How AI is taking weather forecasting and climate modeling by storm

Computer-generated weather forecasts have become steadily more accurate over the past 50+ years, thanks to better satellite observations and powerful computers allowing more faithful, detailed representation of the complex physics of the earth system.  Similar computer models are used for understanding past climates and predicting future climate change, a defining issue of the 21st century.  In the past two years, AI has surpassed the skill of the best global weather forecast models.  Can AI help us more skillfully model climate, even though the future will not be like the past - if so, how will it learn to predict the unseen?  Can it help us plan for coming new extremes of heat, flood, drought, and rising sea levels?  You’ll see promising early steps toward that vision. Indeed, within a few years, AI may become a backbone of weather and climate modeling, saving time and money and making reliable, customized, local climate change information much more broadly accessible to the interested public.  





About Speaker Dr. Chris Bretherton

Chris Bretherton is an atmospheric scientist who studies cloud formation and turbulence and improves how they are simulated in global climate and weather forecast models. Since Sept. 2021, he has led a philanthropically-supported climate model development group at AI2 in Seattle, in collaboration with NOAA GFDL, to use machine learning trained on global cloud-resolving model output to improve the simulation of regional precipitation trends and extremes in climate models. This initiative was started in 2019 at Vulcan Inc.

He is Emeritus Professor of Atmospheric Sciences and Applied Mathematics at the University of Washington. His UW research group has helped lead field experiments and observational analyses and pioneered new frontiers in three-dimensional modeling of fluid flow in and around fields of clouds, including understanding how clouds will respond to and feed back on climate change. Computer code developed by his research group for simulating cloud formation by atmospheric turbulence has been used in the two leading US climate models.

He was a lead author of the Intergovernmental Panel on Climate Change Fifth Assessment Report in 2013, Chair of a 2012 National Academy report entitled A National Strategy for Advancing Climate Modeling, and a former director of the University of Washington Program on Climate Change. In 2012, he received the Jule G. Charney Award, one of the two highest career awards of the American Meteorological Society, and he was the 2019 AMS Haurwitz Lecturer. He is a Fellow of the AMS and AGU, and a member of the National Academy of Sciences and Washington State Academy of Sciences.