Reservoir Flood Routing Considering the Non-Stationarity of Flood Series in North China

Abstract

Owing to the increasing anthropogenic activities, flood frequency analysis and flood control scheduling of reservoir become more and more complicated because of the non-stationarity of flood series. This paper aims to assess the effects that non-stationary flood series, which is affected by land use change of Daqing River Basin (in North China), has induced on the reservoir flood routing. And the analysis has been carried out in Xidayang Reservoir, which is located in south branch of Daqing River Basin of North China. In this paper, the annual maximum peak discharge series and the annual maximum flood volume (1-day, 3-day and 6-day) series of Xidayang Reservoir were selected to detect their variation using three steps change point diagnosis. Mixed distribution was employed to fit the non-stationary flood series, estimating parameters by simulated annealing algorithm. The results indicate that (1) mixed distribution provided a more appropriate and superior fit than conventional distribution (P3 distribution) to non-stationary flood series, (2) design flood values estimated by mixed distribution reduced by 0.03 %–20.24 % with different return periods compared with the results computed by conventional P3 distribution, and (3) through reservoir flood routing, the maximum water level and maximum discharge with various return periods considering the series’ non-stationarity also showed a decrease by up to 0.56 m and 81.68 % respectively compared with the results regardless of the series’ non-stationarity. These results provide a significant theoretical basis for flood control and water resources management in Daqing River Basin.