UMEP Modelling

Environmental Modelling & Software 99 (2018) 70e87

Contents lists available at ScienceDirect

Environmental Modelling & Software

journal homepage: www.elsevier.com/locate/envsoft
Urban Multi-scale Environmental Predictor (UMEP): An integrated

tool for city-based climate services
Fredrik Lindberg a, *, C.S.B. Grimmond b, *, Andrew Gabey b, Bei Huang b, c,
Christoph W. Kent b, Ting Sun b, Natalie E. Theeuwes b, Leena Ja€rvi d, Helen C. Ward b, e,
I. Capel-Timms , Yuanyong Chang , Per Jonsson , Niklas Krave a, b, Dongwei Liu f,
b f g
€stberg g, Lingbo Xue b, i,

D. Meyer b, K. Frans G. Olofson a, Jianguo Tan h, Dag Wa
b, j
Zhe Zhang
a
Department of Earth Sciences, University of Gothenburg, Gothenburg, Sweden
b
Department of Meteorology, University of Reading, Reading, United Kingdom
c
Department of Hydraulic Engineering, Tsinghua University, Beijing, China
d
Department of Physics, University of Helsinki, Finland
e
Institute of Atmospheric and Cryospheric Sciences, University of Innsbruck, Innsbruck, Austria
f
Shanghai Institute of Meteorological Science, Shanghai Meteorological Service, China
g
Tyrens AB, Gothenburg, Sweden
h
Shanghai Climate Centre, Shanghai Meteorological Service, China
i
Information Center, Yangzhou University, China
j
School of Environment and Sustainability, University of Saskatchewan, Saskatoon, Canada
a r t i c l e i n f o a b s t r a c t
Article history: UMEP (Urban Multi-scale Environmental Predictor), a city-based climate service tool, combines models
Received 18 April 2017 and tools essential for climate simulations. Applications are presented to illustrate UMEP's potential in
Received in revised form the identification of heat waves and cold waves; the impact of green infrastructure on runoff; the effects
29 August 2017
of buildings on human thermal stress; solar energy production; and the impact of human activities on
Accepted 29 September 2017
heat emissions. UMEP has broad utility for applications related to outdoor thermal comfort, wind, urban
Available online 27 October 2017
energy consumption and climate change mitigation. It includes tools to enable users to input atmo-
spheric and surface data from multiple sources, to characterise the urban environment, to prepare
Keywords:
QGIS
meteorological data for use in cities, to undertake simulations and consider scenarios, and to compare
Urban climate services and visualise different combinations of climate indicators. An open-source tool, UMEP is designed to be
Heat risk easily updated as new data and tools are developed, and to be accessible to researchers, decision-makers
Solar energy and practitioners.
Green infrastructure © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction climate services to those operating and planning different aspects

of city life (transport, energy demand, water supply etc.) is critical
Urban environments are particularly vulnerable to high impact (Horton et al., 2016; Baklanov et al., 2017). A common toolbox,
weather given the high population densities in many cities and the accessible to researchers, decision-makers and practitioners, offers
associated assets and infrastructure (e.g. as evidenced by the im- great potential for better informed climate-related decisions in
pacts of Hurricane Sandy on New York City, Solecki, 2015). With cities.
weather extremes frequently exceeding climate records, and with Scientists and practitioners from a broad range of disciplines
urban areas growing rapidly, the ability to deliver city-based including architecture (e.g. Ren et al., 2011), climatology (e.g.
Eliasson, 2000), planning (e.g. Alcoforado et al., 2009), engineering
and geography have long been interested in how weather and
climate affects cities and their occupants (Baklanov et al., 2017).
* Corresponding authors.
E-mail addresses: [email protected] (F. Lindberg), [email protected] However, the development and adoption of city-based climate
(C.S.B. Grimmond). services, which require production, translation, transfer,
https://doi.org/10.1016/j.envsoft.2017.09.020
1364-8152/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
F. Lindberg et al. / Environmental Modelling & Software 99 (2018) 70e87 71
communication, and use of climate knowledge and information for enable its use without restriction with respect to cost, license or
urban planning, building design and the operation of cities, is not rights issues. Users are encouraged to contribute to the tool to
straight-forward (Chrysoulakis et al., 2013; Grimmond et al., 2014; enhance and extend its capabilities. One of its major features is the
Masson et al., 2014; Baklanov et al., 2017). Appropriate input data ability for users to interact with spatial information to determine
(surface and atmospheric) can be challenging to access and model parameters, and to edit, map and visualise inputs and re-
specialist formats often make them inaccessible to many end-users sults. For this reason, the software is written as a plug-in to QGIS, a
(Grimmond, 2013). Communication between producers and users cross-platform, free and open source desktop geographic infor-
of climate services has been poor, with outputs often not easily mation system (GIS) application (QGIS Development Team, 2017).
interpretable by non-specialists. Tools that are more user-friendly, UMEP has three main elements (Fig. 1): pre-processor (for inputs
and are technically and economically accessible to users, are of meteorological and surface information); processor (modelling
needed to improve communication across disciplines, researchers system e.g. Urban Land Surface Models, ULSM); and post-processor
and users; to better identify user needs; to ensure common as- (tools to analyse the outputs (individual case and ensemble, in-
sumptions across models; to build capacity to address urban dicators of uncertainty, user applications etc.)). Each element is
climate and weather concerns; and transfer research into practice. described briefly in Table 1, with more complete details presented
Past initiatives have tended to focus on specific processes (e.g. in the online manual (http://www.urban-climate.net/umep/
Herbert et al., 1998) or restricted spatial or temporal scales (e.g. UMEP). UMEP allows users to: integrate atmospheric and surface
Bruse and Fleer, 1998), with applications most often intended for data from multiple sources; take meteorological data measured at
specialist researchers. Many of these studies have focused on water ‘standard’ sites and adapt them to be representative of the urban
and waste water management (e.g. Paton et al., 2014; Saagi et al., environment; use reanalysis or climate prediction data; and
2017) and not on integrated hydro-climatological models appro- compare and visualise results or scenarios for different climate
priate for application at multiple scales (neighbourhood to city) indicators of interest (heat indices, intense precipitation, water/
which account for feedbacks and complex interactions (for example energy demand). This all can be done at a range of spatial scales
the effect of water on heat exchanges as well as on flooding). consistent with end-users’ needs and interests (Table 1). To aid
Here we introduce UMEP (Urban Multi-scale Environmental uptake and use of the model, and to develop capacity in urban
Predictor), an integrated tool for urban climatology and climate- modelling more generally, a series of tutorials have been developed
sensitive planning applications. While elements of UMEP have (http://www.urban-climate.net/umep/UMEP_Manual#Tutorials).
been presented elsewhere (see further discussion below and One key contribution of UMEP is to facilitate the preparation of
summary in Table 1), this is the first full description of UMEP and its input data required for City-Based Climate Services (CBCS). UMEP
potential across a broad range of applications. provides both guidance and tools that enable the preparation and
In its current form, the tool can be used for applications related manipulation of data (Table 1). This is particularly important as
to outdoor thermal comfort, urban energy consumption and climate most end-users are familiar with some, but not the full spectrum of,
change mitigation. UMEP consists of a coupled modelling system data needed for applications. For example, planners are knowl-
which combines state-of-the-art 1-D and 2-D models with systems edgeable about building heights, materials and their spatial
to input data from multiple sources, formats and at different tem- arrangement (i.e. urban surface data) and often have GIS skills, but
poral and spatial scales, and to generate output as data, graphs and they may not necessarily have detailed knowledge of meteorolog-
maps. An important feature of UMEP is its ability to couple relevant ical data. Equally, those knowledgeable of the latter may not be
processes and to use common data across a range of applications. expert of the former. Although remotely sensed data may play a
Here the basic structure of UMEP is described, followed by examples very useful part in CBCS, these data may require further processing
of applications to illustrate the potential of this tool. to be applicable in urban areas. UMEP has been designed to
enhance their integration. The tools within UMEP can also be used
2. UMEP overview to provide data to export to other more complex weather, climate,
hydrological, environment modelling systems. Alternatively, data
UMEP is being developed as a community, open-source tool to from more complex models may be imported into UMEP.
Fig. 1. Structure of UMEP (Urban Multi-scale Environmental Predictor).

72 F. Lindberg et al. / Environmental Modelling & Software 99 (2018) 70e87
Table 1
Description of UMEP components and scales of applicability [C: city; L: local (neighbourhood); M: micro (e.g. street canyon, park)], with an example of where their application
may be useful. Those components with post-processing visualization tools are indicated with (*), benchmarking or other statistics are indicated with (þ), and components
described in detail in the text with (#). Note, the micro scale applications are usable across a whole city, but are likely to be computer intensive.
Component Scales Description Example Applications Key References
(a) Pre-processor
Meteorological Prepare existing C/L/M Formats meteorological data for input to the models, Local weather station data can be manipulated into a n/a
data data# dealing with missing data standard format and subjected to basic quality
control
Download Data C/L Downloads climate re-analysis data; modifies it for If local meteorological data are unavailable, allows Weedon et al.
(WATCH)# use in an urban context, representative of the local reanalysis data for land surface modelling to be used (2011, 2014)
scale Kokkonen et al.
(2017)
Ward et al.
(2017a,b)
Spatial Data Spatial Data C/L/M Downloads spatial data from public servers If spatial population data within a city are n/a
Downloader# unavailable, they can be downloaded from world-
wide online dataset
Tree Generator C/L/M Creation/manipulation of 3D vegetation data Influence of vegetation on thermal comfort and n/a
energy exchanges in a city
LCZ Converter C/L Allows morphometric parameters and land cover Generates surface input data for city- to Stewart and Oke
fractions (see Table 4) to be calculated from Local neighborhood-scale modelling. (2012)
Climate Zone (LCZ) maps generated by WUDAPT Ching et al. (2017)
(http://www.wudapt.org/).
Urban Sky View Factor L/M Amount of the hemisphere with restricted view of the Solar access, urban heat island, thermal comfort Lindberg and
geometry sky modelling. Grimmond
(2011a,b)
Wall Height L/M Height and orientation of buildings and walls Solar access, urban heat island, thermal comfort Lindberg et al.
Aspect# modelling. (2016a)
Urban land Land Cover C/L/M Geodata can be translated into the land cover classes Remote sensing and other geodata sources (e.g. n/a
cover Reclassifier used by all the models MODIS classes) can be modified for use with models
and made appropriate for a local area.
Land Cover L/M Surface cover fractions are determined for an area Analysis and characterization of a measurement site n/a
Fraction (point) (circle of selected diameter) or specific directions
(Table 4)
Land Cover C/L As above, but a grid is used to determine fractions for Data input for an extensive area for modelling. n/a
#
Fraction (grid) multiple areas.
Urban Morphometric L/M Morphometric parameters (Table 4) are determined Interpretation of a measurement location in terms of Kent et al.
Morphology Calculator (point) for an area (circle of selected diameter) or specific surface roughness (2017a,b)
directions can be used. Table 4
Morphometric C/L As above, but a grid is used to determine parameters Data input on surface roughness for an extensive Kent et al.
Calculator (grid) for multiple areas. area for modelling (2017a,b)
#
Table 4
Source Area L/M As above, but determined for an area derived from Interpretation of observations, investigation of Kormann and
(point) source area models. potential siting of instruments Meixner (2001)
Kljun et al. (2015)
Kent et al.
(2017a,b)
SUEWS Prepare# C/L Prepares input data for the SUEWS model (processor) For extensive analysis using SUEWS, input forcing n/a
based on information derived from other pre- data and derived spatial information needed can be
processing tools within UMEP. prepared
(b) Processor
Thermal ExtremeFinder# C Finds extreme high and extreme low events (e.g. heat Identification of extreme temperatures from a long See Table 2
Comfort waves or cold waves) in meteorological time series climate record.
data (Table 2).
Mean Radiant L/M SOLWEIG estimates spatial (2-D) variations of 3-D Tmrt is an important meteorological variable Lindberg et al.
Temperature radiation fluxes and the mean radiant temperature governing the human energy balance and thermal (2008)
#
(SOLWEIG)* (Tmrt) in complex urban settings. Both 3D vegetation comfort outdoors, especially during clear and calm Lindberg and
(trees and bushes) as well as ground cover variations summer days (Mayer and Ho €ppe, 1987) Grimmond
are currently considered in the model. (2011a,b),
Lindberg et al.
(2016b)
Urban Energy Anthropogenic C/L Globally applicable method (low spatial resolution) to Quantifies contributions to energy budget from Allen et al. (2011)
#
Balance Heat (QF) (LQF) calculate QF building energy use and road traffic. Lindberg et al.
(2013a)
Gabey et al. (2017)
Anthropogenic C/L Locally applicable method (high spatial resolution) to Quantifies contributions to energy budget from road Iamarino et al.
Heat (QF) (GQF)# calculate QF traffic, residential and non-residential building (2012)
energy use, with different fuels and vehicle types Gabey et al. (2017)
considered.
Urban Energy C/L Urban land surface model that allows radiation, Simulations of evaporation and heat fluxes for a €rvi et al. (2011,
Ja
and Water energy and water fluxes to be calculated for a single neighborhood or across a city 2014)
Balance (SUEWS; point or area. Ward et al. (2016,
Simple)* # 2017b)
Urban Energy C/L As above, but for multiple areas (e.g. a city with 1000s Assessment of future city plans and impacts of Table 3
and Water of grid squares, or local planning zones (of any shape)). drought, heatwaves, water management, green
Balance (SUEWS; infrastructure.
#
Advanced)*
Table 1 (continued )
Component Scales Description Example Applications Key References
Solar Radiation Solar Energy on L/M Shortwave irradiance on ground, roofs and building Potential for energy production, resource planning. Lindberg et al.
Building walls is estimated based on high resolution Digital (2016b)
Envelopes (SEBE) Surface Models (DSMs) and input meteorological
*# forcing data.
Daily Shadow L/M Shadow maps are derived from buildings and 3D Impact of building form and vegetation on energy Lindberg and
Patterns vegetation. production, outdoor thermal comfort. Planning of Grimmond
parks and outdoor spaces (2011a,b), Ratti
and Richens
(1999)
(c) Post-processor
SEBE L/M Visualization of solar irradiation on building Identify areas of good or poor potential for n/a
(Visualization) envelopes (roofs and walls) generation of solar energy if panels were installed
SOLWEIG L/M SOLWEIG output: point (temporal), model domain Identify: potential hazard areas; potential sites for n/a
Analyzer#þ (spatial/temporal) , park).
outdoor activities (e.g. cafe
SUEWS C/L SUEWS output: point (temporal), model domain Comparison of indicators under different planning Ward and
Analyzer#þ (spatial/temporal) scenarios through time and as maps Grimmond (2017)
Benchmarkingþ C/L/M Statistical tool to compare different datasets Model evaluation n/a
Table 2
The heat-cold- wave indices used in UMEP- ExtremeFinder.
Extreme Reference Index description

Event
Heat- Meehl and Tebaldi (2004) Longest period when maximum temperature is above the 97.5th percentile for at least 3 days; average daily maximum
wave temperature across the event is over the 97.5th percentile; and all days are above the 81st percentile.
Fischer and Sch€ ar (2010) Periods of at least 6 days where maximum temperature exceeds the calendar day 90th percentile.
Vautard et al. (2013) Periods of various length when daily mean temperature is above the 90th percentile.
Schoetter et al. (2014) At least 3 days above the 98th percentile of maximum temperature.
Cold Keevallik and Vint (2015) Cold night: temperature lower than 10th percentile of daily minimum temperatures calculated for a 5-day window centered
-wave on each calendar day in dataset;
Cold wave: six consecutive cold nights.
Srivastava et al. (2009) Minimum temperature is below the normal temperature by 3 C or more, consecutively for 3 days or more.
Busuioc et al. (2010) Cited by At least 6 consecutive days with negative deviations of at least 5 C from the normal value of each calendar day.
Micu (2012)
UMEP has a broad range of capabilities (Table 1). Each of its extreme meteorological conditions are of interest and concern. To
elements may be used independently or in varying combinations. identify these extremes, analysis of a long climatological record is
Users may be interested in the output from tools that are provided required (Table 2). However, if such data are not available for the
in the pre-processor for other modelling applications (e.g. in area of interest, UMEP allows the user to draw on the reanalysis
generating urban surface information or standardised meteoro- dataset WATCH Forcing Data ERA-Interim (WFDEI) (Dee et al. 2011;
logical fields) or in applications that require a chain of tools to Weedon et al., 2011, 2014). This product was selected as it was
provide climate indicators for decision making. As many of the designed to be used for hydrological and land-surface modelling for
individual tools, as well as their evaluations, have been described in climate purposes and has been used in several cities around the
disciplinary focussed papers (see Table 1 references), here we world to explore variations in energy flux partitioning (Best and
present a range of examples, each of which requires the use of Grimmond, 2016).
several tools to obtain a solution. To determine the extreme thermal conditions for a site, the first
step is to use the UMEP Download data (WATCH) (Table 1) to
obtain a meteorological time series for the period and location of
3. UMEP applications
interest. These data can also be used for other UMEP applications
(e.g. section 3.2). For example, they can be downscaled to the area
In this section examples of applications are presented to illus-
of interest using the techniques of Best and Grimmond (2016) and
trate UMEP's potential, specifically in the identification of heat
Ward et al. (2017a,b) (Appendix 1).
waves and cold waves in cities (ExtremeFinder) (bold is used
Currently, ExtremeFinder provides four methods to identify
hereafter to indicate the component of the UMEP plug-in tool) ; the
heat-waves and three for cold-waves (Table 2), as there is no
implications of green infrastructure on runoff (SUEWS); micro-
generally accepted definition of either phenomena (Robinson, 2001;
scale heat stress (SOLWEIG); solar energy production (SEBE); and
Vaidyanathan et al., 2016) different percentiles are used to define
sources of anthropogenic (human-generated) heat (LQF, GQF). Each
extremes (e.g. Table 2). The thresholds for the extremes are based on
application draws on different combinations of UMEP tools.
fixed values or quantiles calculated from the meteorological time
series. Use of a time series spanning decades is therefore recom-
3.1. Application example 1 - identification of extreme thermal mended. The user can modify the fixed thresholds and quantiles
conditions (http://urban-climate.net/umep/UMEP_Manual#Outdoor_
Thermal_Comfort:_ExtremeFinder). Daily values are then evaluated
For many urban planning and human health applications,
to determine if an extreme event has occurred. ExtremeFinder b) SUEWS (Advanced) provides a full version of the model
identifies the dates and daily maximum (average or minimum) appropriate for investigating both spatial and temporal varia-
temperatures of all extreme high (low) events during the period of tions of the urban energy balance.
interest (Fig. 2, yellow boxes) based on the criteria set out by the
method chosen. The SUEWS model has been extensively evaluated for a variety
of locations and situations worldwide (Table 3).
The workflow for an application utilizing SUEWS within UMEP
3.2. Application example 2 - urban energy and water balance fluxes is outlined in Fig. 3. Geodatasets that contain information about the
urban environment are used with the pre-processor tools (Fig. 3,
Energy and water balance fluxes are critical to surface- gray and yellow) to provide the required surface parameters. The
atmosphere interactions in an urban area. The impact of extreme model can be applied to areas of any shape, as in most cities
conditions (heat waves, droughts, floods etc.) are influenced by the planning units have known boundaries already available in vector
state of the urban environment prior to these events, with the ur- polygon format (e.g. boroughs, wards). Alternatively, a square grid
ban energy and water balance varying with different neighbour- can easily be created in QGIS (Fig. 4a).
hood (local-scale) characteristics. The Surface Urban Energy and As land cover (Fig. 3) is a key variable for many calculations, a
Water balance Scheme (SUEWS) is an urban land-surface model method to reclassify data is provided. The UMEP Land Cover
included in the processing part of UMEP (Table 1). The model Reclassifier enables land cover raster grids to be created from
simulates the urban radiation, energy and water balances using sources such as MODIS and then converted to the standard UMEP
commonly measured meteorological variables and information cover types (Table 4). Surface cover fractions are very important
about the surface cover. SUEWS is applicable at the neighborhood given differences in energy and water partitioning that result from
to city scale. In UMEP SUEWS is uncoupled, i.e. advection between underlying differences in moisture availability and surface
grids is not accounted for. UMEP allows SUEWS to be run as a properties.
standalone model, or UMEP can provide the appropriate parame- Accessing reliable sources of land cover information to derive
ters for the use of SUEWS within a 3-D model meso-scale model these parameters at the scale of interest remains a challenge.
such as WRF (Dudhia, 2014). The parameters calculated with UMEP Crowd-sourced data sets such as OpenStreetMap (http://www.
tools can provide the input parameters to a wide range of urban openstreetmap.org) and WUDAPT (http://www.wudapt.org/, see
land surface models (either standalone versions SURFEX (Masson below) offer potential but may be incomplete or inconsistent. Other
et al., 2013), JULES (Best et al., 2011), CLM (Lawrence et al., 2011), sources such as MODIS (https://terra.nasa.gov/about/terra-
SLUCM (Kusaka et al., 2001) or coupled to larger scale models). instruments/modis) are likely to be complete but give low spatial
SUEWS uses an evaporation-interception approach (Grimmond resolution at the sub-km scale. In addition, the number of human-
and Oke, 1991) for an area comprised of seven land cover types altered (urban) classes from such data are limited (3 classes) and
(water, buildings, grass, paved, bare soil, deciduous trees/shrubs the discrimination between land cover classes needs to be made
and evergreen trees/shrubs). The state of each surface type at each based on land use. In the example given here for central London,
time step is calculated from the running water balance of the the land cover information (Fig. 4a) is derived from OS MasterMap®
canopy where the evaporation is calculated from the Penman- Topography Layer (Ordnance Survey, 2010). The Land Cover Frac-
Monteith equation. The soil moisture below each surface type tion (Grid) tool is used to calculate grid-based land cover fractions
(excluding water) is considered. UMEP has the latest version of (Tables 1 and 4) based on the land cover raster grid.
SUEWS (Ward et al., 2017a,b) accessible through two links: Morphometric parameters (Table 4) related to surface rough-
ness can be obtained from the Morphometric Calculator (Grid)
a) SUEWS Simple: provides a useful starting place to introduce using digital surface models (DSM) (Fig. 4b). From these data, the
UMEP and SUEWS. Example data are provided so that users can zero-plane displacement (zd), aerodynamic roughness length (z0)
explore the impact of modifying urban surface characteristics. and other geometric parameters such as mean roughness-element
With SUEWS Simple, the ULSM can be executed for a single height and frontal area index are calculated (Table 4). The rationale
location (area). behind the different methods, and a basis for selecting between
Fig. 2. After heat-wave or cold-wave conditions are identified with ExtremeFinder, a series of graphs are generated including the daily temperatures with the extreme periods
indicated in yellow. In the example here, heatwaves are identified for London (yellow boxes) in the period 1990e2010 using the Meehl and Tebaldi (2004) method (Table 2). (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 3
Studies that have evaluated and applied SUEWS (presented chronologically). Note other sub-component models have been evaluated and applied prior to these references.
City Reference Variables evaluated/application
Evaluation
Vancouver, Canada J€
arvi et al. (2011) Water use, anthropogenic heat flux, radiation, sensible and latent heat fluxes, soil moisture, surface wetness
Los Angeles, USA J€
arvi et al. (2011) Radiation, sensible and latent heat fluxes
Helsinki, Finland J€
arvi et al. (2014) Radiation, sensible and latent heat fluxes, snowmelt, runoff, albedo, snow depth
Montreal, Canada J€
arvi et al. (2014) Radiation, sensible and latent heat fluxes, snowmelt, albedo, snow density, snow depth
Dublin, Ireland Alexander et al. (2015, 2016a) Radiation, sensible and latent heat fluxes
Sacramento, USA Onomura et al. (2015) Boundary layer height, sensible and latent heat flux
Hamburg, Germany Alexander et al. (2016a) Sensible and latent heat fluxes
Melbourne, Australia Alexander et al. (2016a) Sensible and latent heat fluxes
Phoenix, USA Alexander et al. (2016a) Sensible and latent heat fluxes
Swindon, UK Ward et al. (2016) Radiation, sensible and latent heat fluxes, soil moisture, surface wetness
London, UK Ward et al. (2016) Radiation, sensible and latent heat fluxes, soil moisture
Helsinki, Finland Karsisto et al. (2015) Radiation, sensible and latent heat fluxes, snow depth
Shanghai, China Ao et al. (2016) Radiation
Singapore Demuzere et al. (2017) Radiation, sensible and latent heat fluxes
Basel, Switzerland J€
arvi et al. (2017) Radiation, sensible and latent heat fluxes, snow depth
Minneapolis-Saint Paul, J€
arvi et al. (2017) Radiation, sensible and latent heat fluxes, snow depth
USA
Application
Helsinki, Finland Nordbo et al. (2015) Effect of surface cover resolution on SUEWS performance
Dublin Alexander et al. (2016a,b) Impact of urban development pathways on sensible and latent heat fluxes across the city of Dublin
Dublin, Ireland Alexander et al. (2016b) LCZ classification combined with SUEWS
Hamburg, Germany Alexander et al. (2016b) LCZ classification combined with SUEWS
Melbourne, Australia Alexander et al. (2016b) LCZ classification combined with SUEWS
Phoenix, USA Alexander et al. (2016b) LCZ classification combined with SUEWS
Porto, Portugal Rafael et al. (2016) Urban resilience measures under different climate scenarios
London, UK Ward and Grimmond (2017) Impact of urban development and climate mitigation measures on energy partitioning across Greater London
Table 4
Spatially related urban surface parameters determined by UMEP pre-processors.
Plug-in Parameters
Land Cover Types: water, building, paved surfaces, bare soil, deciduous trees, evergreen trees, and grass surfaces
Plan area fraction; the combined fractions for a grid must sum to 1.
Source area Two source area models are included: Kormann and Meixner (2001), Kljun et al. (2015)
Morphometric methods Hav e average roughness-element height Methods for z0 and zd:
lp e plan area index Rule of thumb (Grimmond and Oke, 1999)
lf e frontal area index Raupach (1994)
Hmax e maximum roughness-element height Bottema and Mestayer (1998)
sH e standard deviation of roughness-element heights Macdonald et al. (1998),
z0 e Aerodynamic roughness length Millward-Hopkins et al. (2011)
zd e zero-plane displacement Kanda et al. (2013)
Wall Height and Aspect Wall height
Aspect (orientation )
SVF Calculator Sky View Factor (Values of 1 indicate complete sky access, 0 no sky access. Height of calculations can be varied)
them, is outlined in Kent et al. (2017a,b). Parameters can be derived Downloader). In the example given here, averaged population
for the full surrounding area or for sectors reflecting different wind density between residential and working population is used
directions. (Fig. 4c). Such differentiations are very important in locations such
Applications which involve an assessment of the area sur- as central London for anthropogenic heat flux calculations (Dong
rounding a measurement point can use the Source Area Model et al., 2017; Gabey et al., 2017).
(Point). Currently, there are two turbulent flux source area models Given the challenges of acquiring all the datasets needed (DEM,
included within UMEP (Tables 1 and 3): the analytical model of DSM, and land cover in Fig. 3), local climate zone maps (LCZ;
Kormann and Meixner (2001) and Kljun et al.'s (2015) parameter- Stewart and Oke, 2012) are included in UMEP. From these, a first
isation of a Lagrangian stochastic particle dispersion model. These estimate of input parameters for SUEWS can be made. In UMEP
models indicate the probable surface area contributing to a turbulent the LCZ maps from the WUDAPT database (www.wudapt.org;
flux measurement at a specific point in time and space with imposed Ching et al., 2017) can be translated using the LCZ Converter.
boundary conditions (e.g. meteorological conditions, sources/sinks If more detailed information is available for specific areas, or
of passive scalars or surface characteristics). The results from these becomes available subsequently (e.g. local high resolution DSMs),
models facilitate interpretation of observations, enable improved parameters can be updated.
evaluation of flux models, and/or allow assessments of the appro- The other major input to SUEWS is the meteorological forcing
priateness of siting of new instrumentation (Fig. 5). data (Table A1.1). Such data need to be for above the height of the
Population density (people per hectare) is used in the estima- roughness elements (trees, buildings). A common format is used in
tion of anthropogenic heat flux in SUEWS. If population density all UMEP models (Table A1.1). Most applications require a contin-
datasets are unavailable (e.g. as would be obtained from local uous gap-filled data set. For many urban applications, the start and
census data), the Spatial Data Downloader can be used. This plug- finish of daylight savings is linked to important behavioural pat-
in is directly connected to various Web Coverage Services (WCS) terns (e.g. the shift of rush hour). Therefore, the individual models
including global datasets on population density (http://urban- account for daylight savings if relevant (e.g. timing of anthropo-
climate.net/umep/UMEP_Manual#Spatial_Data:_Spatial_Data_ genic energy use, irrigation). The Preparing Existing Data UMEP
tool (see Appendix 1) supports preparation of the meteorological reducing surface runoff compared to an increase in grass surfaces (a
data and conversion into the format used in all UMEP models. 10% increase in street trees is more effective than a 30% increase in
Once all the required information is pre-processed, SUEWS grass surfaces).
Prepare can arrange the data so the model can be executed. SUEWS
Analyzer (Fig. 3) allows spatial (Fig. 6a), temporal (Fig. 6b) and 3.3. Application example 3 - mean radiant temperature
between variable (Fig. 6c) model results to be explored. In this
example application, central London average daytime sensible heat Temperature-related health problems are expected to increase
fluxes (QH) for four-months period in 2015 are mapped (Fig. 6a), with rising temperature in cities, especially during more extreme
with detail of temporal variations of net all-wave radiation (Q*) and temperatures associated with heat waves. Mean radiant temper-
QH for nine days for one area (grid ID 44) graphed (Fig. 6b). The ature (Tmrt) is one of the most important meteorological variables
relation between Q* and QH for grid ID 44 for the whole time period governing the human energy balance and thermal comfort out-
(Fig. 6c) is also plotted. doors, especially on clear and calm summer days (Mayer and
To illustrate a hydrological application, to examine runoff gen- Ho€ppe, 1987). To provide estimates of thermal comfort/heat
eration in different planning scenarios, SUEWS was run for a highly stress for people, SOLWEIG (SOlar and LongWave Environmental
built-up catchment area (24 ha) in Helsinki for 2010 (Fig. 6def). The Irradiance Geometry model) can be used to calculate Tmrt. In
planning scenarios considered the current land cover/use (base SOLWEIG, both 3D vegetation (trees and bushes), as well as var-
run), and a 10% and 30% increase in areal coverage of street trees iations in ground cover, can be considered (Lindberg and
and grass surfaces at the expense of paved surfaces (Fig. 6f). Results Grimmond, 2011b; Lindberg et al., 2016b). SOLWEIG has been
from these simulations indicate that the increase in the amount of evaluated extensively and applied at urban locations worldwide
street trees (i.e. areal fraction of street trees) is more effective in (Table 5).
Table 5
Evaluation and application studies using the SOLWEIG model.
City Reference Variables Settings
Evaluation
Gothenburg, Sweden Lindberg et al. (2008) Radiant fluxes and Tmrt City square, courtyard
Gothenburg, Sweden Lindberg and Grimmond (2011a) Radiant fluxes and Tmrt City square with tree
Kassel, Germany Lindberg and Grimmond (2011a) Tmrt Street canyon
Freiburg, Germany Lindberg and Grimmond (2011a) Tmrt Four urban sites
London, UK Lindberg et al. (2016b) Radiant fluxes and Tmrt Different urban ground covers
Shanghai, China Chen et al. (2016) Tmrt Very dense urban environment
Hong Kong, China Lau et al. (2016) Tmrt Very dense urban environment
Application
Gothenburg, Sweden Thorsson et al. (2010) Spatial and temporal Tmrt Future climate scenarios
London, UK Lindberg and Grimmond (2011b) Tmrt Intra-urban differences
Gothenburg, Sweden Lindberg et al. (2013a,b) Spatial Tmrt Consideration of weather
Stockholm, Sweden Thorsson et al. (2014) Tmrt Prediction of heat related mortality
Porto, Portugal Lau et al. (2014) Tmrt Effects of urban geometry, climate change
Gothenburg, Sweden Lau et al. (2014) Tmrt Effects of urban geometry, climate change
Frankfurt, Germany Lau et al. (2014) Tmrt Effects of urban geometry, climate change
Berlin, Germany €nicke et al. (2016)
Ja Tmrt City-wide characteristics
Adelaide, Australia Thom et al. (2016) Tmrt Influence of increasing tree cover
Gothenburg, Sweden Lindberg et al. (2016b) Spatial and temporal Tmrt Future climate scenarios
Fig. 3. Workflow and geodata for analysing urban energy balance using the SUEWS model. Bold outlined boxes are mandatory items. In some cases alternatives are shown. Yellow,
orange and red indicates pre-processor, processor and post-processor tools, respectively (consistent with Fig. 1). Grey boxes indicate geodatasets and white boxes other types of
data. DEM e digital elevation model, DSM e digital surface model, LCZ e local climate zones. It is strongly recommended that all geodata used are transformed into the same
projected coordinate system. Model areas need to be defined in a vector polygon layer. For the meteorological forcing, users could manipulate their own data (Metdata Processor),
use the WATCH e WFDEI climatological data set (Download data (WATCH)) or link to their own already-prepared data. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
Fig. 7 shows the UMEP workflow to examine 3D radiant fluxes

and Tmrt. Of the four geodatasets needed, the ground and building
DSM (Fig. 8a) is fundamental. If available, a vegetation DSM (CDSM)
can be added (Lindberg and Grimmond, 2011b). However, as 3D
information on vegetation is sparse, the Tree Generator tool allows
point vector data of tree locations to be transformed into a CDSM.
Ground cover information can be used to estimate outgoing short
and longwave radiation fluxes (Lindberg et al., 2016a,b). To obtain
the appropriate ground cover classes, the Land Cover Reclassifier
(Fig. 7) can be used to obtain the five ground cover classes (water,
buildings, grass, paved and bare soil) used in SOLWEIG (Fig. 8b; no
bare soil present).
The sky view factor (SVF) is the ratio between the radiation
received (or emitted) by a planar surface and the radiation emitted
(or received) by the entire hemispheric environment (Watson and
Johnson, 1987). This dimensionless metric (totally obstructed ¼ 0,
totally unobstructed ¼ 1) is important to human comfort (Fig. 8c),
solar energy and solar access. A pixel-wise sky view factor calcu-
lated in SVF uses ground and building DSMs and/or vegetation DSM
(Fig. 8c).
Solar access and radiative exchanges are impacted by wall
height and aspect. Wall height and aspect (Tables 1 and 4, Fig. 7)
provides wall pixels, with their height (Fig. 8d) and aspect (de-
grees). The latter is a modification of the Goodwin et al. (2009)
linear filter (Lindberg et al., 2015b). To model Tmrt successfully,
building footprint locations must be derived from either the ground
cover grid or from differences between ground heights (DEM) and a
DSM (Fig. 7).
SOLWEIG can be used for an individual time or a time series.
For the latter, points of interest (POI) are added within the
model domain. SOLWEIG Analyzer can be used to provide
spatial (Fig. 8e) and temporal (Fig. 8f) visualizations of results. By
comparing the input geodata (Fig. 8aed) and the results
(Fig. 8eef), the micro-scale influences on the temporal and spatial
patterns can be identified and explained. As shown in Fig. 8f, the
temporal influence of Tmrt is unlike air temperature; it is highly
affected by other variables, such as shortwave and longwave ra-
diation fluxes.
Application of SOLWEIG to explore variations in Tmrt around the
Civic Square (Medborgarplatsen) of central Stockholm shows that
at 2 p.m., shadows from buildings and vegetation are important
(Fig. 9). The highest Tmrt values are next to sunlit walls and on the
sunlit open spaces. Open areas have high values Tmrt due to partly-
cloudy conditions, which increases the proportion of diffuse
shortwave radiation. The high values of Tmrt adjacent to the walls
are related to the emitted longwave radiation and reflected short-
wave radiation from the sunlit walls.
To explore the potential impact of Tmrt, it is useful to consider
critical health thresholds. From analysis of Stockholm County daily
all-cause mortality data (1990e2002), Thorsson et al. (2014) found
that when Tmrt exceeds 59.4 C there is an increase in heat-related
mortality of 10% for those > 80 years of age. Using this threshold the
areas of greatest hazard can be identified (adjacent to the sunlit
buildings, Fig. 9b).
To determine the effects of warmer air temperature
Fig. 4. Examples of input spatial data required to apply SUEWS for central London: (a) (þ2 C, þ4 C) the hazard can be re-analysed with further SOLWEIG
land cover overlain with a polygon grid for (dis-) aggregation (square grid created in simulations (Fig. 9c and d). The hazard increases in both areal and
QGIS, using Vector - > Research Tools - > Vector grid), (b) digital surface models (DSM) temporal extent with both open spaces and areas adjacent to sunlit
and canopy digital surface mode (CDSM) derived from an airborne LiDAR dataset walls being identified (cf. Fig. 9b). Walls have greater influence as
obtained in the summers of 2005 and 2008 (Martin Holt, Infoterra Ltd., personal
air temperatures become warmer, as surface temperature of walls
communication in 2011) and (c) population density (ONS, 2011). Population infor-
mation can be (dis-)aggregated based on the polygon grid, through QGIS tool Zonal increase and emit more longwave radiation while the shortwave
Statistics. radiation in open spaces remains constant.
Fig. 5. The UMEP Source Area Model tool calculating the source area for: (a) individual meteorological conditions and (b) varying conditions to generate a source area climatology.
Required input information includes a measurement location (x, y and z), a surface elevation database and meteorological conditions.
Fig. 6. Results from applications of SUEWS in (aec) London, UK, 17 Jul - 27 Oct 2015 and (def) Helsinki SeptembereNovember 2010. Examples of output from SUEWS Analyzer: (a)
Average daytime sensible heat flux (QH), (b) time series of net radiation (Q*) and QH for grid ID 44, (c) scatterplot between hourly values of Q* and QH grid ID 44. (d) Aerial image of
the water monitoring area (©2011 Kaupunkimittausosasto, Helsinki, Finland), (e) LiDAR derived land cover fractions (Nordbo et al., 2015), and (f) scatterplot between surface runoff
and precipitation for different planning scenarios. Base run is the current case and in the alternative scenarios paved surfaces have been changed to street trees and grass areas. For
visualisation, only events with runoff >4 mm have been plotted.
To investigate patterns at the city scale, the influence of building

and vegetation density on Tmrt across Stockholm is examined at a
pixel resolution of 1 m. The density of buildings and Tmrt show no
strong correlation. However, there is a clear relation between Tmrt at
2 p.m. and vegetation density at the 500 m scale (Fig. 10). This
demonstrates that increasing vegetation in urban areas could
reduce Tmrt and mitigate heat stress.
3.4. Application example 4- Solar Energy on Building Envelopes
The contrast between sunlit and shaded surfaces can explain

micro-scale differences in urban climate, for example spatial vari-
Fig. 7. Workflow and geodata used for analysing mean radiant temperature using
SOLWEIG in UMEP. Bold outlines indicate mandatory items. Colour coding as in Fig. 3.
ation in road surface temperatures (Hu et al., 2015). In UMEP, sunlit
(For interpretation of the references to colour in this figure legend, the reader is fractions are computed using high resolution DSMs and the Shad-
referred to the web version of this article.) owCalculator. The shadow casting algorithm uses sequential
Fig. 8. SOLWEIG (aed) spatial input and (eef) output data displayed with SOLWEIG Analyzer. Inputs include “raw” data (a) digital surface models (DSM and CDSM), (b) ground
cover, and UMEP derived data (see text) (c) sky view factor from buildings and vegetation, (d) wall pixels and height. The pixel resolution here is 1 m. Tmrt ( C) results for 26 July
2006 Gothenburg: (e) at 1 p.m., and (f) hourly Tmrt and air temperature ( C) for the courtyard point of interest (POI). The POI can be represented in any GDAL/OGR (a computer
software library for reading and writing raster and vector geospatial data formats) point vector layer.
Fig. 9. The Civic Square (Medborgarplatsen) in Stockholm, Sweden (a) mean radiant temperature (Tmrt) at 14:00 using meteorological forcing from a hot summer day (28 July 1994)
and (b) hazard map based on number of hours when Tmrt > 59.4 C on one day, (c) same as (b) but with a 2 C and (d) 4 C increase in air temperature.
computation of ‘shadow volumes’ (Ratti and Richens, 1999) with a

raster DSM (Ratti and Richens, 2004; Lindberg and Grimmond,
2010).
To map potential solar energy production, SEBE (Solar Energy
on Building Envelopes) can calculate irradiances at pixel resolution
on building roofs and walls using a 2.5-dimensional model.
Observed solar radiation data are used with high resolution DSMs
to derive accurate irradiances for the surfaces modelled (Lindberg
et al., 2015a,b).
SEBE has been applied to several cities in Sweden (http://
www.urban-climate.net/umep/Example_Applications). Fig. 11
shows a snapshot of an online mapping service where irradi-
ance on roofs in Uppsala (14 km2) has been modelled in UMEP at
0.25 m resolution. The surface data are a combination of airborne
LiDAR data and 3D vector polygons representing roof structures.
For each building, post-processing analysis derived several sta-
tistics including areal extent (m2) of roofs and walls suitable for
solar energy production (Fig. 11a) and the 3D distribution of solar
irradiance (Fig. 11b).
3.5. Application example 5 e anthropogenic heat fluxes
Anthropogenic heat flux (QF), heat released directly by humans Fig. 10. SOLWEIG output showing the relation between average ground surface Tmrt and
and their activities (Sailor, 2011), is a distinct feature of urban vegetation volume for 500 m 500 m grids in Stockholm, Sweden on the 28 July at 2 p.m.
Fig. 11. SEBE results for an online application for rooftop irradiance in central Uppsala, Sweden, (a) pop-up window with statistics for one of the buildings, (b) 3D view of shortwave
irradiance (walls and roof) for the building. (www.uppsala.se/solkarta) Meteorological data used in this example originate from a meso-scale model for solar radiation (STRÅNG,
www.strang.smhi.se) developed by the Swedish Meteorological and Hydrological Institute (SMHI).
areas with significant impact on energy and water exchanges. In shows the spatial distribution of each component, and the
UMEP, in addition to the methods within SUEWS, QF can be resultant larger QF towards the city centre (Fig. 12b, d, f). The
modelled using two standalone approaches (Appendix 2): (i) LQF, assumptions applied here limit LQF to spatial scales of the order
which uses the LUCY methodology (Allen et al., 2011; Lindberg 1 km (Gabey et al., 2017). At spatial scales less than 1 km,
et al., 2013a) and provides simple estimates at low spatial reso- movements of people, for example from home to work, are
lution; and (ii) GQF, which is a reimplementation of GreaterQF important and the structure of the road network needs to be
(Iamarino et al., 2012), and produces high-resolution estimates captured (Gabey et al., 2017). Example output maps from GQF
with greater insights into specific types of energy use. In UMEP, (Fig. 12a, c, e) show order-of-magnitude agreement with LQF,
each has been supplemented with a spatial input data pre- with a notably different spatial structure in QF,T (Fig. 12c, d)
processor that makes use of standard GIS data formats, and out- because road network topology is used rather than population
puts spatially and temporally-resolved QF estimates of traffic count. The effect of using the workday (GQF) rather than resi-
(QF,T), metabolic (QF,M) and building (QF,B) emissions in Universal dential (LQF) population on the daytime metabolic emission
Coordinated Time. The LQF results can be incorporated into the (Fig. 12e, f) is visible as a strong enhancement in the centre of the
meteorological data used to force SUEWS via the refinement stage city. A similar enhancement is also evident in the building
of the Download data (WATCH) tool. emissions (Fig. 12 a, b), which is attributable to the use of
An example of LQF output across Greater London in October spatially-resolved energy consumption in GQF compared with
2015 using Local Super Output Area (LSOA) population data residential population-based attribution in LQF.
UMEP is under active development and refinement. It is

designed as an open source tool, the development team welcomes
all kinds of collaboration through, for example, submission of
comments or issues to the repository (www.bitbucket/fredrik_ucg/
umep/), participation in coding, addition of new features and
development of new tutorials for users. The online manual provides
more details on how to participate (http://www.urban-climate.net/
umep/UMEP_Manual). Planned developments include tools for
pedestrian wind and thermal comfort indices.
Acknowledgements
The Greater London Authority LiDAR dataset are used courtesy

of Matthew Thomas (Greater London Authority) and data from a
NERC/ARSF (GB08/19) flight. The Stockholm data and results are
used courtesy of Anette Jansson (Environmental Office, City of
Stockholm). Financial support was provided by FORMAS e the
Swedish Research Council for Environment, Agricultural Sciences
and Spatial Planning, Met Office Climate Science for Service Part-
nership (CSSP) China as part of the Newton Fund, H2020-EO-1-
2014 Project 637519: URBANFLUXES, NERC Case Studentship (with
RMS) (supported CWK), British Council (funded BH), EPSRC stu-
dentship (ICT), and University of Reading. Work (SG, AG, NK, TS,
HCW) and visits (YYC, DWL) to the UK were supported by the UK-
China Research & Innovation Partnership Fund through the Met
Office Climate Science for Service Partnership (CSSP) China as part
of the Newton Fund. The UMEP plug-in can be downloaded from
http://www.urban-climate.net/umep.
Software availability
Fig. 12. Example outputs from (left) GQF and (right) LQF Greater London on 2nd
October 2015 at 11:00 UTC. (a,b) Building, (c,d) transport and (e,f) metabolic emissions Name of software: Urban Multi-scale Environmental Predictor
are shown to highlight the different spatial distributions that the models produce. The
(UMEP)
same colour scale is used for all maps. Appendix 2 provides more details about the two
methods. Developers: Fredrik Lindberg, Sue Grimmond, Andrew Gabey, Bei
Huang, Christoph W Kent, Ting Sun, Natalie E
Theeuwes, Isabella Capel-Timms, Leena Ja€rvi, Helen C
4. Concluding comments Ward, Yuanyong Chang, Niklas Krave, D Meyer, Frans
€stberg, Lingbo Xue, Zhe
Olofson, Jianguo Tan, Dag Wa
The city based climate service tool UMEP (Urban Multi-scale Zhang
Environmental Predictor) is introduced through a series of appli- Contact: Fredrik Lindberg: [email protected], þ46 31 786 2606
cations. The QGIS plug-in has a coupled modelling system of “state Sue Grimmond: [email protected]. þ44 118 378
of the art” 1-D and 2-D models which can provide estimates of 6248
essential urban climate processes. It also provides tools for deter- Year first available: 2015
mining parameters for more complex 3-D models. A key contri- Hardware required: NA
bution of UMEP is to provide a method to consistently determine Software required: QGIS 2.X
model parameters across a suite of models and applications. This Operation system required: OS independent
serves to ensure consistency in theoretical assumptions between Program language: Python
models, data analysis, observations, evaluation and applications Program size: ~2.5 Mb (compressed), ~9 Mb (uncompressed)
(different scales, applications and end users). Common processing Availability and cost: Open source (no cost)
tools also enable rapid updates when new data become available Repository: www.bitbucket.org/fredrik_ucg/umep
(for example, release of new national statistical data used in the Webpage: www.urban-climate.net/umep
anthropogenic heat flux) or when new parameterisations are
developed (for example, new aerodynamic roughness models of Appendix 1. Meteorological information
Kent et al., 2017a,b) which can then be used to understand flux
measurements (Source Area Model) and to perform energy and UMEP uses a common data format for meteorological data
water balance calculations (SUEWS). (Table A1.1). Preparing Existing Data (Figure A1.1), imports variables
Example applications have been presented to illustrate UMEP's from ASCII files and allows for time related variables (year, day of
potential, specifically of the identification of heat waves and cold year, hour and minute) and other time formats (such as month, day
waves in cities (ExtremeFinder); the implications of green infra- of month etc.) in preparation for analysis. During this process, some
structure on runoff (SUEWS); micro-scale heat stress (SOLWEIG); quality control is performed to ensure the data are within reason-
solar energy production (SEBE); and sources of anthropogenic able limits (Table A1.1, Figure A1.1) and to identify missing time
(human-generated) heat (LQF, GQF). Each application draws on intervals.
different combinations of UMEP tools.
Fig. A1.1. Example dialogue box in UMEP - Meteorological processor.
data. However, for the incoming solar radiation the times of

Download data (WATCH) provides another source of meteo- sunrise and sunset are calculated to ensure that the day length is
rological forcing data (section 3.1), global reanalysis products, if as expected and the interpolated data are adjusted to take this
observed data are not available. However, the coarse spatial reso- into account. For instance, if sunrise and sunset times are 6:00 LT
lution of reanalysis products (e.g. grids of 0.5 ) means that down- (local time) and 19:00 LT, respectively, only the interpolated
scaling is needed prior to their use for local-scale urban values between 6:00 LT and 19:00 LT will have non-zero short
hydrological modelling (Wilby et al., 2000; Fowler et al., 2007; wave radiation, but are rescaled for that period retaining the
Bastola and Misra, 2014). The methods used to prepare the data daily average.
are based on Best and Grimmond (2016), Kokkonen et al. (2017) and The relative humidity (RH) is obtained from the specific hu-
Ward et al. (2017a). These involve: midity (qv , in kg kg-1) as follows:RH ¼ eeas 100
(i) Identification of the area of interest, (ii) download of the data, With ea and es the actual and saturation vapor pressures (Buck,
and (iii) generation of a UMEP-formatted file with data aligned to 1981) given by
the appropriate time zone at hourly intervals.
These UMEP WFDEI data corrections are dependent on the RV qv p
meteorological variables, as some are instantaneous values ea ¼
Rd þ qv ðRv Rd Þ
whereas others are averages for a 3-h period (Weedon et al., 2011.
Ward et al., 2017a). The 3-h data are linearly interpolated to 1-h
8
>
>
>
0 1
>
> TC
>
> B 18:678 T C
>
> B 234:5 C C
>
> B C 1:00072 þ p 3:2 106 þ 5:9 1010 T2 TC 0 C
>
> 6:1121 expB C C
>
> @ TC þ 257:14 A
>
>
>
<
es ¼ 0
>
> 1
>
> T
>
> B 23:036 þ C
T C
>
> B 333:7 C C
>
> B C 1:00022 þ p 3:83 106 þ 6:4 1010 T2 TC < 0 C
>
> 6:1115 expB C C
>
> @ TC þ 279:82 A
>
>
>
>
:
where Rv ¼ 461:5 J K1 kg1 the specific gas constant for water Appendix 2. Anthropogenic heat flux
vapour, Rd ¼ 287:04 J K1 kg1 the specific gas constant for dry
air, p the atmospheric pressure (kPa) and TC the air temperature LQF is designed to provide “out-of-the-box” anthropogenic heat
( C). flux (QF) estimates at 1-h time steps. The minimum user-provided
Other corrections consider the effect of elevation. With a input data required are the spatially-resolved population count and
coarse grid the city or of area of interest may be at a lower a daily mean air temperature time series. Both data sets can be
elevation than the surroundings. As one example, Vancouver, the obtained using tools within UMEP. Heat fluxes are estimated using
real weighted mean WFDEI grid height includes mountains but a top-down methodology that draws on a database of national
much of the city is much lower (Kokkonen et al., 2017). Thus, energy consumption, population and vehicle ownership statistics.
consideration needs to be given to: (i) the elevation of the area of LQF attributes this energy consumption and traffic based on local
interest, (ii) the height of the roughness elements (buildings, population count variations, and estimates QF in each population
trees), and (iii) the appropriate height for the forcing data for the area. It is possible to replace the national data with provincial or
simulation. In UMEP, the WFDEI temperature and pressure values smaller regions if the data are available.
are adjusted to the simulation height using environmental lapse The database contains diurnal variations for metabolism,
rate (G ¼ 6.5 C km1) and the hypsometric equation (Weedon traffic flow and building energy consumption, which are
et al. 2011). RH, used by SUEWS, is calculated from the WFDEI optionally overridden with user-specified versions. Weekend/
specific humidity assuming it is constant with altitude to avoid weekday variations are captured for buildings (QF,B) and trans-
supersaturation. Interpolation and altitude corrections may port (QF,T), while the user-specified versions provide control over
require a spin-up period to avoid interpolation errors and each day of the week. Day-to-day building energy consumption
missing data points at the beginning and at the end of the time is estimated using a database-held, country-specific temperature
series. Thus, longer periods should be used for analysis than the response function and assumption about the prevalence of air-
period of specific interest. conditioning (Lindberg et al., 2013a). A blanket weekend traffic
reduction is also applied to capture day-of-week traffic flow
changes.
Table A1.1
Meteorological data used in various components of UMEP are formatted in the order indicated. The Metdata Pre-processor tool does simple quality control to ensure data are
within acceptable ranges. Not all variables are required in UMEP (with the exception of the four time-related columns): All e any that use meteorological data, SU e SUEWS, SO
e SOLWEIG, SA e source area model, SE e SEBE, for some application these data are required (Bold). The variables extracted by Download data (WATCH) are indicated
(column 5). When data are not needed it can be assigned 999 to indicate no data available. Day of year is used instead of date within a month for simplicity. The hour and
minute is local standard time and refers to time ending.
No. Description (units) Accepted range Used by WATCH
1 Year [YYYY] All

2 Day of year [DOY] All
3 Hour [H] All
4 Minute [M] All
5 Net all-wave radiation [W m2] 200 to 800 SU
6 Sensible heat flux [W m2] 200 to 750 SU
7 Latent heat flux [W m2] 100 to 650 SU
8 Storage heat flux [W m2] 200 to 650 SU
9 Anthropogenic heat flux [W m2] 0 to 1500 SU
10 Wind speed [m s1] 0.001 to 60 SU/SA X
11 Relative Humidity [%] 5 to 100 SU/SO/SE X
12 Air temperature [ C] 30 to 55 SU/SO/SE X
13 Barometric pressure [kPa] 90 to 107 SU/SO X
14 Rainfall [mm] 0 to 30 (per 5 min) SU X
15 Incoming shortwave radiation [W m2] 0 to 1200 SU/SO/SE X
16 Snow [mm] 0 to 300 (per 5 min) SU X
17 Incoming longwave radiation [W m2] 100 to 600 SU X
18 Cloud fraction [tenths] 0 to 1 SU
19 External water use [m3] 0 to 10 (per 5 min) SU
20 Observed soil moisture [m3 m3 or kg kg1] 0.01 to 0.5 SU
21 Observed leaf area index [m2 m2] 0 to 15 SU
22 Diffuse shortwave radiation [W m2] 0 to 600 SO/SE
23 Direct shortwave radiation [W m2] 0 to 1200 SO/SE
24 Wind direction [ ] 0 to 360 SU/SA
25 Friction velocity [m s1] All SA
26 Standard Deviation (sigma) of transverse-wind velocity [m s1] All SA
27 Obukhov Length [m] All SA
28 Boundary layer height [m] All SA
Table A2.1
Input data files required by the LQF model
Dataset Type Remarks
Population count Polygon shapefile Basis of spatial variations of QF components

Daily mean temperature CSV file(s) Basis of day-to-day building energy consumption variations
Traffic diurnal profile CSV file(s) Optional
Building energy diurnal profile CSV file(s) Optional
LQF database Spatialite file Downloaded from UMEP website; can be edited by user
GQF produces QF estimates at higher spatial resolution and 30- References

min time steps, but requires comprehensive input data (Table A2.2)
to support this. The shift in population distribution during working Alcoforado, M.J., Andrade, H., Lopes, A., Vasconcelos, J., 2009. Application of climatic
guidelines to urban planning: the example of Lisbon (Portugal). Landsc. Urban
days is captured using residential and workday population data- Plan. 90 (1e2), 56e65.
sets, with morning and evening transition periods. Road transport Alexander, P.J., Mills, G., Fealy, R., 2015. Using LCZ data to run an urban energy
emissions are calculated using a vector map containing every road balance model. Urban Clim. 13, 14e37.
Alexander, P.J., Bechtel, B., Chow, W.T.L., Fealy, R., Mills, G., 2016a. Linking urban
link, along with the road class label and traffic flow broken down by climate classification with an urban energy and water budget model: Multi-site
vehicle type. Maps of residential and non-residential gas and and multi-seasonal evaluation. Urban Clim. 17, 196e215.
electricity consumption across the city are used to estimate build- Alexander, P.J., Fealy, R., Mills, G., 2016b. Simulating the impact of different urban
development pathways on the local climate: a scenario-based analysis in the
ing emissions. The model therefore sub-divides QF,T by vehicle type, greater Dublin region, Ireland. Landsc. Urban Plan. 152, 72e89.
and QF,B by fuel and consumer, and estimates the spatial distribu- Allen, L., Lindberg, F., Grimmond, C.S.B., 2011. Global to city scale urban anthropo-
tion of transport emissions independently of the population dis- genic heat flux: model and variability. Int. J. Climatol. 31 (13), 1990e2005.
Ao, X., Grimmond, C.S.B., Chang, Y., Liu, D., Tang, Y., Hu, P., Wang, Y., Zou, J., Tan, J.,
tribution. Model spatial resolution is dictated by the input
2016. Heat, water and carbon exchanges in the tall megacity of Shanghai:
residential and workday population datasets, with the annual en- challenges and results. Int. J. Climatol. 36 (14), 4608e4624.
ergy consumption data disaggregated to common spatial units Baklanov, A., Grimmond, C.S.B., Carlson, D., Terblanche, D., Tang, X., Bouchet, V.,
Lee, B., Langendijk, G., Kolli, R.K., Hovsepyan, A., 2017. From urban meteorology,
based on population.
climate and environment research to integrated city services. Urban Clim.
Day-to-day variations in building energy consumption and https://doi.org/10.1016/j.uclim.2017.05.004.
hence QF,B are captured using empirical demand data from utility Bastola, S., Misra, V., 2014. Evaluation of dynamically downscaled reanalysis pre-
companies, and half-hourly variations are obtained from week-long cipitation data for hydrological application. Hydrol. Process. 28, 1989e2002.
https://doi.org/10.1002/hyp.9734.
diurnal profiles, with separate profiles specified for each energy Best, M.J., Grimmond, C.S.B., 2016. Modelling the Bowen ratio at a number of urban
consumption category. Temporal variations in transport are gov- sites over a range of vegetation cover. J. Hydrometeorol. http://journals.
erned by a week-long diurnal profile for each vehicle type, in which ametsoc.org/doi/10.1175/JHM-D-15-0126.1.
Best, M.J., Pryor, M., Clark, D.B., Rooney, G.G., Essery, R.L.H., Me nard, C.B.,
some days have greater mean values than others to reflect day-to- Edwards, J.M., Hendry, M.A., Porson, A., Gedney, N., Mercado, L.M., Sitch, S.,
day traffic variations. This segmentation means the temporal evo- Blyth, E., Boucher, O., Cox, P.M., Grimmond, C.S.B., Harding, R.J., 2011. The joint
lution of each QF component in a modelled area depends on the UK land environment simulator (JULES), model description e Part 1: energy and
water fluxes. Geosci. Model Dev. 4, 677e699.
balance of vehicle types, energy consumers and the residential to Bottema, M., Mestayer, P.G., 1998. Urban roughness mappingevalidation techniques
workday population ratio. GQF also partitions QF,T, QF,B and meta- and some first results. J. Wind Eng. Ind. Aerodyn. 74, 163e173.
bolism (QF,M) into sensible, latent and (for buildings) wastewater Bruse, M., Fleer, H., 1998. Simulating surface-plant-air interactions inside urban
environments with a three dimensional numerical model. Environ. Model.
fractions. These can be included or excluded from the modelled QF
Softw. 13 (3e4), 373e384.
at model run time. Buck, A.L., 1981. New equations for computing vapor pressure and enhancement
factor. J. Appl. Meteorol. 20 (12), 1527e1532.
Busuioc, A., Caian, M., Cheval, S., Bojariu, R., Boroneanţ, C., Baciu, M., Dumitrescu, Al,
Table A2.2 2010. Variabilitatea Şi Schimbarea Climei În Rom^ ania, Edit. ProUniversitaria,
Input data sources required by the GQF model. Bucureşti, p. 226 (cited by Micu 2012).
Chen, L., Yu, B., Yang, F., Mayer, H., 2016. Intra-urban differences of mean radiant
Data source Type Remarks temperature in different urban settings in Shanghai and implications for heat
stress under heat waves: a GIS-based approach. Energy Build. 130, 829e842.
Residential Polygon Used for metabolism estimates and to spatially Ching, J., Mills, G., Bechtel, B., See, L., Feddema, J., Wang, X., Ren, C., Brousse, O.,
population count shapefile disaggregate energy consumption data to Martilli, A., Neophytou, M., Mouzourides, P., Stewart, I., Hanna, A., Ng, E.,
Workday population Polygon common spatial units. Foley, M., Alexander, P., Aliaga, D., Niyogi, D., Shreevastava, A., Bhalachandran, P.,
count shapefile Masson, V., Hidalgo, J., Fung, J., Andrade, M., Baklanov, A., Dai, W., Milcinski, G.,
Road network and Line Each road segment with road classification Demuzere, M., Brunsell, N., Pesaresi, M., Miao, S., Mu, Q., Chen, F., Theeuwes, N.,
traffic flow shapefile and annual average daily total traffic for each 2017. World urban database and access portal tools (WUDAPT), an urban
vehicle type weather, climate and environmental modeling infrastructure for the anthro-
Residential energy Shapefiles Spatially disaggregated using residential pocene. Bull. Am. Meteorol. Soc. (under review).
Chrysoulakis, N., Lopes, M., San Jose , R., Grimmond, C.S.B., Jones, M.B., Magliulo, V.,
consumption: population
Klostermann, J.E.M., Synnefa, A., Mitraka, Z., Castro, E.A., Gonza lez, A., Vogt, R.,
Residential
Vesala, T., Spano, D., Pigeon, G., Freer-Smith, P., Staszewski, T., Hodges, N.,
electricity
Mills, G., Cartalis, C., 2013. Sustainable urban metabolism as a link between bio-
Residential gas physical sciences and urban planning: the BRIDGE project. Landsc. Urban Plan.
Residential economy 112, 100e117. https://doi.org/10.1016/j.landurbplan.2012.12.005.
7 Dee, D.P., Uppala, S.M., Simmons, A.J., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U.,
Non-residential Shapefiles kWh per year, spatially-resolved within city Balmaseda, M.A., Balsamo, G., Bauer, P., Bechtold, P., Beljarrs, A.C.M., van de
energy Spatially disaggregated using workday Berg, L., Bidlot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Geer, A.J.,
consumption: population Haimberger, L., Healy, S.B., Hersbach, H., Holm, E.V., Isaksen, L., Kallberg, P.,
Non-residential Kohler, M., Matricardi, M., McNally, A.P., Monge-Sanz, B.M., Morcrette, J.-J.,
electricity Park, B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thepaut, J.-N., Vitart, F., 2011.
Non-residential gas The ERA-Interim reanalysis: configuration and performance of the data
Non-residential assimilation system. Q. J. R. Meteorol. Soc. 137, 553e597.
Demuzere, M., Harshan, S., Ja €rvi, L., Roth, M., Grimmond, C.S.B., Masson, V., 2017.
other
Impact of urban canopy models and external parameters on the modelled ur-
Daily energy CSV file Relative variations used to scale annual to
ban energy balance in a tropical city. Q. J. R. Meteorol. Soc. https://doi.org/
consumption time daily values.
10.1002/qj.3028.
series Applied to both residential and non- Dong, Y., Varquez, A.C.G., Kanda, M., 2017. Global anthropogenic heat flux database
Gas residential sectors. with high spatial resolution. Atmos. Environ. 150, 276e294. https://doi.org/
Electricity 10.1016/j.atmosenv.2016.11.040.
Vehicle diurnal CSV file Time series for each vehicle type Dudhia, J., 2014. A history of mesoscale model Development. Asia-Pac. J. Atmos. Sci.
profiles 50, 121e131.
Building energy CSV file Time series for energy consumption dataset Eliasson, I., 2000. The use of climate knowledge in urban planning. Landsc. Urban
diurnal profiles Plan. 48, 31e44.
Metabolic diurnal CSV file Time series defining metabolic rate per person Fischer, E.M., Scha€r, C., 2010. Consistent geographical patterns of changes in high-
profiles and transition between workday and impact European heatwaves. Nat. Geosci. 3 (6), 398e403.
residential population. Fowler, H.J., Blenkinsop, S., Tebaldi, C., 2007. Linking climate change modelling to
Vehicle fuel CSV file Amount of fuel consumed per vehicle class impacts studies: recent advances in downscaling techniques for hydrological
modelling. Int. J. Climatol. 27, 1547e1578. https://doi.org/10.1002/joc.1556.
efficiencies and fuel type
Gabey, A., Grimmond, C.S.B., Capel-Timms, I., 2017. Anthropogenic Heat Flux: at Lindberg, F., Grimmond, C.S.B., 2011a. The influence of vegetation and building
what spatial resolution is a simple model reasonable? Theor. Appl. Climatol. morphology on shadow patterns and mean radiant temperature in urban areas:
(under review). model development and evaluation. Theor. Appl. Climatol. 105 (3), 311e323.
Goodwin, N.R., Coops, N.C., Tooke, T.R., Christen, A., Voogt, J.A., 2009. Characterizing Lindberg, F., Grimmond, C.S.B., 2011b. Nature of vegetation and building
urban surface cover and structure with airborne lidar technology. Can. J. Remote morphology characteristics across a city: influence on shadow patterns and
Sens. 35, 297e309. mean radiant temperatures in London. Urban Ecosyst. 14 (4), 617e634.
Grimmond, C.S.B., 2013. Observing London: Weather Data Needed for London to Lindberg, F., Holmer, B., Thorsson, S., 2008. SOLWEIG 1.0 e modelling spatial vari-
Thrive, London Climate Change Partnership, p. 82. http://climatelondon.org.uk. ations of 3D radiant fluxes and mean radiant temperature in complex urban
London Climate Change Partnership, Lloyds, Met Office. settings. Int. J. Biometeorol. 52, 697e713.
Grimmond, C.S.B., Oke, T.R., 1991. An evaporation-interception model for urban Lindberg, F., Grimmond, C.S.B., Yogeswaran, N., Kotthaus, S., Allen, L., 2013a. Impact
areas. Water Resour. Res. 27, 1739e1755. of city changes and weather on anthropogenic heat flux in Europe 1995e2015.
Grimmond, C.S.B., Oke, T.R., 1999. Aerodynamic properties of urban areas derived Urban Clim. 4, 1e15.
from analysis of surface form. J. Appl. Meteorol. 38, 1262e1292. Lindberg, F., Holmer, B., Thorsson, S., Rayner, D., 2013b. Characteristics of the mean
Grimmond, C.S.B., Tang, X., Baklanov, A., 2014. Towards integrated urban weather, radiant temperature in high latitude citiesdimplications for sensitive climate
environment and climate services. WMO Bull. 63 (1), 10e14. https:// planning applications. Int. J. Biometeorol. 58 (5), 613e627.
googledrive.com/host/0BwdvoC9AeWjUazhkNTdXRXUzOEU/bulletin_63-1_en. Lindberg, F., Jonsson, P., Honjo, T., Wa €stberg, D., 2015b. Solar energy on building
pdf. http://library.wmo.int/opac/index.php?lvl¼notice_display&id¼15965. envelopes - 3D modelling in a 2D environment. Sol. Energy 115, 369e378.
Herbert, J.M., Johnson, G.T., Arnfield, A.J., 1998. Modelling the thermal climate in city Lindberg, F., Grimmond, C.S.B., Martilli, A., 2015a. Sunlit fractions on urban facets -
canyons. Environ. Model. Softw. 13 (3), 267e277. impact of spatial resolution and approach. Urban Clim. 12, 65e84. https://
Horton, R., Rosenzweig, C., Solecki, W., Bader, D., Sohl, L., 2016. Climate science for doi.org/10.1016/j.uclim.2014.11.006.
decision-making in the New York metropolitan region. In: Parris, A.S., Lindberg, F., Onomura, S., Grimmond, C.S.B., 2016a. Influence of ground surface
Garfin, G.M., Dow, K., Meyer, R., Close, S.L. (Eds.), Climate in Context: Science characteristics on the mean radiant temperature in urban areas. Int. J. Bio-
and Society Partnering for Adaptation. John Wiley & Sons, Ltd, Chichester, UK. meteorol. 60 (9), 1439e1452.
https://doi.org/10.1002/9781118474785.ch3. Lindberg, F., Lau, K., Rayner, D., Thorsson, S., 2016b. The impact of urban planning
Hu, Y., Almkvist, E., Lindberg, F., Bogren, J., Gustavsson, T., 2015. The use of screening strategies on heat stress in a climate-change perspective. Sustain. Cities Soc. 25,
effects in modelling route-based daytime road surface temperature. Theor. 1e12.
Appl. Climatol. 125 (1), 303e319. Macdonald, R., Griffiths, R., Hall, D., 1998. An improved method for the estimation of
Iamarino, M., Beevers, S., Grimmond, C.S.B., 2012. High-resolution (space, time) surface roughness of obstacle arrays. Atmos. Environ. 32, 1857e1864.
anthropogenic heat emissions: London 1970e2025. Int. J. Climatol. 32 (11), Masson, V., Le Moigne, P., Martin, E., Faroux, S., Alias, A., Alkama, R., Belamari, S.,
1754e1767. Barbu, A., Boone, A., Bouyssel, F., Brousseau, P., Brun, E., Calvet, J.C., Carrer, D.,
J€
anicke, B., Meier, F., Lindberg, F., Schubert, S., Scherer, D., 2016. Towards city-wide, Decharme, B., Delire, C., Donier, S., Essaouini, K., Gibelin, A.L., Giordani, H.,
building-resolving analysis of mean radiant temperature. Urban Clim. 15, 83e98. Habets, F., Jidane, M., Kerdraon, G., Kourzeneva, E., Lafaysse, M., Lafont, S.,
€rvi, L., Grimmond, C.S.B., Christen, A., 2011. The surface urban energy and water
Ja LebeaupinBrossier, C., Lemons, A., Mahfouf, J.-F., Marguinaud, P., Mokhtari, M.,
balance scheme (SUEWS): evaluation in los angeles and vancouver. J. Hydrol. Morin, S., Pigeon, G., Salgado, R., Seity, Y., Taillefer, F., Tanguy, G., Tulet, P.,
411, 219e237. Vincendon, B., Vionnet, V., Voldoire, A., 2013. The SURFEXv7.2 land and ocean
€rvi, L., Grimmond, C.S.B., Taka, M., Nordbo, A., Seta
Ja €l a
€, H., Strachan, I.B., 2014. surface platform for coupled or offline simulation of earth surface variables and
Development of the surface urban energy and water balance scheme (SUEWS) fluxes. Geosci. Model Dev. 6, 929e960.
for cold climate cities. Geosci. Model Dev. 7, 1691e1711. https://doi.org/10.5194/ Masson, V., Marchadier, C., Adolphe, L., Aguejdad, R., Avner, P., Bonhomme, M.,
gmd-7-1691-2014. Bretagne, G., Briottet, X., Bueno, B., de Munck, C., Doukari, O., Hallegatte, S.,
€rvi, L., Grimmond, C.S.B., McFadden, J.P., Christen, A., Strachan, I., Taka, M.,
Ja Hidalgo, J., Houet, T., Le Bras, J., Lemonsu, A., Long, N., Moine, M.-P., Morel, T.,
Warsta, L., Heimann, M., 2017. Warming effects on the urban hydrology in cold Nolorgues, Pigeon, L.G., Salagnac, J.-L., Viguie, V., Zibouche, K., 2014. Adapting cities
climate regions. Sci. Rep. 7, 5833. https://doi.org/10.1038/s41598-017-05733-y. to climate change: a systemic modelling approach. Urban Clim. 10, 407e429.
Kanda, M., Inagaki, A., Miyamoto, T., Gryschka, M., Raasch, S., 2013. A new aero- Mayer, H., Ho €ppe, P., 1987. Thermal comfort of man in different urban environ-
dynamic parametrization for real urban surfaces. Bound.-Layer Meteorol. 148, ments. Theor. Appl. Clim. 38, 43e49.
357e377. Meehl, G.A., Tebaldi, C., 2004. More intense, more frequent, and longer lasting heat
Karsisto, P., Fortelius, C., Demuzere, M., Grimmond, C.S.B., Oleson, K., Kouznetsov, R., waves in the 21st century. Science 305 (5686), 994e997.
Masson, V., J€ arvi, L., 2015. Seasonal surface urban energy balance and winter- Micu, D., 2012. Cold waves in the Romanian Carpathians, an indicator of negative
time stability simulated using three land-surface models in the high-latitude temperature extremes. In: Air and Water Components of the Environment
city Helsinki. Q. J. R. Meteorol. Soc. 142 (694), 401e417. https://doi.org/ Conference. http://aerapa.conference.ubbcluj.ro/2012/pdf/13%20D%20Micu.pdf.
10.1002/qj.2659. Millward-Hopkins, J., Tomlin, A., Ma, L., Ingham, D., Pourkashanian, M., 2011. Esti-
Keevallik, S., Vint, K., 2015. Temperature extremes and detection of heat and cold mating aerodynamic parameters of urban-like surfaces with heterogeneous
waves at three sites in Estonia. Proc. of the Est. Acad. Sci vol. 64 (No. 4), building heights. Bound-Layer Meteorol. 141, 443e465.
473e479. Nordbo, A., Karsisto, P., Matikainen, L., Wood, C.R., J€ arvi, L., 2015. Urban surface
Kent, C.W., Grimmond, C.S.B., Barlow, J., Gatey, D., Kotthaus, S., Lindberg, F., cover determined with airborne lidar at 2 m resolution - implications for sur-
Halios, C.H., 2017a. Evaluation of urban local-scale aerodynamic parameters: face energy balance modelling. Urban Clim. 13, 52e72.
implications for the vertical profile of wind speed and for source areas. Onomura, S., Grimmond, C.S.B., Lindberg, F., Holmer, B., Thorsson, S., 2015. Mete-
Boundary-Layer Meteorol. 164, 183e213. orological forcing data for urban outdoor thermal comfort models from a
Kent, C.W., Grimmond, C.S.B., Gatey, D., 2017b. Aerodynamic roughness parameters coupled convective boundary layer and surface energy balance scheme. Urban
in cities: inclusion of vegetation. J. Wind Eng. Industrial Aerodynamics 169, Clim. 11, 1e23.
168e176. ONS, 2011. Population Office for National Statistics. Residents, Workplace and
Kljun, N., Calanca, P., Rotach, M.W., Schmid, H.P., 2015. A simple two-dimensional Daytime Population e UK 2001 National Census [Online] Available from:
parameterisation for Flux Footprint Prediction (FFP). Geosci. Model Dev. 8, https://www.nomisweb.co.uk/census/2011. (Accessed 25 February 2013).
3695e3713. https://doi.org/10.5194/gmd-8-3695-2015, 2015. Ordnance_Survey, 2010. Crown Database Right 2010. An Ordnance Survey/EDINA
Kokkonen, T., Grimmond, C.S.B., Ra €ty, O., Ward, H.C., Christen, A., Oke, T., Supplied Service.
Kotthaus, S., Ja €rvi, L., 2017. Sensitivity of surface urban energy and water bal- Paton, F.L., Dandy, G.C., Maier, H.R., 2014. Integrated framework for assessing urban
ance scheme (SUEWS) to downscaling of reanalysis forcing data. Urban Clim. water supply security of systems with non-traditional sources under climate
https://doi.org/10.1016/j.uclim.2017.05.001. change. Environ. Model. Softw. 60, 302e319.
Kormann, R., Meixner, F.X., 2001. An analytical footprint model for non-neutral QGIS Development Team, 2017. QGIS Geographic Information System. Open Source
stratification. Bound.-Layer Meteorol. 99, 207e224. Geospatial Foundation Project. http://www.qgis.org/.
Kusaka, H., Kondo, H., Kikegawa, Y., Kimura, F., 2001. A simple single-layer urban Rafael, S., Martins, H., Sa, E., Carvalho, D., Borrego, C., Lopes, M., 2016. Influence of
canopy model for atmospheric models: comp. multi-layer slab models. Bound.- urban resilience measures in the magnitude and behaviour of energy fluxes in
Layer Meteorol. 101, 329e358. https://doi.org/10.1023/A:1019207923078. the city of Porto (Portugal) under a climate change scenario. Sci. Total Environ.
Lau, K., Lindberg, F., Rayner, D., Thorsson, S., 2014. The effect of urban geometry on 566e567, 1500e1510.
mean radiant temperature under future climate change: a study of three Eu- Ratti, C.F., Richens, P., 1999. Urban texture analysis with image processing tech-
ropean cities. Int. J. Biometeorol. 59 (7), 799e814. niques. In: Proceedings of the CAADFutures99 (Atlanta, GA).
Lau, K.K.-L., Ren, C., Ho, J., Ng, E., 2016. Numerical modelling of mean radiant Ratti, C.F., Richens, P., 2004. Raster analysis of urban form. Environ. Plann B Plann
temperature in high-density sub-tropical urban environment. Energy Build. 114, Des. 31, 297e309.
80e86. Raupach, M., 1994. Simplified expressions for vegetation roughness length and
Lawrence, D.M., Oleson, K.W., Flanner, M.G., Thornton, P.E., Swenson, S.C., zero-plane displacement as functions of canopy height and area index. Bound-
Lawrence, P.J., Zeng, X., Yang, Z.L., Levis, S., Sakaguchi, K., Bonan, G.B., Layer Meteorol. 71, 211e216.
Slater, A.G., 2011. Parameterization improvements and functional and structural Ren, C., Ng, E., Katzschner, L., 2011. Urban climatic map studies: a review. Int. J.
advances in version 4 of the Community Land Model. J. Adv. Model. Earth Syst. Climatol. 31 (15), 2213e2233.
3, 1e27. https://doi.org/10.1029/2011MS000045. Robinson, P.J., 2001. On the definition of a heat wave. J. Appl. Meteor. 40, 762e775,
Lindberg, F., Grimmond, C.S.B., 2010. Continuous sky view factor maps from high 10.1175/1520-0450(2001)040<0762:OTDOAH>2.0.CO;2.
resolution urban digital elevation models. Clim. Res. 42, 177e183. Saagi, R., Flores-Alsina, X., Kroll, S., Gernaey, K.V., Jeppsson, U., 2017. A model library
for simulation and benchmarking of integrated urban wastewater systems. Scinocca, J., Sobolowski, S., Suklitsch, M., Teichmann, C., Warrach-Sagi, K.,
Environ. Model. Softw. 93, 282e295. Wulkmeyer, C., Yiou, P., 2013. The simulation of European heat waves from an
Sailor, D.J., 2011. A review of methods for estimating anthropogenic heat and ensemble of regional climate models within the EURO-CORDEX project. Clim.
moisture emissions in the urban environment. Int. J. Climatol. 31 (2), 189e199. Dyn. 41, 2555e2575. https://doi.org/10.1007/s00382-013-1714-z.
Schoetter, R., Caiaux, J., Douville, H., 2014. Changes of western European heat wave Ward, H.C., Grimmond, C.S.B., 2017. Assessing the impact of changes in sur-
characteristics projected by the CMIP5 ensemble. Clim. Dyn. 1e16. face cover, human behaviour and climate on energy partitioning across
Solecki, W., 2015. Hurricane Sandy in New York, extreme climate events and the Greater London. Landsc. Urban Plan. 165, 142e161. https://doi.org/10.1016/
urbanization of climate change: perspectives in the context of sub-Saharan j.landurbplan.2017.04.001.
African cities. Curr. Opin. Environ. Sustain. 13, 88e94. https://doi.org/10.1016/ Ward, H.C., Kotthaus, S., Ja€rvi, L., Grimmond, C.S.B., 2016. Surface urban energy and
j.cosust.2015.02.007. water balance scheme (SUEWS): development and evaluation at two UK sites.
Srivastava, A.K., Rajeevan, M., Kshirsagar, S.R., 2009. Development of a high reso- Urban Clim. 18, 1e32. https://doi.org/10.1016/j.uclim.2016.05.001.
lution daily gridded temperature data set (1969e2005) for the Indian region. Ward, H.C., Tan, Y., Gabey, A.M., Kotthaus, S., Morrison, W.T.J., Grimmond, C.S.B.,
Atmosph. Sci. Lett. 10, 249e254. https://doi.org/10.1002/asl.232. 2017a. Impact of temporal resolution of precipitation forcing data on modelled
Stewart, I.D., Oke, T.R., 2012. Local climate zones for urban temperature studies. urban-atmosphere exchanges and surface conditions. Int. J. Climatol. https://
Bull. Am. Meteorol. Soc. 93 (12), 1879e1900. doi.org/10.1002/joc.5200.
Thom, J.K., Coutts, A.M., Broadbent, A.M., Tapper, N.J., 2016. The influence of Ward, H.C., Ja €rvi, L., Sun, T., Onomura, S., Lindberg, F., Olofson, F., Gabey, A.,
increasing tree cover on mean radiant temperature across a mixed develop- Grimmond, C.S.B., 2017b. SUEWS Manual V2017b. Department of Meteorology,
ment suburb in Adelaide, Australia. Urban For. Urban Green. 20, 233e242. University of Reading, Reading, UK. http://urban-climate.net/umep/SUEWS.
€rklund, J., Holmer, B., Rayner, D., 2010. Potential changes in
Thorsson, S., Lindberg, F., Bjo Watson, I.D., Johnson, G.T., 1987. Graphical estimation of sky view-factors in urban
outdoor thermal comfort conditions in Gothenburg, Sweden due to climate environments. J. Climatol. 7, 193e219.
change: the influence of urban geometry. Int. J. Climatol. 31 (2), 324e335. Weedon, G.P., Gomes, S., Viterbo, P., Shuttleworth, W.J., Blyth, E., Osterle, € H.,
Thorsson, S., Rocklo€v, J., Konarska, J., Lindberg, F., Holmer, B., Dousset, B., Rayner, D., Adam, J.C., Bellouin, N., Boucher, O., Best, M., 2011. Creation of the WATCH
2014. Mean radiant temperature - a predictor of heat related mortality. Urban Forcing data and its use to assess global and regional reference crop evapora-
Clim. 10 (2), 332e345. tion over land during the twentieth century. J. Hydrom. 12, 823e848.
Vaidyanathan, A., Kegler, S.R., Saha, S.S., Mulholland, J.A., 2016. A statistical Weedon, G.P., Balsamo, G., Bellouin, N., Gomes, S., Best, M., Viterbo, P., 2014. The
framework to evaluate extreme weather definitions from a health perspective: WFDEI meteorological forcing data set: WATCH Forcing Data methodology
a demonstration based on extreme heat events. Bull. Am. Meteorol. Soc. 97 (10), applied to ERA-Interim reanalysis data. Water Resour. Res. 50, 7505e7514.
1817e1830. Wilby, R.L., Hay, L.E., Gutowski Jr., W.J., Arrit, R.W., Takle, E.S., Pan, Z., Leavesley, G.H.,
Vautard, C., Gobiet, A., Jacob, D., Belda, M., Colette, A., De que
, M., Ferna ndez, J., Clark, M.P., 2000. Hydrological responses to dynamically and statistically
GarcíaDíez, M., Goergen, K., Güttler, I., Halenka, T., Karacostas, T., Katragkou, E., downscaled climate model output. Geophys. Res. Lett. 27, 1199e1202.
Keuler, K., Kotlarski, S., Mayer, S., van Meijgaard, M., Nikulin, G., Patar ci
c, M.,

UMEP Modelling

Uploaded by

Copyright:

Available Formats

UMEP Modelling

Uploaded by

Document Information

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

UMEP Modelling

Uploaded by

Copyright:

Available Formats

Environmental Modelling & Software 99 (2018) 70e87

Contents lists available at ScienceDirect

Environmental Modelling & Software

Urban Multi-scale Environmental Predictor (UMEP): An integrated

€stberg g, Lingbo Xue b, i,

1. Introduction climate services to those operating and planning different aspects

Fig. 1. Structure of UMEP (Urban Multi-scale Environmental Predictor).

Component Scales Description Example Applications Key References

Component Scales Description Example Applications Key References

Extreme Reference Index description

City Reference Variables evaluated/application

City Reference Variables Settings

Fig. 7 shows the UMEP workﬂow to examine 3D radiant ﬂuxes

To investigate patterns at the city scale, the inﬂuence of building

3.4. Application example 4- Solar Energy on Building Envelopes

The contrast between sunlit and shaded surfaces can explain

computation of ‘shadow volumes’ (Ratti and Richens, 1999) with a

3.5. Application example 5 e anthropogenic heat ﬂuxes

UMEP is under active development and reﬁnement. It is

The Greater London Authority LiDAR dataset are used courtesy

Fig. A1.1. Example dialogue box in UMEP - Meteorological processor.

data. However, for the incoming solar radiation the times of

No. Description (units) Accepted range Used by WATCH

1 Year [YYYY] All

Dataset Type Remarks

Population count Polygon shapeﬁle Basis of spatial variations of QF components

GQF produces QF estimates at higher spatial resolution and 30- References

You might also like