Towards optimizing the operation of controlled flood detention basins

Floods are one of the most hazardous natural phenomena worldwide and they are predicted to increase both in intensity and frequency due to climate change. This necessitates comprehensive flood risk mitigation measures that are planned and controlled from a regional as well as a strategic trans-regional perspective.Controlled flood detention basins can be effective measures for dealing with extreme flood events [1]. By temporally storing water in the detention basin, the discharge in a river is reduced. If the water is removed from the river at the optimal time, this should reduce the peak water level at downstream locations and hence the flood risk.However, the identification of the optimal operation of flood detention basins is a non-trivial as well as non-deterministic problem. Flood forecast uncertainty, dilatation of the wave along the river channel and the uncertainty in the breaching process turns the polder operation into a stochastic optimization problem with multiple possible optimization targets. Hence, this optimization belongs to the class of sequential decision problems under uncertainty. In this contribution, we utilize a developed dynamic-probabilistic flood risk model [2] to analyze and optimize different control strategies as well as the effect of uncertainties on the optimality of the detention basin operation. We consider the case of a single detention basin as well as that of multiple detention basins that are arranged in series. [1] De Kork, J.-L.; Grossmann, M. (2009): Large-scale assessment of flood risk and the effects of mitigation measures along the Elbe River. Natural Hazards (2010) 52:143-166.[2] Ruf, M., Hoffmann, A., Straub, D. (2023): Application of a decision sensitivity measure for the cost-benefit analysis of a flood polder at the Bavarian Danube. 14th International Conference on Applications of Statistics and Probability in Civil Engineering (ICASP 14). Dublin, Ireland.