When
Noon – 1 p.m., Today
Where
This talk from January 2024 presents an overview of the speed at which machine learning methods have become competitive, if not better, than traditional numerical weather forecasting systems. At the end, this talk then addresses the potential impacts of machine learning on the other foundation of weather prediction, data assimilation, which combines real, noisy observations with these weather models to improve forecasts. This presentation is a nice example of how machine learning works its way into existing mathematical and scientific disciplines, the speed at which it evolves, and what to look to in the future.