Generative climate downscaling enables high-resolution compound risk assessment by preserving multivariate dependencies

TL;DR

This paper introduces a diffusion-based multivariate generative framework for climate downscaling that preserves inter-variable dependencies, significantly enhancing the accuracy of high-resolution compound risk assessments.

Contribution

It presents a novel multivariate generative approach combined with bias correction that outperforms existing methods in preserving variable dependencies during climate downscaling.

Findings

Reduces inter-variable correlation errors by over fourfold.

Improves univariate and spatial accuracy in climate projections.

Enhances detection of severe drought conditions.

Abstract

Physics-based climate projections using general circulation models are essential for assessing future risks, but their coarse resolution limits regional decision-making. Statistical downscaling can efficiently add detail, yet many methods treat variables independently, degrading inter-variable relationships that govern compound hazards such as heat stress, drought, and wildfire. Here we show that a diffusion-based multivariate generative framework, combined with bias correction, recovers degraded inter-variable correlations even under a 50$\times$ increase in linear resolution. When applied to five meteorological variables over Japan, the framework reduces inter-variable correlation errors by more than fourfold relative to existing baselines while improving both univariate and spatial accuracy, leading to more accurate detection of severe drought. These results demonstrate that multivariate generative downscaling improves the reliability of compound risk assessment under large resolution gaps.