restoration projects, 131134
Buckman Terrace (Portland, OR), watershed
restoration projects, 134141
C
B
Basins, see Detention storage, basins/ponds;
Infiltration
Beltway Plaza Expansion (MD), low-impact design in,
299301
Bioretention, 125126
combined with wetland systems, see Wetland(s);
Wetland ponds
design manual, Prince Georges County (MD),
298299
designs for, 131, 298
eco-roofs, 137142
infiltration gardens, 131134
stormwater planters, 136137
swales, 128131, 135136, 150151
California
Rio Hondo Spreading Grounds, 2123
San Francisco Bay Area, low-impact design
manual, 310
San Francisco Bay area watershed management
plans, 459474
Cambridge (MA), Fresh Pond Reservation restoration,
407429
Canada
193202
Gander Lake, Newfoundland, landwater
interaction modeling, 577596
lakeshore capacity planning, 647679, see also
Lakeshore capacity planning
689
690
timber management/land use, see also Ontario,
Canada, timber management/land use
erosion plumes study, 635643
GIS modeling, 601611
Ontario Great Lakes Remedial Action Plan, 445455
CanadaUnited States Great Lakes Water Quality
Agreement, 446
CeDES (Conservation Development Evaluation
System), 6263
Chandler Pond restoration, Boston (MA)
biolog installation in, 332, 333
community involvement in, 326
condition prior to, 317319
demonstration project for, 324326
design planning for, 322324
management options for, 325
planning for, and dredging, 319324
plantings and bank stabilization in, 326329
protection of, 330332, 334
plantings depth zones in, 331
waterfowl fence for, 334, 335
watershed management planning in, 332338
Channel formation, river, 435438
Chesapeake Bay Riparian Handbook: A Guide for
Establishing and Maintaining Riparian Forest
Buffers (Palone and Todd), 6
Comparative hydrology, see Subdivision design,
residential
Connecticut
Harbor Brook restoration, Meriden, 384, 390393
Merrick Brook restoration, Scotland, 349350
Naugatuck River restoration, 383
river restoration, 379
channel daylighting in, 386
dam removal in, 382383
dechannelization in, 383386
flow management in, 387
geomorphic considerations in, 388390
habitat improvements in, 387
public participation in, 390
river front access planning in, 386387
science of, evolution of, 388390
watersheds in, 380382
Conservation Development Evaluation System
(CeDES), 6263
Conveyances, 13
Cornell University, Ithaca (NY),
bioretention/wetlands system
construction of, for parking lot runoff, 161162
educational opportunities afforded by, 171172
enhancement of, for parking lot runoff at,
159161
habitat improvements at, 172
maintenance plan for, 171
parking lot design for Ornithology Laboratory
complex, 158159
plant selection for, 164165
soils installation in, 170171
soils profiles/creation for, 165170
swale construction in, for parking lot runoff,
162164

Curve number (CN), 35, 109
redevelopment, 111112
Custer Park, Portland (OR), swale design, 151152
D
Design
interdisciplinary, 343344
earth science application in, 346347
ecologic assessment in, 344346
goals and objectives of, 344346
process of, 344
sustainability and, 341343, 347349
low-impact development, 97122, see also Lowimpact development (LID)
open space, 4950, 52, 5658
comparative analysis of, 4960
in low-density subdivision, 5054
in medium-density subdivision, 5460
other comparisons in, 6061
residential development, see also Subdivision
design
in Prince Georges County (MD), 97122
Designing Wetlands: Principles and Practices for Landscape
Architects and Land-Use Planners (France), 4
Detention storage, 109, 113114
basins/ponds,1314, 3132
critical data acquisition for, 34
design goals for, 3334
ecologic effect of, 3638
flood control function of, 35
functions of, 3233
in limited space, 40
mosquito breeding in, 37
muddy discharge from, 4244
outflow control structures of, 3738
slope design for, 4042
water quality function of, 3536
Halls pond restoration, Brookline (MA)
approach to, and process of, 220223
environmental issues before, 216220
wetland plant list for, 233
Digital elevation model (DEM), 513
creating, 520522, 529, 606608
Duck Crossing (MD), redesign comparisons,
5054
DuPont Victoria Wetlands, Victoria (TX), 286290
E
Eco-roofs, 137142
Edgewood Crossroads, Pittsburgh (PA), watershed
restoration, 7882
Educational opportunities
Low-impact development (LID), 120
Sterrett school site, Pittsburgh (PA), 82
Art Creek at, 83
cisterns for roof runoff from, 8284
porous paving at, 85
restorative design for, 8285
Index
swale at, 85
water garden at, 83, 85
201202
Boston (MA), 231
Cornell University, Ithaca (NY), 171172
DuPont Victoria Wetlands, Victoria (TX), 286, 288,
290
Exploration Place, Wichita (KS), 244
Farmington waterway (MN), 191
Greenbush Bay Natural Resource Area, Kent (WA),
281
Sweetwater Wetlands, Tucson (AZ),
284
Wakodahatchee Wetlands, Palm Beach County
(FL), 293294
Escher catchment (Germany), 118
Environmental quality indexing, 453455
Evapotranspiration, 1517
Experimental Sewer System (Japan), 118119
Educational opportunities afforded by, 244
maintenance of, 243
park design concepts for, 237
plantings unwieldiness of, 239240
second planting of, 243
site descriptions, 236237
wetland construction for, 240243
planting process in, 241243
wetland design for, 238239
integrated with reflecting, 237238
plantings in, 239240
691
Wetlands-Based Indirect Potable Reuse Project, 205
advanced wastewater treatment system
development for, 207
baseline monitoring of wetlands in, 207208
goals of, 205
hydrologic features of West Palm Beach, 205206
hydrologic modeling in, 208
Wetland Reuse Site in, 211
wetlands demonstration project for, 207
Frederick County (MD), low-impact subdivision,
305306
Fresh Pond Reservation restoration, Cambridge (MA),
407408
current and anticipated capital projects at, 414
history of, 408
landscape restoration and maintenance plan for,
413, 415
bikeway corridor improvements, 425427
erosion control and slope stabilization, 419420
golf course drainage plans, 425
northeast sector, 424425
perimeter fencing, 422424
road resurfacing, 422424
vegetation, 415419
water purification facility landscape
improvements, 420, 422
natural resource inventory of, 408
vegetative cover types, 412, 413
original plan for, 412
stewardship plan for, 408410, 428429
recommendations of, 410, 413
G
F
aesthetic unidirectional opportunities of,
191
design of, 175,177181
educational opportunities provided by, 191
habitats provided by, 190
maintenance of, 190
municipal approval of, 177
outcomes of,183187
planting recommendations for, 188, 189190
site for, 176177
soils descriptions for, 189
stability of,189190
stormwater treatment provided by, 190
vegetation for wetlands and swales of, 188190
water systems of,189
and water flow, 181183
Filter strip, definition of, 361362
Florida
Florida Aquarium, Tampa Bay, low-impact design,
307
Wakodahatchee Wetlands, Palm Beach County,
290294
Watershed Interactive Network Plan in,
545552
Galilee, Israel, aquifer recharge management,

615632
Gander Lake, Newfoundland, landwater interaction
models, 577579
comparisons of, 589, 592
development of, 580584
GIS maps used, 582, 584
land use changes, evaluating effect of,
592595
land use model, 579580, 584
predicted land use impact, 592, 593
suspended solids contribution, 585586
spatially explicit model, 579580, 584
inherent sensitivity to soil loss, 590592
landscape features, 589
predicted land use impact, 592, 594
slope sensitivity, 589590
suspended solids contribution, 586588
Geographic information systems (GIS)
data commonly available from, 560561
data manipulation by, 559560
data obtained by,559, 560
definition of, 559
in aquifer recharge modeling, 615632
in land cover change modeling, 515516, see also
Land cover change, modeling
692
in landwater interaction, 580592, see also
Landwater interaction
in watershed natural resource inventory mapping,
557558, see also Watershed Natural
Resource Inventory mapping
repositories of data from, 561
schematic diagram of, 559
topology analysis by, 559560
Global positioning system (GPS), 562563
in watershed natural resource inventory mapping,
557
Great Lakes, see Ontario Great Lakes Remedial Action
Plan; Rouge River National Wet Weather
Demonstration Project
Green River Natural Resource Area, Kent (WA),
278281
Greenways
as infrastructure, 395396, 399, 404405
environmental education and, 403
solid waste recycling and, 402
transportation design, 399400
H
Halls pond restoration, Brookline (MA)
environmental issues before, 216220
wetland plant list for, 233
Harbor Brook river restoration, Meriden (CT), 390393
Hearthstone Quarry brook restoration, Chicopee
(MA), 351
High Point (NC), watershed/water supply protection,
low-impact design, 306307
Hills pond restoration, Arlington (MA)
goals of, 223
Homebush Bay area, Sydney, Australia
design concepts for restoration of, 247250
Northern Water Feature, 252260
pollution control, 251252
resource conservation, 251
species conservation, 250251
water cycle strategy, 252
restoration of, 247
site description of, 247250
Hunter Park, Pittsburgh (PA), watershed restoration,
7478
Hydrologic Soil Group (HSG), 514
definitions, Natural Resources Conservation
Service, 530
estimation, using field capacity measurements,
524527
Hydrology, comparative, see Subdivision design,
residential
I
Impervious cover
as indicator of watershed quality, 104107
effects of, on watersheds, 1112, 1718, 49, 616617

in suburban setting, 4964, see also Open space
design
in urban setting,1127, see also Urban watershed(s)
Infiltration, 17
as basis of stormwater management,
27
basins for, 1819,see also Detention storage;
Retention storage
designing, 19, 2023
ecologic use of, 21, 22
examples of, 2023
human use of, 21
during storm event, 2627
effects of, 1718
gardens for,131134
swales for, 2021, see also Swale(s)
vegetation and,125153, see also Bioretention
Infrastructure
definitions of, 396397
greenways as, 395396, 399
in United States, 397399
Integrated Watershed Management in the Global Ecosystem
(Rattan, Ed.), 6
Integrating Stormwater into the Urban Fabric (American
Association of Landscape Architects), 1
Interdisciplinary design, 343344
Israel, see Galilee, Israel
J
Johnson County Wetlands Mitigation Bank project
(KS), 479481, 489
Blue River floodplain, typical cross section, 483
Blue River Watershed Map, 481
concept and plan development of, 481484
construction phases of, 484486
design of, 484
plantings for, 485
review and approval of, 486, 488
city and state exemptions in, 488489
water budget calculations for, monthly, 487
K
Kansas
Johnson County Watershed Map, 478
Johnson County Wetlands Mitigation Bank project,
479489
Kansas Urban Resource Assessment Project,
477479
Wilderness Valley project, 479489
L
Lake(s)
Gander Lake, Newfoundland, landwater
interaction modeling, 577596
Muskoka River watershed, Ontario, lakeshore
capacity planning, 647679
Index
Lake Chocorua, 569571
Squam Lake, 564569
Rouge River National Wet Weather Demonstration
Project, 491511
Lakeshore capacity planning, Ontario, 650
background of, 649650
development capacity, in Muskoka River
Watershed, 676677
municipal responsibilities in, 663664
need for, in Muskoka River Watershed, 647648
shoreline development policy in, 651
objectives, 675676
trophic status modeling in, 647662, see also Lake
trophic status modeling
water management and, 663
water quality objectives in, 663
alternative, 670671
baseline, modeled assessment, 668669
baselines, and measured water quality, 667668
filters in, 671672
phosphorus, total, 664668
proportional increases, 669670
watershed based phosphorus limits, implications,
673674
Lake trophic status modeling, Muskoka River
Watershed, 647, 651
accuracy of, 660662
calibration in, 654
anthropogenic phosphorus sources, 656, 659660
development distance from shoreline, 658659
natural phosphorus sources, 654656
phosphorous retention in soils, 656658
DillonRigler model, 650
Ontario Lakeshore Capacity Model, 650
Ontario Trophic Status model, 652
results of, 660, 676677
technical basis of, 651653
validation of, 653
monitoring programs, 653654
value of, 677679
Land cover change, modeling, 513515
algorithm in, 519520
data sources for, 517519, 561562
digital elevation model creation in, 520522
geographic information systems in, 515516
hydrologic soil group estimation in, 515, 524527
land cover classification in, 522523
precipitation grid formation in, 524
remote sensing technology in, 515516
soil moisture indices in, 516517
estimation process, 527529
Land cover classification, data collection and
modeling, 522523, 529, 532, 534
LANDSAT multi-spectral scanner, and thematic
mapper, 518, 561562
Landscape(s), 15
Landwater interaction
in urban watersheds, 1518, see also Bioretention
model comparisons in, 577579, 584592
model development in, 580584
693
delivery ratios, 583584, 584, 586
GIS map combinations used in spatially explicit
model, 582, 584
landscape features, 589
slope sensitivity, 589590
soil loss values, 583
Universal Soil Loss Equation, 580581
Low-impact development (LID)
aesthetic and educational aspects of, 120
case studies of, 297298
bioretention demonstration project, University of
Virginia, 310
commercial applications, Tampa Bay (FL), 307, 309
in Maryland, 298306, see also Duck Crossing
(MD); Frederick County (MD); Prince
Georges County (MD)
infill development in Beltway Plaza Expansion,
299301
international, 311
Jordan Cove (CT), 311
Maryland State Highway Administration, 311
national, 311
parking lot retrofit, Prince Georges County Office
Complex, 301302
subdivision, Frederick County (MD), 305306
U.S. Environmental Protection Agency, 311
watershed/water supply protection, High Point
(NC), 306307
compared with conventional stormwater
management, 302304
costs of, 121
design approach in,108109
conservation, 109
detention storage, 113114
minimization, 109110
pollution prevention in, 112113
redevelopment curve number in, 111112
runoff volume maintenance in, 111112
time of concentration in, predevelopment,
110111
Maryland Department of Environment, 306
national (EPA), 304305
Prince Georges County, 298299, 302
San Francisco Bay Area Stormwater
Management Agencies Association, 310
ecosystem protection in, 103104
environmental impact of, 100101
examples of, national and international, 117120
habitat restoration in, 102107
implementation of,121122
in current development trends, 101102
in Prince Georges County (MD), 9799, 297298,
see also Prince Georges County (MD)
in restorative development, 100
in urban watershed restoration, 114, 116117
integrated management practices in, 114
political pressures and, 101
purpose of, 99100
watershed protection in, 102103
Lyons residential infiltration strategies (France),
119120
694
M
Maryland
Department of Environment, design manual (2000),
306
Duck Crossing, redesign comparisons, 5054
Frederick County, low-impact subdivision, 305306
Prince Georges County, low-impact development
in, 9799, 297298, see also Low-impact
development (LID)
Beltway Plaza Expansion, infill development,
299301
cost comparisons with conventional stormwater
management, 302304
Count Office Complex, parking lot retrofit, 301302
design manuals published on, 298299, 302
Massachusetts
Chandler Pond, 317338
Halls Pond and Hills Pond, 215231
Cambridge, Fresh Pond Reservation restoration,
407429
Hearthstone Quarry Brook restoration, Chicopee,
351
Salem Salt Marsh restoration, Salem, 351353
stormwater pond design in urban, 3132
critical data acquisition for, 34
ecologic effect of, 3638
flood control function of, 35
functions of, 3233
in limited space, 40
mosquito breeding in, 37
muddy discharge from, 4244
outflow control structures of, 3738
slope design for, 4042
water quality function of, 3536
Merrick Brook restoration, Scotland (CT), 349350
Michigan, see Rouge River National Wet Weather
Demonstration Project
Minnesota, see Farmington waterway (MN)
Modeling
aquifer recharge management, Galilee, Israel,
615632
hydrographic, Northern Water Feature (Sydney,
Australia), 257, 258, 259
Project, 496497
Wetlands-Based Indirect Potable Reuse Project
(FL), 208
hydrologic soils/land cover change, developing
countries, 513534
lake shore capacity, Muskoka River Watershed,
Ontario, Canada, 647679
Gander Lake, Newfoundland, Canada, 577596
San Francisco Bay area, 461466
SUNOM (Simplified Urban Nutrient Output
Model), 50, 5354, 59
Mosquito breeding, in detention basins/ponds, 3738
Muskoka River Watershed, 647648
development capacity, based on modeling, 676677
phosphorus distribution in, 666667
quaternary geology of, 657

subwatersheds of, 660
N
Natural resource inventory(ies), see Watershed natural
resource inventory mapping
Natural Resources Conservation Service (NRCS), U. S.,
514
as GIS data repository, 561
hydrologic soil group classifications, 530
modeling from field measurements, 524527,
527529
Natural Step, 342343
New Hampshire
shoreland buffers, 361362
attitudes toward resource protection, 371373
attitudes toward value of surface waters, wetlands
and shoreland property, 370371
concepts of, 374375
definition of, 361
effectiveness of, 362363
establishing and maintaining, 368369
habitat provided, 363
historical interest in, 362
hydrologic effects of, 364
opinions of appropriate protective methods,
373374
public opinion survey results regarding, 370374
recreational opportunities provided by, 364
water quality maintenance by, 364365
watershed natural resource inventory mapping,
557558
critical watershed analysis, 573575
information systems and data analysis, 557563
Lake Chocorua, 569571
Squam Lake, 564569
wetland/riparian buffer zones, Deerfield, 571573
New York, see Cornell University, Ithaca (NY)
Nine Mile Run watershed, Pittsburgh (PA)
Edgewood Crossroads site in, 7778
infiltration basins in, 78, 80, 81
open space and catchment design for, 81
porous pavement use in, 81
redirection of roof runoff in, 81
restoration of, 7882
habitat restoration in, 88
Hunter Park site of, 74
bioretention areas in, 74
cover modification in, 77, 78
stream restoration in, 74, 77
swales in, 74, 77
industrial use and restoration planning for,
6972
Regent Square Gateway site in, 8586
multiple use design for, 8688
sample designs for restoration of, 7273
status of restoration of, 91
Index
Sterrett school site in, 82
Art Creek at, 83
cisterns for roof runoff from, 8284
porous paving at, 85
swale at, 85
water garden at, 83, 85
Northern Water Feature (Sydney, Australia), 248,
252253, 254
continuous deflection separation systems of, 253
design team for, 256
hydrographic model of, 257
hydrographic models of, 257, 258, 259
plantings for, 256, 257
stormwater treatment in, 253260
wetland ponds of, 255, 258
O
Ontario, Canada, timber management/land use,
601602
areas of concern in, 602603
harvestable timber in, results, 609611
background information, 635637
buffer adequacy findings, 641643
characteristics of erosion plumes, 638639
data collection, 637638
factors influencing erosion, 639641
accuracy of, 608
concept and approach, 603604, 605
data manipulation/GIS formatting, 606
digital elevation model creation, 606608
digital topography data, 604605
forest cover data, 605, 608
slope determination, in riparian zones, 608
study area, 602603, 604
timber grid merged with areas of concern grid, 608,
609
criteria for delisting, 449
status of, 449450
background of, 445446
environmental quality indexing in, 453455
funding concerns in, 452
habitat rehabilitation in, 451452
progress tracking methods in, 446449
progress variability in, 450452
public involvement in, 451452
Open space design, residential, 4950
advantages of, 6162
development costs, 54, 5960, 6162
Duck Crossing (MD), 5054
hydrology, 5253, 59
nutrient output, 5354, 59
Stonehill Estates (Virginia), 5460
evaluation of, 6263
in watershed protection, 6364
low density, 5054
medium density, 5460
other comparative studies of, 6061
695
Oregon, see Portland (OR)
Outflow control structures, of detention basins/ponds,
3738, see also specific projects
P
Parking lot design
bioretentive structures in, 131153, see also Portland
(OR)
bioretentive-wetland system in, 155172, see also
Cornell University, Ithaca (NY)
in parking lot expansion, Beltway Plaza, 301302
in suburban areas, see Low-impact development
(LID); Open space design
in urban areas, see Nine Mile Run watershed,
Pittsburgh (PA)
retrofit of, Prince Georges County Office Complex,
301302
Pavements, porous, 2326, see also specific projects
Peachtree Creek watershed, Atlanta (GA), 1517
Pembroke subdivision, Frederick County (MD),
305306
Pittsburgh (PA), see Nine Mile Run watershed,
Pittsburgh (PA)
Plant selection, in design, see also specific projects
bioretentive structures, Portland (OR), 128153
wetlands, and bioretention swales, 271
Abbotsford Park, British Columbia, 197198
Ithaca (NY), 164165
Northern Water Feature (Sydney, Australia), 257
Boston (MA), 326329
Brookline (MA), 233
Pollutants, in stormwater, 616
removal efficiency for, 3536
Ponds
stormwater, 3145,see also Detention storage,
basins/ponds
wetland, see Wetland(s); Wetland ponds
Porous pavements, 2326
Portland (OR)
bioretentive structures and design in, 125126, 131
Buckman Height watershed restoration projects in,
131134
Buckman Terrace watershed restoration projects in,
134141
Bureau of Environmental Services test swales,
128131
Bureau of Environmental Services Water Pollution
Control Laboratory, 141142
retention pond at, 142147
vegetative filter at, 147148
Custer Park,151152
Oregon Museum of Science and Industry, riparian
zone management, 148151
rainfall analysis of, 126
stormwater analysis, pre- and post-development,
126128
watershed restoration projects in, 131141
Precipitation, 15
Precipitation grid formation, 524
696
Prince Georges County (MD), low-impact
development in, 9799, 297298, see also
Low-impact development (LID)
Beltway Plaza Expansion, infill development,
299301
cost comparisons with conventional stormwater
management, 302304
Count Office Complex, parking lot retrofit, 301302
design manuals published on, 298299, 302
Q
Qualitative habitat Index, 104
R
Rain gardens, see Bioretention
Redesign comparisons, residential subdivision, 4964,
see also Subdivision design, residential
Regent Square Gateway, Pittsburgh (PA), watershed
restoration, 8590
Remote sensing technology
in land cover change modeling, 515516, see also
in watershed natural resource inventory, 561562,
see also Watershed natural resource
inventory
Retention storage, 109,111112, see also Wetland(s);
Wetland ponds
basins/ponds,142147
Rio Hondo Spreading Grounds, 2123
Riparian Management in Forests of the Continental Eastern
United States (Verry, Hornbeck, and Dolloff,
Eds.), 5
Riparian zones, 56, see also Lake(s); Wetland(s);
Wetland ponds
193202
definition, 362
New Hampshire, see under New Hampshire
Newfoundland, Canada, see Gander Lake,
Newfoundland
Ontario, Canada, see Ontario, Canada, timber
management/land use; Ontario, erosion
plumes study
Green River Natural Resource Area, Kent (WA),
278281
Hearthstone Quarry Brook, Chicopee (MA), 351
Johnson County Wetlands Mitigation Bank project
(KS), 479489
Merrick Brook, Scotland (CT), 349350
natural resource inventory mapping, 571573, see
also Watershed natural resource inventory
mapping
Nine Mile Run watershed, Pittsburgh (PA), 4964
Oregon Museum of Science and Industry on
Willamette River, 148151
restoration of, see River restoration, Connecticut
river behavior, as system, 433435
channel evolution in, 435438

connectivity, 433
energy dissipation and free flow, 435438
human activities impact, 380, 381
management practices regarding, 431433, 438442
natural processes of flow, 431
nitrogen removal, 434
periodic ecosystem disturbances, 434435
Rosgen stream classification system, 440, 441
watershed management, 380382
Salem Salt Marsh restoration, Salem (MA), 351353
timber cutting/land use management, see Gander
Lake, Newfoundland; Ontario, Canada,
timber management/land use; Ontario,
erosion plumes study
River restoration, Connecticut, 379
channel daylighting in, 386
dam removal in,382383
dechannelization in, 383386
flow management in, 387
geomorphic considerations in, 388390
habitat improvements in, 387
Harbor Brook, 384, 390393
Naugatuck, 383
river front access planning in, 386387
science of, evolution of, 388390
watersheds in, 380382
Roof design
for redirection of runoff
in Lyons, France, 8284
in Nine Mile Run (Pittsburgh) watershed, 81,
8284
for retention/detention,3839, see also Bioretention;
Eco-roofs
Rosgen stream classification system, 440, 441
Project, Detroit (MI), 491493, 510511
combined sewer overflow management in,
494495, 502503
computer modeling in, 496497
consensus building in, 496497
cost effectiveness of, 496
data collection effort in, 495496
data management/information system in, 506
financial arrangements in, 508509
greenway concepts for, 500
initiation of, 493
institutional arrangements in, 508509
lessons learned in, 498, 500509
luck and political advantages in, 509
monitoring versus modeling in, 506508
public education and involvement in, 497, 498,
504506
regulatory framework formation for, 497498
Rouge Gateway Master Plan, 499
Rouge River Watershed, 493
as Area of Concern in Great Lakes system, 494
characteristics of, 493494
stormwater management, 503504
water flow and water quality issue in, 509
watershed approach to, 500502
Index
Runoff
parking lot, see Parking lot design
roof, see Roof design
volume maintenance,111112
Runoff curve number (RCN), 514
grid formation, 519520, 527529
S
Salem Salt Marsh restoration, Salem (MA), 351353
San Francisco Bay area watershed management plans,
459460
baseline studies for, 463
GIS database, 463464
land use/water quality impact models, 464465
map/tool kits, 465
resource sensitivity zones, 465466
water quality vulnerability zones, 465
land management challenges for, 470
implementation and funding, 472, 474
policy adjustment over time, 470471
short-term versus long-term planning,
471472
planning issues for, key, 466
fire hazard management, 466467
golf course management, 468470
road management, 467
trail management, 467468
planning process in
database establishment, 461
environmental impact reports, 462
goal establishment, 461
management plan development, 462
newsletters, 463
plan evaluation and selection, 461462
public involvement, 461, 462463
public opinion surveys, 463
resource vulnerability maps, 461
web page, 463
San Francisco Public Utilities Commission vision
of, 460461
upstream activities monitoring, 470
Science instruction, see Educational opportunities
Sediment transport, logging roads, 635643
Setback, definition of, 361
Shoreland buffers, see Riparian zones, buffers
Simplified Urban Nutrient Output Model (SUNOM),
see SUNOM (Simplified Urban Nutrient
Output Model)
Soil moisture, modeling, 513515
data sources for, 517519
digital elevation model creation in, 520522
geographic information systems in, 515516
hydrologic soil group estimation in, 515, 524527
indices from, 516517
estimation process, 527529
flowchart, 526
land cover classification in, 522523
modeling algorithm in, 519520
precipitation grid formation in, 524
remote sensing technology in, 515516
697
Squam Lake (NH), watershed natural resource
inventory of, 564
conventional land capability analysis,
564565
geographic data display, 565566
geographic information systems visualization,
565566
geographic referencing and spatial analysis,
566567
in-lake resource inventories, 567
integration of data layers, 567
results, and implementation, 569
useful geographic information system products,
567569
Start at the Source. Design Guidance Manual for
Stormwater Quality Protection (Bay Area
Stormwater Management Agencies
Association), 2
Sterrett school, Pittsburgh (PA), watershed restoration,
8285
Stonehill Estates (VA), redesign comparisons, 5460
Stormwater discharge permits, 32
Stormwater management, 13,155157
bioretention in,125153, see also Bioretention
detention basins in, 1314, 3145, see also Detention
storage
development alternative, and impacts, 617619,
620
for flood control, see Abbotsford wetland park,
British Columbia; Farmington waterway
(MN)
in suburban watersheds, 4964, see also Low-impact
development (LID); Open space design;
Subdivision design
in urban watersheds, 1127, see also Urban
watershed(s)
infiltration basins in, 1823, see also Infiltration
low-impact development in, 97122, see also Lowimpact development (LID)
porous pavements in, 2326, see also Porous
pavements
roof detention in,3839, see also Eco-roofs
wetlands in, 34,14, see also Wetland(s)
Stormwater parks, see Wetland(s)
Stormwater planters, 136137
Stormwater ponds, 3145, see also Detention storage
Stormwater restoration, see Bioretention; Detention
storage; Infiltration; Retention storage;
Stormwater management; Wetland(s)
in suburban watersheds, see Low-impact
development (LID); Open space design;
Subdivision design
in urban watersheds, see Urban watershed(s)
Stream quality, studies of, 105
Stream restoration, 105, 385, see also River restoration
Subdivision design, residential, 4950, see also
Farmington waterway (MN); Frederick
County (MD); Low-impact development
(LID); Prince Georges County (MD)
development costs, 54, 5960
698
Duck Crossing (MD), 5054

hydrology, 5253, 59
nutrient output, 5354, 59
Stonehill Estates (Virginia), 5460
low density, 5054
medium density, 5460
open space, 52, 5658
other comparisons in, 6061
SUNOM (Simplified Urban Nutrient Output Model),
50
in low-density residential modeling, 5354
in medium-density residential modeling, 59
Sustainability, watershed
designing for, 341343, 347349
earth science applications in, 346347
ecologic assessments in, 344346
examples of, 349353
goals and objectives of, 344346
interdisciplinary, 343344
process of, 344
planning for, 541542
in Florida, 545552
process of, 548549
regional, 542545
Watershed Interactive Network Plan for, in Florida,
549553
Swale(s), infiltration/bioretention, 2021
design tests, Portland (OR), 128131
for parking lot runoff, 157158, 162164
in low-impact design, 108114
in urban watershed restoration, see Fresh Pond
Reservation restoration, Cambridge (MA);
Nine Mile Run watershed, Pittsburgh (PA)
Portland (OR), 135136, 150151
Sweetwater Wetlands, Tucson (AZ), 281286
Sydney (Australia), see Homebush Bay area (Sydney,
Australia)
Wakodahatchee Wetlands, Palm Beach County (FL),

290294
Washington, see Green River Natural Resource Area,
Kent (WA)
Water design, 1
Water gardens, 142148, see also Bioretention
Water planning, 1
Water sensitive design, 910, 359360
Water wall, 8284
Watershed(s), 67, 341
assessment of, 477489, see also Watershed(s),
modeling of
greenways and design of, 400402
impervious cover as quality indicator for,
104107
in river restoration, 380382, see also Riparian zones
in suburban areas,4964, see also Farmington
waterway (MN); Johnson County Wetlands
Mitigation Bank project (KS); Open space
design; Prince Georges County (MD);
Wetland(s); Wetland ponds
in urban areas,1127, 6769, 107108, see also Urban
watershed(s)
Texas, see DuPont Victoria Wetlands, Victoria (TX)

Thailand, Sisaket and Chachoengsao Provinces
digital elevation models, 527, 528
land cover change in, 519, 520
classifications, 522523, 529, 532, 534
modeling, 513529, 530531, 533, see also Soil
moisture, modeling
study areas and data sources, 517519, 522
rainfall gauge stations in, 521
relative soil moisture index grids, 533
runoff curve number grids, 531
Thermal battery, 8283
Timber management/land use, see Gander Lake,
Newfoundland, landwater interaction
models; Ontario, Canada, timber
management/land use
Time of concentration (Tc), 109
predevelopment, maintaining, 110111
Top Ten Watershed Lessons Learned (EPA), 7
Trophic status modeling, see Lake trophic status
modeling
Universal Soil Loss Equation, 580581

Urban watershed(s), 1112,107108
landscape dynamics in, 1518
management/restoration of
Boston (MA), 332338
detention basins and,1314, 3145
Galilee, Israel, 615632
High Point (NC), 306307
Homebush Bay area (Sydney, Australia),
247260
infiltration basins and, 1823
low-impact development in, 100, 114, 116117
Nine Mile Run, Pittsburgh (PA), 6994
Peachtree Creek, Atlanta (GA), 1517
policies for, 8891
porous pavements and, 2326
Portland (OR), 125153
Project, Detroit (MI), 491511
San Francisco Bay area, 459474
wetland ponds and, 14
stormwater discharge in, 1213
surface and subsurface processes of, 27
V
Vegetative filters, 147148, see also Bioretention; Plant
selection
Vernal pool hydrology, effects of bioretention on,
172
Virginia
Stonehill Estates, redesign comparisons, 5460
University of, long-term bioretention study, 310
Index
water quality improvement for water supply, see
Homebush Bay area (Sydney, Australia);
Sweetwater Wetlands, Tucson (AZ);
Wakodahatchee Wetlands, Palm Beach
County (FL); Wetlands-Based Indirect
Potable Reuse Project
lake shore capacity planning in, 647679, see also
Lake shore capacity planning
low-impact development in, 100, 114, 116117, see
also Low-impact development (LID)
modeling of aquifer recharge, 615632
hydrologic soils/land cover, 513534
landwater interaction, comparison of models,
577596
natural resource inventory mapping, 557575
natural resource inventory of, see Watershed
natural resource inventory mapping
restoration of, 6769, 9193, see also Riparian zones;
Urban watershed(s); Wetland(s); Wetland
ponds
low-impact development in, 100, 114, 116117,
see also Low-impact development (LID)
sustainability of, see also under Sustainability
designing for, 341349
planning for, 541553
Watershed natural resource inventory mapping, 557
critical land and resources analysis, 573575, see also
data collection and analysis in
global information systems, 557561
global positioning systems, 562563
remote sensing technology, 561562
in New Hampshire, 558
Lake Chocorua (NH), 569571
Squam Lake (NH), 564
conventional land capability analysis, 564565
geographic data display, 565566
geographic information systems visualization,
565566
geographic referencing and spatial analysis,
566567
in-lake resource inventories, 567
integration of data layers, 567
results, and implementation, 569
useful geographic information system products,
567569
wetland/riparian buffers, Deerfield (NH), 571573
West Palm Beach (FL), Wetlands Based Potable Reuse
Project, 205211
Wetland(s), and wetland parks, 34, 263264, 275276,
294, see also Wetland ponds
buffers of, 361375, see also Riparian zones, buffers
in
construction of, 271272, see also specific wetland
project
plans typically required in, 273275
design criteria for, typical, 266267
for science instruction, 235244, see also Exploration
Place, Wichita (KS)
goals and objectives in creation of, 277278, see also
specific wetland project
699
maintenance of, 272, 275, see also specific wetland
project
of Cornell University, Ithaca (NY), 155172, see also
Cornell University, Ithaca (NY)
of DuPont Victoria Wetlands, Victoria (TX), 286290
of Farmington (MN) waterway, 175191, see also
Farmington waterway (MN)
of Fishtrap Creek, Abbotsford, British Columbia,
193202, see also Abbotsford wetland park,
British Columbia
of Green River Natural Resource Area (Kent, WA),
278281
of Johnson County Wetlands Mitigation Bank
project (KS), 479489
of Sweetwater Wetlands, Tucson (AZ), 281286
of Wakodahatchee Wetlands, Palm Beach County
(FL), 290294
planning and feasibility assessment for, 264271
data requirements, 269
plant selection for, 271, see also specific wetland
project
site selection criteria for, 268269
water balance design for, 270
water quality improvement for urban water supply,
see Homebush Bay area (Sydney, Australia);
Sweetwater Wetlands, Tucson (AZ);
Wakodahatchee Wetlands, Palm Beach
County (FL); Wetlands-Based Indirect
Potable Reuse Project
Wetland ponds, 14, see also projects under Wetland(s)
buffers of,361375, see also Riparian zones, buffers
in
of Northern Water Feature, Sydney, Australia,
255
restoration of, in Boston (MA) area, 215216
Chandler Pond, 317338, see also Chandler Pond
restoration
funds needed for, 229
Halls Pond, 216223
Hills Pond, 223228
professional consultation in, 229
public educational process in, 231
soils testing in, 228229
vegetation management plan required in, 231
wetland expertise required for, 231
Wetland treatment system, 205206
Wetlands and Urbanization (Azous and Horner, Eds.), 3
Wetlands-Based Indirect Potable Reuse Project,
205
advanced wastewater treatment system
development for, 207
baseline monitoring of wetlands in, 207208
goals of, 205
hydrologic features of West Palm Beach (FL),
205206
hydrologic modeling in, 208
Wetland Reuse Site in, 211
wetlands demonstration project for, 207
Wilderness Valley project, see Johnson County
Wetlands Mitigation Bank project (KS)

Water Sensitive Planning and Design PDF

Uploaded by

Copyright:

Available Formats

Water Sensitive Planning and Design PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Water Sensitive Planning and Design PDF

Uploaded by

Copyright:

Available Formats

What is the purpose of the book?

What is the purpose of the book?

What are some of the topics covered in the book?

What are some of the topics covered in the book?

HANDBOOK OF

L1562 FMFrame Page 2 Tuesday, April 9, 2002 12:53 PM

Robert L. France, B.Sc., M.Sc., Ph.D.

L1562 FMFrame Page 4 Tuesday, April 9, 2002 12:53 PM

Library of Congress Cataloging-in-Publication Data

Visit the CRC Press Web site at www.crcpress.com

L1562 FMFrame Page 5 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 6 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 7 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 8 Tuesday, April 9, 2002 12:53 PM

Koichiro Matsuura, the Director-General of UNESCO, concluded his message on the

L1562 FMFrame Page 9 Tuesday, April 9, 2002 12:53 PM

Purpose of the book

Structure of the book

L1562 FMFrame Page 10 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 11 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 12 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 13 Tuesday, April 9, 2002 12:53 PM

About the Editor

L1562 FMFrame Page 14 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 15 Tuesday, April 9, 2002 12:53 PM

Margot Young Cantwell

L1562 FMFrame Page 16 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 17 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 18 Tuesday, April 9, 2002 12:53 PM

L1562 FMFrame Page 19 Tuesday, April 9, 2002 12:53 PM

Stormwater management and stormwater restoration...............................................11

Successful stormwater management ponds (Massachusetts) ...................................31

Open spaces and impervious surfaces: Model development principles

Post-industrial watersheds: Retrofits and restorative redevelopment

L1562 FMFrame Page 20 Tuesday, April 9, 2002 12:53 PM

Low-impact development: An alternative stormwater

Water gardens as stormwater infrastructure (Portland, Oregon) ..........................125

Retaining water: Technical support for capturing parking lot runoff

A productive stormwater park (Farmington, Minnesota) .......................................175

A stormwater wetland becomes a nature park

L1562 FMFrame Page 21 Tuesday, April 9, 2002 12:53 PM

I.15 Applications of low-impact development techniques (Maryland) .......................297

Shoreline buffers: Protecting water quality and biological diversity

River restoration planning (Connecticut)...................................................................379

Greenways as green infrastructure in the new millennium ..................................395

Natural resource stewardship planning and design: Fresh Pond Reservation

L1562 FMFrame Page 22 Tuesday, April 9, 2002 12:53 PM

Treating rivers as systems to meet multiple objectives...........................................431

Watershed management plans: Bridging from science to policy

Watershed assessment planning process assessment

Urban watershed management (Detroit, Michigan) ................................................491

II.10 Modeling a soil moisture index using geographic information

L1562 FMFrame Page 23 Tuesday, April 9, 2002 12:53 PM

II.14 Spatial investigation of applying Ontarios timber management guidelines:

L1562 FMFrame Page 24 Tuesday, April 9, 2002 12:53 PM

L1562-1Ch00Frame Page 1 Wednesday, April 3, 2002 12:40 PM

Background Perspectives of water

L1562-1Ch00Frame Page 2 Wednesday, April 3, 2002 12:40 PM

Handbook of water sensitive planning and design