Exploiting Machine Learning to Enhance Earth System Modeling and Analysis Across Scales

Machine Learning (ML) in Climate Science represents a promising and growing path of research that could contribute to substantially enhancing our understanding of the Earth system and reducing uncertainty in climate projections. ML has already started to make a tremendous contribution to Earth system science and is rapidly evolving to include new methodologies and applications. Climate modeling and analysis are facing new demands to enhance projections and climate information. This GRC conference will explore how to best push the frontiers of ML beyond state-of-the-art approaches, in particular in the fields of (1) developing hybrid (physics+ML) Earth system models with greater fidelity, (2) providing capabilities for climate extremes through large ensembles with emulators as well as enhancing detection and attribution methods, and (3) advancing climate model analysis and benchmarking. Utilizing this potential requires addressing (4) key ML challenges, in particular generalization, uncertainty quantification, explainable artificial intelligence, and causality that will be discussed as well. This interdisciplinary conference brings together ML and climate scientists, as well as the private sector, to accelerate progress towards actionable climate science.