Research summary
Flooding is Canada’s most frequent and costly natural hazard—and the climate crisis is expected to increase damages from flooding tenfold by the end of the century. Accurate flood forecasting is essential for prevention and disaster response, but current forecasting methods are limited. As Canada Research Chair in Hydrological Ensemble Forecasting, Dr. Marie-Amélie Boucher is advancing flood prediction by improving ensemble forecasting systems, which have been shown to provide better results than traditional models.
Boucher and her research team are enhancing predictability across time and space, integrating machine learning to refine hydrological models, and leveraging citizen science to improve forecasting in data-scarce areas. By developing innovative techniques—including the use of artificial intelligence to interpret water level photos—their work will provide more reliable tools to support flood preparedness. These advances will help protect Canadian communities, reduce infrastructure damage, and strengthen resilience against increasingly severe weather events.