Background

These articles explain the concepts, theory, and design decisions behind scikit-downscale and statistical downscaling methods.

Note

Background articles are explanatory and conceptual. They provide context and deeper understanding but are not task-oriented.

Coming Soon

We are actively developing background articles. Future topics will include:

Statistical Downscaling Concepts

  • What is statistical downscaling?

  • Bias correction vs. spatial disaggregation

  • When to use statistical vs. dynamical downscaling

  • Understanding quantile mapping

  • Stationarity assumptions in downscaling

Method Comparisons

  • Comparing downscaling methods: strengths and weaknesses

  • Performance metrics for evaluating downscaling

  • Handling extremes in downscaling

  • Temporal and spatial consistency

Design and Architecture

  • Why scikit-learn’s API?

  • Pointwise vs. spatial downscaling approaches

  • Integration with Xarray and Dask

  • Extensibility and custom methods

External Resources

For broader context on statistical downscaling: