Xclim: Climate Data Processing and Analysis for Everyone
summary
Library claims reproducible climate indicators at scale and demonstrates it: built on xarray and dask with Numba and Pint integrations, Xclim implements 187 indicators, ensemble selection, calendar harmonization and bias adjustment, and powers Environment and Climate Change Canada’s ClimateData.ca platform. Demos show unit-consistent results across Celsius/Fahrenheit/Kelvin, ensemble percentiles computed on hundreds of gigabytes in about a minute with chunked Dask arrays, and operational checks for metadata, missing-data tolerances, and localized attributes.
AI-written summary · about
chapters (14)
- 0:00Xclim introduction
- 1:27Climate change urgency
- 3:11Climate data platforms benefits
- 4:57Xclim for reproducible science
- 7:44Xclim features and foundations
- 10:12Indicator modules explained
- 12:59Bias adjustment in models
- 13:47Live demo: indicator calculations
- 16:56Demo: handling missing data
- 17:59Demo: metadata standards
- 18:44Demo: climate ensembles
- 21:38Xclim users and impact
- 22:23Xclim future plans
- 23:27Q&A
related talks
38mClimate Data, Scientific Data Visualization and Clojure
35mClimate change analyzing remote sensing data with Python
44mA Beginner's Guide to Weather & Climate Data
40mData-centric exploration using intake, dask, hvplot, and more
01h 18mUsing NASA EarthData Cloud & Python to Model Climate Risks
31mUnderstanding Climate Change with Python: From the IPCC Report to Climate Action
30mOn the Fly and On the Flight: Scientific Data Analysis Beyond the Beamline
39mItaly-Finland climate round trip with Python