IISER MOHALI

Abstract: Recent advances in artificial intelligence are rapidly transforming our ability to observe, model, and predict the Earth system. This talk will explore the development of AI-based multi-scale Earth system modelling approaches that connect global atmospheric dynamics with regional and local-scale processes to enable high-resolution prediction across a wide range of forecast horizons. The presentation will highlight AI-driven approaches for high-resolution short-range weather prediction, including downscaling and correction of coarse-resolution forecasts to better represent precipitation, temperature, land–atmosphere interactions, and extreme events. It will also discuss opportunities for extending AI-based prediction toward longer lead times, where learning large-scale circulation patterns, land and ocean memory, teleconnections, and other sources of predictability becomes increasingly important. Particular emphasis will be placed on combining AI, observations, remote sensing, and physics-based Earth system models rather than viewing AI and traditional modelling as competing approaches. Emerging techniques such as physics-informed learning, generative modelling, and multi-scale architectures provide new opportunities to bridge spatial and temporal scales while substantially reducing computational costs. Through examples spanning weather and climate applications, the talk will discuss both the potential and the remaining challenges of building AI-enabled prediction systems that are accurate, physically consistent, and useful for decision-making. Ultimately, these developments point toward a new generation of seamless Earth system prediction systems, connecting high-resolution early warning with longer-term preparedness for weather and climate risks.

About the speaker: Dr. Manmeet Singh is an Assistant Professor in the Department of Earth, Environmental, and Atmospheric Sciences at Western Kentucky University. His interdisciplinary research lies at the intersection of climate science, artificial intelligence, and Earth system modeling. Before joining WKU, Dr. Singh was a Distinguished Postdoctoral Fellow at The University of Texas at Austin, where he worked on AI-augmented weather and climate prediction systems to improve forecasts related to extreme weather, health impacts, and disasters. He also served as a Scientist at the Indian Institute of Tropical Meteorology (IITM), contributing to national climate modeling efforts and leading several projects on aerosol-monsoon dynamics and climate data downscaling. Dr. Singh holds a Ph.D. in Climate Studies from the Indian Institute of Technology Bombay.