Study On Flood Forecasting Method Of Chaersen Reservoir

Flood disaster is one of the world’s worst natural disasters, poses a serious threat for people’s lives and property. Flood forecasting is an important non-engineering measure of basin flood protection. Chaersen reservoir is a flood protection and hennessy water conservation project midstream and upstream of Taoer River basin. No effective flood forecasting scheme is applied in Chaersen reservoir basin, because of extremely nonuniform land surface condition and rainfall patterns, complicated temperature and weather, flood forecasting of Chaersen reservoir is complex and difficult. So, Studying on flood forecasting of Chaersen reservoir is extremely important. This paper takes Chaersen reservoir basin as the study to make the real-time flood forecasting method, then the system of lots of flood forecasting models is proposed. The main contents and conclusions are as follows:(1) The XAJ model, the DHF model and the API model are chosen for Chaersen reservoir based on the analysis of hydrological characteristics and land surface condition in Chaersen reservoir basin. The parameters of model are optimized with genetic algorithm and manual calibration.(2) Chaersen reservoir basin is used for studying on applicability of hydrological models as an unit. Simulation accuracy of flood forecasting model is evaluated from runoff, peak flow, peak time and uncertainty factor.(3) Chaersen reservoir basin is divided into sub-units, each of which is under control of each rainfall station, for flood forecasting considering the propagation time of flood in river. In order to improves the accuracy of simulation, Flood forecasting of Suolun basin is researched.But, flood forecasting of Suolun basin is ineffective, like the whole basin. The reason is that its rainfall stations are unrepresentative. In order to avoid this problem, study on flood forecasting of Chaersen reservoir is based on actual flow process of Suolun hydrological station.(4) Based on Browser/Server frame, this paper develops and designs the real-time flood forecasting system with Java, JavaBean, HTML, JSP language and MySQL database. The system could simulate historical flood efficiently, which also can be added to other flood forecasting model in the future.