Jumeau numérique probabiliste pour la prévision, l’alerte et la prévention des risques naturels à déclenchement rapide dans les petits bassins versants de montagne

Description

Rapid-onset natural phenomena (flash floods, sudden rises in water levels, torrential events) pose a major challenge for mountainous regions. Their dynamics are characterized by significant uncertainties, very short response times, and limited data availability, making it difficult to apply traditional deterministic approaches in an operational context.

This thesis is part of the European Interreg POCTEFA RISKRAPID project dedicated to the development of forecasting, warning, and prevention systems for rapidly occurring natural hazards in small mountain watersheds. The project brings together French and Spanish partners to develop new approaches to observation, modeling, and decision support for natural hazard management.

The scientific objective of the thesis is to develop a probabilistic digital twin capable of representing, predicting, and explaining the evolution of rapid hydrological phenomena based on a distributed...