EG26: ML across Earth System Modelling

[Simon Driscoll]

Dear all,

Happy New Year! We are delighted to announce that abstract submission is open for our EGU 2026 session: “Developments in Machine Learning Across Earth System Modeling: Subgrid-Scale Parameterizations, Emulation and Hybrid Modeling”.

Confirmed Invited Speaker: Prof Richard Turner/Dr Anna Allen (Cambridge/Turing), including the state-of-the-art developments around end-to-end AI forecasting

The deadline is 15 January 2026, 13:00 CET, and you can submit your abstract here: https://www.egu26.eu/session/55729

Please share widely among colleagues and friends. We hope it is exciting to all researchers interested in AI and the environment. All are welcome and we look forward to seeing you there!

Session Conveners: Simon Driscoll (Cambridge), Sebastian Schemm (Cambridge), Tom Beucler (Lausanne), Pritthijit Nath (Cambridge)

Session Description: Machine learning is rapidly transforming how we model the Earth system. We invite contributions on all aspects of ML-driven parameterization, emulation, and hybrid modelling, including (but not limited to):

- Subgrid-scale parameterization via machine learning
- Emulators of physical processes, model components, or whole weather and climate models (including end-to-end learning)
- Hybrid ML-physics modelling frameworks
- Foundation Models
- Reinforcement learning
- Physics-informed neural networks, neural operators, and differentiable programming
- Verification of data-driven models
- Physical behaviour, encoding and analysis of AI parameterizations, emulators and whole models
- Calibration and parameter optimization using ML
- Coupling of ML models with physical models
- Cross-domain applications (atmosphere, ocean, cryosphere, land).

This session provides a critical overview of current progress and emerging directions in the application of ML across all these domains.

Location: Vienna, Austria (3–8 May 2026) & Online