Emulating Cloud Droplets in a Climate Model with Sam Silva

Abstract: Clouds exert an enormous influence on the Earth’s climate, playing a major role in the global water cycle and controlling the amount of radiation that both reaches the surface of the earth and is lost back to space. The physical processes that control how clouds interact with the climate system (e.g., cloud brightness, cloud lifetime, etc.) fundamentally occur at the scale of an individual cloud droplet (i.e., microns, 10-6 meters) within seconds or less, while cloud systems can span 1,000s of kilometers (106 meters) and exist for days (105 seconds). Reconciling these massive differences in scale is a current grand challenge in the cloud and climate research community. As it currently stands, clouds are the largest uncertainty in our modern understanding of the physical climate system, limiting our ability to predict future climate states and assess climate solutions.In this seminar I will present results from recent work using A.I. techniques to better constrain modern understanding of the influence of clouds on the Earth System. Specifically, I will discuss recent results on the application of emulation methods to develop improved representations of aerosol-cloud interactions in climate models, and describe efforts toward online integration of deep neural network cloud predictions into a state of the art climate model.


Bio: Sam Silva is currently an assistant professor in the Department of Earth Sciences and the Department of Civil and Environmental Engineering at the University of Southern California. Prior to his faculty position, he worked as a data scientist focused on machine learning and climate science the Pacific Northwest National Laboratory, a U.S. Department of Energy research laboratory. He received a Ph.D. in Environmental Engineering and Computation from the Massachusetts Institute of Technology, and an M.S. in Atmospheric Science and B.S. in Physics from the University of Arizona. His research is focused on air pollution and climate change, with particular interest in the convergence of traditional computational methods with modern data science and artificial intelligence techniques.

Paper: https://gmd.copernicus.org/articles/1...