Diffusion-Based Refinement for Kilometer-Scale Probabilistic Precipitation Nowcasting

arXiv:2608.30205 · cs.LG, physics.ao-ph · Submitted 2026-08-31 · Read on arXiv

cs.LG, physics.ao-ph

Submitted: 2026-08-31

Updated: 2026-08-31

Comments: Submitted to npj Climate and Atmospheric Science. Includes Supplementary Information

Code: https://github.com/tony890048/exPreCast

License: http://creativecommons.org/licenses/by/4.0/

The gist: Localized extreme precipitation is a major trigger of urban flash floods and landslides, yet producing nowcasts that combine fine spatial detail with probabilistic uncertainty remains challenging.

Terminology

Abstract

Localized extreme precipitation is a major trigger of urban flash floods and landslides, yet producing nowcasts that combine fine spatial detail with probabilistic uncertainty remains challenging. Here we introduce exPreCast-ENS, a conditional residual diffusion framework that transforms the deterministic 4 km radar nowcaster exPreCast into a 1 km probabilistic ensemble while correcting systematic forecast errors. Conditioning on both the forecast and preceding radar observations lets the ensemble-mean correct the baseline rather than perturb it, while members represent unresolved fine-scale variability. Over the Korean Peninsula, skill improves with ensemble size. In two high-impact events in 2023, a 30-member ensemble recovers 38-47% of heavy-rain pixels missed by exPreCast while retaining approximately 95% of its correct detections and alarming on under 1% of the pixels it correctly left clear. The method generates a 1-h forecast in 3.4 s on a single GPU and yields consistent improvements on the French regional MeteoNet radar dataset.

Related papers