This notebook is part of the tutorials in the ASP summer school.

In the S2S verification tutorial, we use climpred https://climpred.readthedocs.io/en/stable/ to verify subseasonal-to-seasonal (S2S) forecasts against observations.

• Evaluate metrics across models and model versions
• Predictability limits
• State-dependent predictability
• contribute to climpred

• 'ASP_data_catalog.yml': intake catalog, see data_access_with_intake.ipynb
• s2s-climpred.yaml: conda environment file
• cluster.ipynb: Start a PBS-Cluster on casper, needed for big data (10GB+)
• climpred_*.ipynb: Jupyter notebooks about climpred and student projects

It is recommented to use the enviroment s2s-climpred on NCAR_casper.

Else create your own environment:

  conda activate
  conda env create -f s2s-climpred.yaml
  # update existing
  # conda env update -f s2s-climpred.yaml
  conda activate s2s-climpred

• xarray: working horse for geospatial data in python
  • documentation: xarray.pydata.org/
  • tutorial: https://xarray-contrib.github.io/xarray-tutorial/
• xskillscore: is built on top of xarray and provides metrics to climpred
  • documentation: https://xskillscore.readthedocs.io/en/stable/
  • quick-start: https://mybinder.org/v2/gh/xarray-contrib/xskillscore/master?urlpath=lab
• climpred:
  • documentation: https://climpred.readthedocs.io/en/stable/
  • data model: https://climpred.readthedocs.io/en/stable/setting-up-data.html
  • classes: https://climpred.readthedocs.io/en/stable/prediction-ensemble-object.html
  • list of initialized public datasets to work with: https://climpred.readthedocs.io/en/stable/initialized-datasets.html
  • terminology: https://climpred.readthedocs.io/en/stable/terminology.html
  • alignment: https://climpred.readthedocs.io/en/stable/alignment.html

Consider...

• raising an issue, which can be transferred to discussions
• reaching out on slack

• Aaron Spring
• Judith Berner
• Abby Jaye