Flood modeling and management using emerging Artificial Intelligence models

Contact: Ali Pourzangbar

Floods result in extensive damage, loss of life, and have far-reaching negative effects on public health and the economy in the world.

In July 2021, devastating floods impacted Europe, leading to over 220 deaths, with Germany facing significant destruction. The scale, frequency and impact of floods are not only increasing globally but also becoming complex due to the interplay of climate change and uncontrolled urbanization. Historical records, such as those of the flood peaks at the Ahr river overlooked by conventional models, illustrate the inadequacy of current flood models specially in predicting extreme flash floods.

Recently, Artificial Intelligence (AI) have achieved notable progress in water science. Accordingly, the main goal of this project is to create an AI tool that helps in mapping areas at risk of (flash)floods and landslides. It will also identify infrastructures at risk of damage from these disasters.

The developed model incorporates the effects of (1) large debris on flood dynamics and (2) evacuees’ behavior on the number of casualties. The configuration of the developed model will be fine-tuned in various locations to ensure its generalizability. It is expected that this AI toolbox will help reduce the number of people affected and the economic losses caused by floods.    

Publications:

Pourzangbar, A., Oberle, P., Kron, A., Franca, M.J. On the application of machine learning into flood modeling: data consideration and modeling algorithm. In Proceedings: SimHydro: New modelling paradigms for water issues? Chatou, France, 9-10 May 2023.