The Work
Bommer built a climate-cooling emulator that couples a hybrid neural model with a physics-based temperature correction to simulate stratospheric aerosol injection.
What Worked
The model reproduced Pinatubo's stratospheric warming and surface cooling from climatology alone, matching 50 hPa warming to within 0.16 K.
What Has Been Achieved
It reconstructs the 1991 Pinatubo eruption from climatology alone, matching observed stratospheric warming to within 0.16 K and getting the anomaly sign right across all 37 pressure levels, and runs a decade of daily scenario data in under eight hours on a single GPU.
What Was Learned
Retraining on high-resolution climate archives was blocked by how little open data exists at the needed resolution. Remains a community wide blocker.
About
Philine is completing a PhD in Explainable AI (XAI) for Climate Science from TU Berlin. She built QuantusXClimate, the first-ever tutorial on XAI evaluation, comparison, and selection for climate teams. She led a projects developing frontier XAI methods and building S2S weather forecasting models, submitting to top-tier ML conferences and interdisciplinary journals.
Lab & Advisors
.jpg)
Gabi Hegerl
University of Edinburgh
Professor of Climate System Science, School of GeoSciences, University of Edinburgh.
Professor Gabi Hegerl, CBE FRS FRSE, is a Professor of Climate System Science, School of GeoSciences, University of Edinburgh. Prof Hegerl's research focuses on the causes of climate change and the causes and consequences of extreme events. She has also been involved in IPCC reports on climate change and work with the world climate research programme on setting priorities for climate research. She is also a fellow of the Royal Society and Royal Society of Edinburgh.
