On January 13, 2026, the SCOR Foundation and the Geolearning Chair, which it co-finances, held a workshop at SCOR’s offices in Paris.

Running from 2022 to 2027, the Geolearning Chair aims to apply geostatistics, extreme event theory and Machine Learning to climate challenges. It is co-financed by the SCOR Foundation for Science alongside ANDRA, CCR and BNP-Paribas.

This workshop offered a diverse and vivid overview of the research projects pursued within the Chair, focusing on five in particular. You can find presentations of these below.

Grégoire Jacquemin, a third-year PhD student, showed how some high-impact compound climate events can be projected into future climates. Based on recent mathematical developments (see Jacquemin et al., 2026), he showed how the frequency of such events is expected to vary during the 21st century.

Post-doctoral fellow Antoine Heranval presented a novel approach to analyzing the dynamics of extreme climate events using the theory of point processes.

Rita Maatouk, a second-year PhD student, presented a mathematical framework for the analysis of extreme flows on river networks using causal graph networks.

Antoine Doizé, a third-year PhD student, presented a new stochastic model for rainfall occurrence, able to accurately account for long periods of drought.

Finally, Tiziano Fassina, a first-year year PhD student, presented his initial findings regarding the simulation of extreme rainfall events using score-based generative AI methods.

Learn more about the Geolearning Chair on the SCOR Foundation Project page and 
at chaire-geolearning.org.