AI-empowered Next-generation Multiscale Climate Modeling for Mitigation and Adaptation

Join us for the fifth session of the IIASA-wide seminar series, AI for Climate Science!

We are pleased to announce the next talk in the series, featuring Veronika Eyring from DLR, Germany and University of Bremen, Germany.

Veronika Eyring is Head of the Earth System Model Evaluation and Analysis Department at the German Aerospace Center (DLR) Institute of Atmospheric Physics and Professor of Climate Modelling at the University of Bremen. She maintains a strong collaboration with the National Center for Atmospheric Research (NCAR, USA) as Affiliate Scientist, with the DLR Climate Informatics Group  that she founded in 2017, and with the team of the European Research Council (ERC) Synergy Grant on “Understanding and Modelling the Earth System with Machine Learning (USMILE)”. Eyring’s research focuses on Earth system modeling and process-oriented model evaluation and analysis, including artificial intelligence (AI) techniques, to better understand the Earth system and climate change, and to improve the models. She has authored many peer-reviewed journal articles and has contributed to the Intergovernmental Panel on Climate Change (IPCC) climate and World Meteorological Organization (WMO) ozone assessments since 2004. Veronika is involved in the World Climate Research Programme (WCRP) since many years, for example through her role as Chair of the Coupled Model Intercomparison Project (CMIP) Panel (2014-2020). She was PI of the Earth System Model Evaluation Tool (ESMValTool) until 2020 and is a member of the European Lab for Learning & Intelligent Systems (ELLIS) since 2019. Veronika received the Gottfried Wilhelm Leibniz Prize in 2021 for her significant contributions to improving the understanding and accuracy of climate projections through process-oriented modeling and model evaluation.

For online participation, a registration is necessary.

Please register here.

Please note that the seminar is going to be recorded.

Title:
AI-empowered Next-generation Multiscale Climate Modeling for Mitigation and Adaptation

Abstract:
Earth System Models (ESMs) are fundamental to understanding and projecting climate change. They have continued to improve, but systematic errors and large uncertainties in their projections remain. A large contribution to this uncertainty stems from the representation of processes such as clouds and convection that occur at scales smaller than the resolved model grid. This impacts the models’ ability to accurately project global and regional climate change, climate variability, and extremes. High-resolution, cloud resolving models with horizontal grid resolution of a few kilometers alleviate many biases of coarse-resolution models, but their high computational costs limit their applicability to run multiple decades and large ensembles. Yet short simulations from high-resolution models can serve as information to develop machine learning (ML)-based parametrisations that are then incorporated into hybrid (physics+ML) ESMs that promise to have significantly reduced systematic errors and enhanced projection capability compared to current ESMs. We argue that an AI-empowered multiscale approach is needed, making use of km-scale climate models and hybrid ESMs that include essential Earth system processes and feedbacks yet are still fast enough to deliver large ensembles for better quantification of internal variability and extremes. The talk presents progress in hybrid modelling work with the ICOsahedral Non-hydrostatic (ICON) atmospheric model from the European Research Council (ERC) Synergy Grant on "Understanding and Modelling the Earth System with Machine Learning (USMILE)" as well as key challenges and visions how to enhance climate modeling with ML (Eyring et al., 2024a,b).