Study On Flood Forecasting In Taihang Piedmont Basin Based On Atmospheric-Hydrological Coupling Model

Whether flood forecasting is effective or not depends on two important factors:forecasting accuracy and forecast period.If higher precision flood forecast and longer forecast period are obtained,the value of flood forecast will be greater.At present,in order to extend the forecast period of flood forecasting and ensure its forecasting accuracy,the effective way is to establish the coupled atmospheric-hydrological model.WRF(Weather Research & Forecasting Model)is a new generation of mesoscale numerical weather prediction model.It is widely used because of its convenient design and mature physical process parameters of atmospheric system.Taihang Pedimont basin,backed by Taihang Mountains,is located at the junction of Loess Plateau and North China Plain.The terrain fluctuates greatly,and flood disasters occur from time to time.Therefore,taking the Taihang basin as the research area,this paper uses the measured rainfall and flood data of the basin above Duanzhuang station to carry out the flood prediction research based on WRF numerical weather prediction and HEC-HMS(Hydrologic Modeling System)model.The rainfall results of climate model are used to input into the hydrological model for one-way coupling,and the coupling model is analyzed to establish the realtime flood forecasting scheme of the Taihang basin.The main research contents and results of this paper are as follows:(1)Firstly,the necessity of optimizing the physical parameter scheme in numerical weather model is analyzed.Then,an index evaluation system based on RE、POD、CSI、FAR、RMSE is established to test the capability of precipitation forecast of a climate model.Based on grey correlation degree and TOPSIS(Technique for Order Preference by Similarity to an Ideal Solution)method,the multi-attribute decision model is established to verify the results and construct a more scientific and reasonable index system.(2)Taking the Taihang Piedmont basin as the research area,the numerical weather prediction model was established to simulate typical rainfall events in different magnitudes.From the perspectives of accumulated area mean rainfall and spatio-temporal point rainfall,the established index system and multi-attribute decision-making model are applied to select the best combination of parameterized schemes for WRF model.The results show that physical parameters combination MYJ-BMJ-WSM6 and MYJ-GD-Lin are good for simulation in the study area,but they have different advantages in time and space simulation as well as in different rainfall magnitudes.(3)Based on HEC-HMS model,a flood forecasting model of Taihang Piedmont basin is established to realize one-way coupling with numerical weather forecasting.By converting the rainfall results of WRF model in time and space to drive the hydrological model,a coupled atmospheric-hydrological flood prediction model is constructed.The flood rolling forecast is carried out,considering the actual flood occur time and model calculation time.The results show that coupling model can prolong the forecast period and has good forecasting accuracy.Through the further test of rainfall prediction results,it can be seen that the simulation effect is mainly affected by the prediction error of climate model on rainfall magnitude and rainfall falling area.At the same time,the reservoir storage in the basin also further expands the influence of climate model error on the coupling model.