On Global Applicability and Location Transferability of Generative Deep Learning Models for Precipitation Downscaling

TL;DR

This paper investigates the ability of generative deep learning models for precipitation downscaling to generalize across different geographic regions using a global dataset and hierarchical data splitting.

Contribution

It introduces a global evaluation framework for assessing the transferability of generative downscaling models across diverse regions.

Findings

Models show varying performance across regions.

Global models can generalize reasonably well.

Region-specific tuning improves accuracy.

Abstract

Deep learning offers promising capabilities for the statistical downscaling of climate and weather forecasts, with generative approaches showing particular success in capturing fine-scale precipitation patterns. However, most existing models are region-specific, and their ability to generalize to unseen geographic areas remains largely unexplored. In this study, we evaluate the generalization performance of generative downscaling models across diverse regions. Using a global framework, we employ ERA5 reanalysis data as predictors and IMERG precipitation estimates at $0.1^\circ$ resolution as targets. A hierarchical location-based data split enables a systematic assessment of model performance across 15 regions around the world.