Regional Climate Model Emulation with Diffusion Approaches: What is the Added Value of Generative Machine Learning?

2026-06-12

AI Summary

This study addresses the challenge of efficiently emulating high-resolution precipitation fields generated by regional climate models (RCMs) while adequately quantifying their uncertainties, particularly for extreme events. To this end, we propose ParamDiffusion, a two-stage diffusion-based generative model driven by large-scale variables from global climate models, offering a computationally efficient alternative to costly RCM simulations. We introduce an innovative, climate-science-oriented validation framework to systematically evaluate, for the first time, the added value of diffusion models over deterministic approaches in precipitation emulation. Results demonstrate that ParamDiffusion accurately reproduces key statistical characteristics of precipitation—including tail behavior and spatially compounding extremes—and generates physically plausible, high-detail precipitation fields. However, its ability to fully capture uncertainty in the most extreme events remains an area for improvement.

Abstract

Emulators provide a cost-effective alternative to regional climate models (RCMs) by capturing their dynamical downscaling function. They link large-scale predictors simulated by global climate models (GCMs) to RCM-simulated high-resolution fields of the target variable, here precipitation. Machine learning methods, typically deep learning, are cheaper than running RCMs in computation time and energy. Among them, generative models are appealing because they can simulate ensembles of local high-resolution fields consistent with the predictors. This ensemble, which we call the uncertainty envelope, remains to be properly assessed for added value. Here, we make three contributions. First, we introduce ParamDiffusion, a new two-stage diffusion-based framework, and compare it with a state-of-the-art diffusion approach. Second, we expand standard validation through a comprehensive framework aligned with climate-science needs, examining specific precipitation events, including extremes. Third, within this framework, we assess the added value of diffusion approaches relative to deterministic methods. We intercompare four deep-learning models: a deterministic model designed to capture the precipitation tail; a parametric probabilistic model based on it; a recently proposed diffusion approach; and ParamDiffusion, which couples the parametric model with a diffusion model. Our results show that diffusion-based approaches reproduce climatological precipitation statistics with high skill, including distributional tails and spatially compounded extremes, while generating spatially detailed fields. However, none of the assessed models consistently accounts for the most extreme RCM-simulated events within its uncertainty envelope. Diffusion models are therefore promising for probabilistic RCM emulation, but progress is still required before they can reliably represent high-impact precipitation extremes.

Problem

• regional climate model emulation
• diffusion models
• precipitation extremes
• generative machine learning
• uncertainty envelope

Innovation

• diffusion models
• regional climate model emulation
• generative machine learning
• extreme precipitation
• uncertainty quantification