Effectiveness of flood retention basins under hydrologic uncertainty

Abstract

Flood retention basins are widely applied to improve flood safety. The effectiveness of such basins is more sensitive for (hydrologic) uncertainties compared to other flood protection measures. This paper shows how uncertainty can be included in the effectiveness of two major types of flood retention: with fixed sills and adjustable gates. Fixed sills divert flood waters optimally only in a small range of flood levels. If the sill is optimized for a design flood, it has limited impact on events with a different peak discharge or volume. Therefore, the flood level reduction is often overestimated using design floods. Adjustable gates operated based on flood forecasts offer the possibility to divert water under a wider range of conditions and therefore these are potentially more effective and robust, depending on the forecast uncertainty. The probabilistic analysis for fixed sills includes flood duration statistics from a stochastic weather generator. For adjustable gates, we developed a decision support algorithm for the optimal moment to open the gates based on an ensemble forecast, and quantified the expected flood reduction. A case study along the Rhine river shows that an approach without hydrologic uncertainty would overestimate the effectiveness with a factor 2 to 3.