Multiobjective Optimization for Improved Management of Flood Risk


Effective flood risk management requires consideration of a range of different mitigation measures. Depending on the location, these could include structural or nonstructural measures as well as maintenance regimes for existing levee systems. Risk analysis models are used to quantify the benefits, in terms of risk reduction, when introducing different measures; further investigation is required to identify the most appropriate solution to implement. Effective flood risk management decision making requires consideration of a range of performance criteria. Determining the better performing strategies, according to multiple criteria, can be a challenge. This article describes the development of a decision support system that couples a multiobjective optimization algorithm with a flood risk analysis model and an automated cost model. The system has the ability to generate potential mitigation measures that are implemented at different points in time. It then optimizes the performance of the mitigation measures against multiple criteria. The decision support system is applied to an area of the Thames Estuary and the results obtained demonstrate the benefits multiobjective optimization can bring to flood risk management.