CRPS-LAM: Regional ensemble weather forecasting from matching marginals
This work addresses the problem of slow sampling times in regional ensemble weather forecasting for meteorologists, offering a more efficient incremental improvement over existing methods.
The paper tackles the computational expense of diffusion-based ensemble weather forecasting models by introducing CRPS-LAM, a probabilistic regional forecasting model that generates ensemble members in a single forward pass, achieving sampling speeds up to 39 times faster while matching the low errors of diffusion models on the MEPS dataset.
Machine learning for weather prediction increasingly relies on ensemble methods to provide probabilistic forecasts. Diffusion-based models have shown strong performance in Limited-Area Modeling (LAM) but remain computationally expensive at sampling time. Building on the success of global weather forecasting models trained based on Continuous Ranked Probability Score (CRPS), we introduce CRPS-LAM, a probabilistic LAM forecasting model trained with a CRPS-based objective. By sampling and injecting a single latent noise vector into the model, CRPS-LAM generates ensemble members in a single forward pass, achieving sampling speeds up to 39 times faster than a diffusion-based model. We evaluate the model on the MEPS regional dataset, where CRPS-LAM matches the low errors of diffusion models. By retaining also fine-scale forecast details, the method stands out as an effective approach for probabilistic regional weather forecasting