Downscaling and Bias-correction of Climate Model Simulations

Total time needed: ~5 hours
This recipe gives an introduction to the ClimAlign method of downscaling/upsampling climate model output. This is a non-trivial method, from both, the ML and the climate modeling point of view.
Potential Use Cases
Bias-correction and downscaling of climate projection data to higher resolution, e.g. for impact modeling
Who is This For ?
INTERMEDIATEData scientitst interested in climate science; climate and climate change impact modelers interested in more advanced machine learning
Click on each of the following annotated items to see details.
ARTICLE 1. Computing the Climate
  • How do climate models work?
  • Why should we trust climate model predictions?
60 minutes
ARTICLE 2. Embeddings of Weather
  • Are embeddings useful for meteorology?
40 minutes
ARTICLE 3. ClimAlign
  • How can we downscale spatial structure?
180 minutes

Concepts Covered

0 comment