A Deep Dive into Google DeepMind | Re-released
from Weather Geeks
RECORDED JANUARY 22, 2025; Originally released FEBRUARY 12, 2025 Guests: Dr. Ilan Price, Senior Research Scientist, & Matt Willson, Research Engineer From the Euro to the GFS to the Canadian, there are multitudes of models that forecasters use to predict our daily weather. There are models for short-term severe forecasting, 10-day outlooks from your local news, and even models that predict our climate years into the future. As technology advances, so do all of these models and the technology we are focusing on today on Weather Geeks is AI. While it may seem like a buzzword these days, it can be used to enhance our industry and help us all reach our common goal: saving lives and property. We are thrilled to welcome Dr. Ilan Price to discuss GenCast, Google’s weather forecasting model that is entirely powered by AI. How does it stack up to the models we know and love? The answer may surprise you… Chapters 00:00 Introduction to AI in Weather Forecasting 02:10 Meet the Experts: Ilan Price and Matthew Wilson 06:34 Understanding GenCast: The AI Weather Model 10:47 Machine Learning vs Traditional Forecasting 13:22 Data Sources and Ethical Considerations 15:10 Handling Extreme Weather Events 21:15 Validation and Verification of GenCast 23:26 Impact of GenCast on Weather Forecasting
Embed this episode
What this episode covers
In this re-release episode of Weather Geeks, Dr. Marshall Shepherd discusses the advancements in weather forecasting through AI with experts Dr. Ilan Price and Matthew Willson from Google DeepMind. They delve into the GenCast model, which utilizes machine learning to provide probabilistic weather forecasts that outperform traditional models. The conversation covers the importance of data sources, the handling of extreme weather events, and the validation process of the model, highlighting its potential impact on future weather predictions.
No transcript yet.