arXiv 2506.10772
Skillful joint probabilistic weather forecasting from marginals
By Ferran Alet, Ilan Price, et al.
Published 2025-06-12
Citation lineage
Review the prior work and downstream research connected to this paper.
Machine learning (ML)-based weather models have rapidly risen to prominence due to their greater accuracy and speed than traditional forecasts based on numerical weather prediction (NWP), recently outperforming traditional ensembles in global probabilistic weather forecasting. This paper presents FGN, a simple, scalable and flexible modeling approach which significantly outperforms the current state-of-the-art model…