A Generative Machine Learning Approach for Improving Precipitation from Earth System Models

TL;DR

This paper introduces a generative machine learning method that enhances Earth system model outputs by improving spatial patterns, temporal dynamics, and reducing biases, surpassing traditional bias correction techniques.

Contribution

It presents a novel combination of unpaired domain translation and super-resolution models to improve ESM simulation accuracy and detail.

Findings

Realistic spatial patterns achieved

Temporal dynamics improved

Distributional biases reduced

Abstract

Quantifying the impacts of anthropogenic global warming requires accurate Earth system model (ESM) simulations. Statistical bias correction and downscaling can be applied to reduce errors and increase the resolution of ESMs. However, existing methods, such as quantile mapping, cannot effectively improve spatial patterns or temporal dynamics. We address this problem with a purely generative machine learning approach, combining unpaired domain translation with a super-resolution foundation model. Our results show realistic spatial patterns and temporal dynamics as well as reduced distributional biases in the processed ESM simulation.