RainAtlas: A Multi-Continental Dataset for Precipitation Downscaling

2026-09-30

AI Summary

This study addresses the scarcity of high-resolution precipitation observations and the limited cross-regional generalizability of machine learning models by integrating ERA5 reanalysis data with multi-source direct observations. We construct a standardized hourly precipitation dataset unified to a 2 km grid spanning three continents, providing approximately 210,000 low-to-high resolution paired samples. Leveraging this platform for precipitation downscaling modeling and multi-metric evaluation, we reveal significant performance fluctuations of existing state-of-the-art models over unseen geographical regions. This work establishes a novel paradigm for integrating cross-domain heterogeneous precipitation data, offering a reliable benchmarking platform for extreme rainfall prediction research.

Abstract

Extreme rainfall events are increasing in intensity and frequency as climate change accelerates. While kilometer-scale precipitation forecasts are critical for supporting local decision-making, the limited availability of high-resolution precipitation observations hinders their accuracy, especially in under-resourced regions. Machine learning models are widely used to downscale precipitation data to km-scale, but their application to unseen geographies presents challenges. First, processing raw high-resolution precipitation datasets across regions requires significant engineering and domain expertise. Second, generalization across regions remains difficult. To help overcome these barriers, we release RainAtlas, a large-scale, ML-ready and multi-continental dataset for precipitation downscaling. Covering three continents, RainAtlas harmonizes heterogeneous hourly km-scale observations to a common 2-km grid. Each regional partition contains around 210,000 aligned low- and high-resolution precipitation pairs, respectively from ERA5 reanalysis and direct observations. We benchmark state-of-the-art ML-based downscaling models across RainAtlas using a wide range of metrics. Our evaluation reveals substantial variance in out-of-domain generalization depending on the training regions. This underscores the need for cross-regional, multi-source km-scale evaluation, establishing RainAtlas as a well-positioned benchmark for precipitation downscaling research.

Problem

• precipitation downscaling
• high-resolution precipitation
• cross-regional generalization
• extreme rainfall
• machine learning

Innovation

• Precipitation Downscaling
• Multi-continental Dataset
• Machine Learning Benchmark
• Out-of-domain Generalization
• High-resolution Observations