Author(s): F. Jordan; J.-L. Boillat; J. Dubois; A. J. Schleiss
Linked Author(s): Anton J. Schleiss
Keywords: No Keywords
Abstract: During the past 20 years, the region Wallis has been hit by three major flood events. The Rhone River and some of its tributaries upstream of Lake Geneva flooded the plain and caused severe damages leading to human deaths. In order to reduce the risk of such natural disasters, a flood forecasting and management system is developed, which allows during flood events identifying the optimal operation of numerous hydropower plants and dams located on the Rhone catchment area. A 72 hours lead time flood forecast based on quantitative precipitation, temperature forecast and real-time measurements is developed. Taking into account snow and glacier melt as well as soil discharge, the model will provide every 12 hours new flood forecasts at the outlets of the 239 sub-catchments covering a surface 2 of 5500 km. Computing the predicted inflows at the most important alpine reservoirs as well as the predicted hydrographs at different points in the hydraulic network, an optimization model computes and defines the most efficient strategies for minimizing the flood damages. This model takes into account the characteristics of the existing hydropower schemes and their reservoirs. It proposes preventive turbine and pump operation as well as water release through gates and allows the use of all operational water intakes during the flood peak. The preliminary results are promising and reveal a significant flood discharge reduction with preventing operations 30 hours in advance. The necessity of a reliable flood prediction model is also highlighted in order to minimize the cost of inappropriate preventive operations.