Stochastic coastal flood risk modelling in the east coast of Africa

Extreme coastal flood events can have devastating impacts in densely populated and low-lying coastal areas, affecting societies, economies, and the environment. Flood risk assessments play a key role in reducing the potential impacts of these events. At global scale, coastal flood risk assessments allow determining the prime price definition of (re-)insurance companies, establishing of climate adaptation and risk reduction measures and understanding flood hazard and risk in data-scarce regions.

Flood risk assessments at large to global scales, however, have generally been based on extreme sea levels estimated for specific return periods, combined with static flood modelling approaches. These traditional approaches are computationally efficient but at large scales they neglect the spatial patterns of flood events, leading to miss-estimation of the risk. Stochastic flood modelling approaches, instead, can become an alternative to capture the spatiotemporal dependency of events.

In this study we analyse the added value of a stochastic coastal flood modelling approach over a traditional return period-based approach for 1000 years of synthetic tropical cyclone events in the east coast of Africa. Synthetic tropical cyclone events from the Synthetic Tropical cyclOne geneRation Model (STORM) combined with the Global Tide and Surge Model (GTSM) will be used to simulate water level timeseries. The Super Fast INundation of CoastS (SFINCS) hydrodynamic flood model together with an impact model will be used to derive the flood risk.