AGU Call for Abstracts: Machine Learning Across ESM: Subgrid-Scale Parameterizations, Emulation and Hybrid Modeling

[Simon Driscoll]

Dear colleagues,

We would like to advertise our 2025 AGU Meeting session, "Developments in Machine Learning Across Earth System Modeling: Subgrid-Scale Parameterizations, Emulation and Hybrid Modeling": https://agu.confex.com/agu/agu25/prelim.cgi/Session/251817. 

We are very excited to announce our invited speakers Ching-Yao Lai (Stanford), Sophie Abramian (Columbia) and Adam Subel (NYU). Your submissions are due by Wednesday, 30 July 23:59 EDT/03:59 UTC (and can be done by accessing the above link)

Confirmed talks include discussions on how Koopman variational autoencoders can help uncover underlying dynamical structure of atmospheric processes, assessing the physical behaviour of ocean emulators and more. Topics include (but are not limited to):

– Subgrid-scale parameterization via machine learning

– Emulators of physical processes, model components, or whole models

– Hybrid ML-physics modeling frameworks

– Physics-informed neural networks, neural operators, and differentiable programming

– Reinforcement learning

– Cross-domain applications (atmosphere, ocean, cryosphere, land).

Check out the link above for all details and we are excited to see you there! 

Best wishes,

Simon Driscoll, Sara Shamekh, Akshay Subramaniam, Aakash Sane, Karan Jakhar