EGU General Assembly 2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

The key role of causal discovery to improve data-driven parameterizations in climate models

Fernando Iglesias-Suarez1, Veronika Eyring1,2, Pierre Gentine3,4, Tom Beucler5, Michael Pritchard6,7, Jakob Runge8,9, and Breixo Solino-Fernandez1
Fernando Iglesias-Suarez et al.
  • 1Deutsches Zentrum für Luft- und Raumfahrt (DLR), Institut für Physik der Atmosphäre, Germering, Germany (
  • 2University of Bremen, Institute of Environmental Physics (IUP), Bremen, Germany
  • 3Department of Earth and Environmental Engineering, Center for Learning the Earth with Artificial intelligence and Physics (LEAP), Columbia University, New York, USA
  • 4Earth and Environmental Engineering, Earth and Environmental Sciences, Learning the Earth with Artificial intelligence and Physics (LEAP) Science and Technology Center, Columbia University, New York, USA
  • 5University of Lausanne, Institute of Earth Surface Dynamics, Lausanne, Switzerland
  • 6University of California, Department of Earth System Science, Irvine, USA
  • 7NVIDIA Corporation, Santa Clara, USA
  • 8Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR), Institute of Data Science, Jena, Germany
  • 9Technische Universität Berlin, Institute of Computer Engineering and Microelectronics, Berlin, Germany

Earth system models are fundamental to understanding and projecting climate change, although there are considerable biases and uncertainties in their projections. A large contribution to this uncertainty stems from differences in the representation of clouds and convection occurring at scales smaller than the resolved model grid. These long-standing deficiencies in cloud parameterizations have motivated developments of computationally costly global high-resolution cloud resolving models, that can explicitly resolve clouds and convection. Deep learning can learn such explicitly resolved processes from cloud resolving models. While unconstrained neural networks often learn non-physical relationships that can lead to instabilities in climate simulations, causally-informed deep learning can mitigate this problem by identifying direct physical drivers of subgrid-scale processes. Both unconstrained and causally-informed neural networks are developed using a superparameterized climate model in which deep convection is explicitly resolved, and are coupled to the climate model. Prognostic climate simulations with causally-informed neural network parameterization are stable, accurately represent mean climate and variability of the original climate model, and clearly outperform its non-causal counterpart. Combining causal discovery and deep learning is a promising approach to improve data-driven parameterizations (informed by causally-consistent physical fields) for both their design and trustworthiness.

How to cite: Iglesias-Suarez, F., Eyring, V., Gentine, P., Beucler, T., Pritchard, M., Runge, J., and Solino-Fernandez, B.: The key role of causal discovery to improve data-driven parameterizations in climate models, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-6450,, 2023.