Flood protection along the main rivers in the Netherlands is based on design discharges with an exceedance probability of 1/1250. The estimation of the design discharges is currently based on the statistical analysis of measured peak discharges at Borgharen (Meuse) and Lobith (Rhine). However, this approach faces various problems. Because of changes in the upstream basin, the river geometry and climate, the discharge record is potentially non-homogeneous and it is uncertain how representative the measured records are of the present situation. Another problem is the choice of probability distributions. Finally, the current approach provides no information about the volume and duration of the considered flood event. Therefore, a new methodology has been proposed to provide a better physical basis for the estimation of the design discharge of the Dutch rivers. This methodology consists of a stochastic multivariate weather generator, which generates long simultaneous records of precipitation and temperature over the basin, and a hydrological/ hydraulic model to simulate discharge. This presentation summarises the first results of the new method for the Rhine and Meuse basins, and discusses its prospects and limitations.
R Leander, H Buiteveld, MJM de Wit, TA Buishand. Application of a weather generator to simulate extreme river discharges in the Rhine and Meuse basins