Accelerating Discovery in Climate Physics with Machine Learning

The oceans are a crucial component of the climate system. They store and redistribute most of the excess heat from anthropogenic emissions. Climate simulations have been essential for understanding and predicting this ocean warming. However, uncertainty remains regarding the causes and pace of future ocean warming due to inadequate representations of unresolved processes, such as clouds and ocean turbulence, in global climate models.

In this Presidential Lecture, Laure Zanna will discuss the potential for machine learning to accelerate the discovery of physics principles and governing equations for multiscale climate processes such as ocean turbulence and how these discoveries can help improve the simulations of ocean heat transport and climate models.

Zanna is a climate physicist and professor at NYU’s Courant Institute of Mathematical Sciences. She received the 2020 Fofonoff Award from the American Meteorological Society “for exceptional creativity in the development and application of new concepts in ocean and climate dynamics.” She is the lead principal investigator of M²LInES, an international collaboration sponsored by Schmidt Futures dedicated to improving climate models using scientific machine learning. She is also the geoscience director for the National Science Foundation–funded Science and Technology Center LEAP (Learning the Earth with AI and Physics). She is also on the advisory boards of KITP, CESM and several NSF-funded Institutes.