PhD Position in Spatio-Temporal Modelling of Extreme Wildfires

The University of Lausanne is offering a fully funded PhD position within the Institute of Earth Surface Dynamics, focusing on spatio-temporal modelling of extreme wildfires. Funded by the Swiss National Science Foundation under the ExtFire project, the research aims to integrate non-stationary extreme value theory, causal inference, and tail-sensitive deep learning. The successful candidate will work under the supervision of Dr. Marj Tonini and co-PI Prof. Valérie Chavez-Demoulin in collaboration with an international network of researchers. The role involves processing large environmental datasets, developing predictive models, and contributing to scientific publications and teaching activities.

• Harmonize and integrate large-scale open datasets covering wildfires, meteorology, and land surface characteristics into a unified pan-European infrastructure.
• Develop innovative statistical and spatio-temporal models to analyze extreme wildfire events characterized by rapid spread or massive burned areas.
• Construct baseline machine learning models and datasets to evaluate the predictability of extreme environmental events.
• Apply explainable artificial intelligence techniques, such as SHAP, to determine key environmental drivers influencing extreme fires across different regions.
• Author scientific research papers, disseminate findings at international conferences, and assist with departmental teaching duties for up to ten percent of the time.