AI in Climate Science: From Emulation to New Discoveries

Simulating Earth’s climate is a tough task, involving representing many complex processes such as turbulence, convection and terrestrial processes such as phenology. Recently, AI has proven to be a powerful tool in overcoming these roadblocks.

In this Presidential Lecture, Pierre Gentine will present the role AI is playing in improving simulations of Earth’s climate. He’ll discuss how AI can be used to make new physical discoveries on complex multiscale and nonlinear physics by extracting new knowledge from ever-increasing datasets. These approaches, he says, are revolutionizing the way we analyze the wealth of observational data (e.g., satellite observations) and simulation data to advance our understanding of the complex climate system.

Gentine is the Maurice Ewing and J. Lamar Worzel Professor of Geophysics in Columbia University’s Department of Earth and Environmental Engineering and Professor of Earth and Environmental Sciences. He obtained an M.Sc. and Ph.D. from MIT. He is the director of the National Science Foundation’s Center for Learning the Earth with Artificial Intelligence and Physics (LEAP). His research investigates the continental hydrologic cycle using multiscale modeling and big data in the context of rising carbon dioxide concentrations.