Just weather scoring improves fast and accurate radar precipitation forecasts

Just Weather Scoring: Efficient End-to-end Nowcasting with Distributional Diffusion

Machine LearningComputer Vision and Pattern Recognition

Summary

Predicting short-term rainfall is important but challenging and slow with current AI methods. The authors introduce Just Weather Scoring, a new model that learns to predict future radar images of rain directly, without extra compression steps, making it faster and simpler. Their method also trains to match probability forecasts better and uses fewer steps to make predictions. Tests show it achieves top accuracy while running much faster and using fewer resources.

What this means in practice

  • For weather forecasting teams: Create faster and more accurate short-term radar-based rainfall predictions for operational weather services using an integrated end-to-end model.
  • For urban emergency planners: Implement rapid, probabilistic nowcasting systems to better prepare for localized heavy rain events and floods.

Authors

Jannik Wiese, Johannes Schusterbauer, Tommaso Martorella, Björn Ommer

Abstract

Generative diffusion models are well-suited for probabilistic precipitation nowcasting, but existing approaches often rely on separately trained compression or deterministic forecasting components and remain costly at inference due to iterative denoising. We introduce Just Weather Scoring (JWS), a single-stage, end-to-end diffusion model which addresses both issues by forecasting directly in radar space and enabling few-step generation. Radar-space modeling greatly simplifies training and inference and eliminates uncertainty arising from lossy compression. JWS combines Masked Asynchronous Diffusion, a timestep-sampling scheme that preserves clean context while adapting diffusion training to high-dimensional spatio-temporal data, with a simple scoring-rule objective that aligns training with probabilistic forecasting and unlocks few-step generation. On the SEVIR and MeteoNet benchmarks, JWS achieves state-of-the-art probabilistic forecasting performance at reduced training and inference cost. Even our smallest model remains competitive using substantially fewer parameters and more than 17x faster inference.