Flood Forecasting System For Sanjiangkou Area Of Liao River Basin

This paper is focused on forecasting the peak flow and flood hydrograph for each control profile of Sanjiangkou region, Liaohe river. A brief introduction to the development of flood forecasting technology at home and abroad is presented at first. Strategy coupling the deterministic hydrological model and statistical-stochastic forecasting method is determined to adopt in building the forecasting schemes, which is based on the aborative investigation and analysis of characteristics of the studied catchment, including its natural geography, rainfall events, flood events, and the run of water conservancy projects etc. Simultaneously, the practical requirements on flood prevention for the studied basin are concerned in selecting various forecasting models.Wangben station is a controlling gauge located in the downstream of east of Liaohe river (ELH), the length of the river, above of which, is 279km and the basin area is about 10418 km2. The watershed hydrological model is selected as the rainfall-runoff forecasting model to predict the flood hydrograph at each controlling cross-section within in ELH, considering that river system is well developed, the effect of water conservancy projects is very small and hydrological data are available in practice. Xin'anjiang model has been widely applied in most regions of China for the merits of clear concept and explicit physical meaning of each parameter. However, as the model utilizes a saturated runoff mechanism, the model is more suitable for humid areas. While the ELH catchment belongs to semi-humid and semi-arid region, thus runoff with an infiltration excess mechanism would occur in the formation of runoff. With consideration of this feature, the mode of infiltration excess is coupled with the Xin'anjiang model, namely Xin'anjing-Shanbei model, and then applied in ELH catchment. Forecasting models are calibrated by historical rainfall-runoff events and forecasting precisions are investigated, results indicate that the Xin'anjiang-Shanbei model could comparatively well applied in semi-humid and semi-arid regions with the existence of both saturated and infiltration excess runoff Zhengjiatun station is the outlet of the west of Liaohe river (WLH), above of which, the length of the river is 170km with a total area about 91368 km2. The characteristics of WLH catchment are as follows: the river system is not well developed, and the branch, Wulijiemuren river will route runoff downward only when a big flood event occurs; there are a number of brakes which control the flood routing along the river; the condition of ground surface is very complex, including grasslands, swamps and deserts; Hydrological gauges within the catchment are scarcely and short of rainfall-runoff data. Therefore, a linear dynamic system model isutilized to forecast the flood hydrographs for main profiles, and a statistically based approach namely corresponding discharge method is calibrated to offer the prediction for flood peak flow at the same time.New computer software technologies are adopted to actualize the software system for flood forecasting of Sanjiangkou region, including GIS, RDB, VB and other computer tools. The software provides a convenient and practical tool for the study basin for purpose of flood prevention.