Bayesian data fusion for the space-time forecasting of precipitations

Should you take your umbrella with you today? Rain forecasting is not always accurate as you may have noticed... but researchers are working on it! During his postdoctoral research, Dr. Dominique Fasbender developed an innovative statistical approach aimed at predicting precipitation at a local scale. The challenge of rainfall prediction is that multiple parameters need to be accounted for, including atmospheric conditions but also topography. Dr. Fasbender's approach allows the integration of all these parameters. In the future, it could be useful at several levels. It might help improve early warning systems* in the case of extreme rainfall events. His work might also be used to study how climate change may increase the risk of local extreme events happening. It would then help provide recommendations to private and public authorities in charge of land-use planning.