




























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.
| Subjects: | Atmospheric and Oceanic Physics (physics.ao-ph); Machine Learning (cs.LG); Applications (stat.AP) |
| Cite as: | arXiv:2605.11531 [physics.ao-ph] |
| (or arXiv:2605.11531v1 [physics.ao-ph] for this version) | |
| https://doi.org/10.48550/arXiv.2605.11531 arXiv-issued DOI via DataCite (pending registration) |
From: Takuro Kutsuna [view email]
[v1]
Tue, 12 May 2026 04:56:59 UTC (5,957 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理,仅供阅读参考。 原文来自 — 版权归原作者所有。