Author(s): Adele Young; Biswa Bhattacharya; Emma Daniels; Chris Zevenbergen
Linked Author(s): Biswa Bhattacharya
Keywords: Urban Flood Forecasting Early Warning Systems Bayesian Decision Framework Disaster Risk Reduction
Abstract: This study explores flood anticipatory actions in data-scarce urban settings using a Bayesian Decision Framework, focusing on Alexandria, Egypt. Flood forecasts are generated using a coupled ensemble Weather Research and Forecasting (WRF) and a MIKE urban inundation model. Actions are guided by probability density functions of flood depth and loss functions. The framework enables decisions to be updated 12–72 hours before events by selecting the actions that minimise expected losses. Results show that such probabilistic approaches can improve decision-making under uncertainty compared to ensemble means, but consideration is needed for a suitable loss function, which represents the decision maker's preference.
Year: 2025