Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the current state and meaning of Earth System Modeling in this comprehensive 2-hour lecture by Raghu Murtugudde from the International Centre for Theoretical Sciences. Delve into how Earth System Models (ESMs) serve as crucial tools for understanding Earth's future climate under various anthropogenic emission scenarios and socio-economic policies. Learn about the simplified modeling approaches used to represent various Earth system components including climate, hydrology, ecology, biogeochemistry, and human activities at global scales with applications to local and regional dynamics. Understand the essential role of ESMs in simulating carbon cycles and environmental conditions that drive climate-ecosystem variability and change through physics-based differential equations. Examine the computational challenges and parameterization limitations of traditional ESMs, particularly for localized processes that can lead to simulation inaccuracies. Discover how recent advances in Machine Learning, including Physics-Inspired Neural Networks (PINN) and Generative Modeling, are being integrated to develop low-cost surrogate models and improve parameterization accuracy. Gain insights into the revolutionary impact of data-driven weather forecasting models powered by deep learning, including Transformer-based systems like FourCastNet, DLWP, and Pangu, as well as Graph Neural Networks like GraphCast that are now being adopted operationally by meteorological agencies. This lecture forms part of the Advanced Machine Learning for Earth System Modeling program, providing foundational knowledge for understanding how artificial intelligence is transforming climate science and Earth system research.
