Demystifying Data-Driven Probabilistic Medium-Range Weather Forecasting

TL;DR

This paper shows that effective probabilistic medium-range weather forecasting can be achieved with a simple, scalable framework that does not rely on complex architectures or specialized training, outperforming existing models.

Contribution

The authors introduce a versatile, scalable framework for probabilistic weather forecasting that works across various estimators without complex design, improving accuracy over existing models.

Findings

Achieves significant improvements on most forecast variables.

Robust across different probabilistic estimators like diffusion models and CRPS.

Eliminates the need for tailored training strategies.

Abstract

The recent revolution in data-driven methods for weather forecasting has lead to a fragmented landscape of complex, bespoke architectures and training strategies, obscuring the fundamental drivers of forecast accuracy. Here, we demonstrate that state-of-the-art probabilistic skill requires neither intricate architectural constraints nor specialized training heuristics. We introduce a scalable framework for learning multi-scale atmospheric dynamics by combining a directly downsampled latent space with a history-conditioned local projector that resolves high-resolution physics. We find that our framework design is robust to the choice of probabilistic estimator, seamlessly supporting stochastic interpolants, diffusion models, and CRPS-based ensemble training. Validated against the Integrated Forecasting System and the deep learning probabilistic model GenCast, our framework achieves statistically significant improvements on most of the variables. These results suggest scaling a general-purpose model is sufficient for state-of-the-art medium-range prediction, eliminating the need for tailored training recipes and proving effective across the full spectrum of probabilistic frameworks.