We would like to invite you to a KPA-IMfESS talk given by Prof. Philip Stier, Department of Physics, University Oxford, UK with the title:
Towards constraining global cloud-aerosol interactions combining km-scale climate models with machine learning
Date: Wednesday, November 19, 2025 at 10:15 a.m.
Location: University of Cologne, Institute for Geophysics and Meteorology (4th floor, lecture hall 4.001, Höninger Weg 100, Köln)
It will also be streamed online. For the LogIn details, please get in touch with us.
Philip Stier is Professor of Atmospheric Physics in the Department of Physics, where he leads the Climate Processes Group, and a Fellow of Reuben College. He also serves as Director of Intelligent Earth – Oxford\’s UKRI AI Centre for Doctoral Training in AI for the Environment and serves on the steering group of the Oxford Climate Research Network.
His research addresses physical climate processes in the context of anthropogenic perturbations to the Earth system as the underlying cause of climate change and air pollution. Focal points of his research are cloud and aerosol physics, their interactions and their role in the climate system. In his Climate Processes Group they combine complex numerical models with Earth observations and AI / Machine Learning to advance their theoretical understanding and predictability of the climate system.
Abstract:
Aerosol-cloud interactions have persistently remained the single greatest uncertainty in anthropogenic perturbations to the climate system. The associated radiative effects have traditionally been assessed by global general circulation models (GCMs), aiming to model the chain of microphysical processes from aerosols acting as cloud condensation via cloud microphysics to the global energy balance. However, this relies on a complete representation of a very complex process chain and has been shown to be subject to large and persistent uncertainties. Moreover, significant structural limitations remain, in particular related to the intrinsic sub-grid scale coupling of aerosols and clouds. The emergence of global km-scale models provides an opportunity to overcome some of these limitations and to explore previously unresolved aerosol phenomena explicitly, including aerosol-convection interactions. However, they remain insufficiently constrained by observations.
In this presentation, I will highlight the potential of global km-scale climate models to make progress on our understanding of cloud-aerosol interactions, with a particular focus on aerosol-convection interactions. Starting from regional km-scale simulations using ICON with idealised aerosol perturbations from the MACv2-SP plume model, we demonstrate the potential to extend this work to global kilometre-scale modelling. This work motivates our development of the reduced complexity aerosol model HAM-lite, suitable for long-term global km-scale modelling of aerosol-convection interactions using ICON. I will present novel approaches for the evaluation of km-scale models, including cloud tracking and machine learning. Their systematic application will provide the foundation for future observationally constrained assessments.