Intelligent Earth Seminar: Ilan Price (DeepMind)

Probabilistic weather forecasting with machine learning

Recent advances in machine learning (ML)-based weather prediction (MLWP) have produced ML-based models which exhibit less forecast error than single NWP simulations. However, these advances have focused primarily on single, deterministic forecasts which fail to represent uncertainty and estimate risk, and overall these MLWP models have remained less accurate and reliable than state-of-the-art operational NWP ensemble forecasts. This talk will present our recent work on GenCast, the first machine learning based probabilistic weather model to exhibit greater skill than the top operational medium-range weather forecasts.