| Mountain flood disaster is one of the most serious natural disasters in China,which hinders the economic and social development of our country.Hydrological models can predict flood process and peak discharge in time,but the accuracy of flood prediction depends on the reliability of model parameters.Therefore,accurate and reliable model parameters are the basis of flood prediction and an important link of flood flood prevention and disaster mitigation.This paper takes the upper Baihe river basin controlled by the Baitugang Hydrological Station in Nanzhao County as the research area,uses the Flash Flood Modul Smiulation Stytem and the Xin’anjiang model to simulate 22 floods in the basin,and uses the simple polygon evolution global search algorithm(SCE-UA)and The adaptive wind-driven optimization algorithm(AWDO)conducts calibration research on model parameters,and compares the advantages and disadvantages of the two models and the two parameter calibration methods.The purpose of this paper is to study the applicability of the Flash Flood Modul Smiulation Stytem for mountain flood prediction in small watershed,and to explore a faster and more accurate parameter calibration method.The main research contents are as follows:1)Analyze changes in the characteristics of runoff series in the study area According to the runoff data of Baitugang Hydrological Station from 1971 to 2017,the evolution trend and evolution law of its long-term series are analyzed and studied.The results show that the runoff sequence in the upper reaches of the Baihe River shows an increasing trend;the mutation points of the runoff sequence are 1973,1981,and 1987;there are two types of periodic changes in the runoff sequence: one is 6a-9a,About 8a is the center of oscillation;the second is 12a-16 a,and about 14 a is the center of oscillation.2)The Flash Flood Modul Smiulation Stytem was established,GLUE method and Sobol method are used to analyze the sensitivity of model parameters and simulate floods to evaluate the applicability of the model in the study area.The results show that the main sensitive parameters of this study are pref＿flow＿den,layer＿top＿depth,fastcoef＿lin,initial＿water＿content.The average Nash efficiency coefficient of the 22 floods is 0.711,and the average peak relative error percentage is 6.80%.The simulated flood process has good consistency with the measured flood process.It can be considered that the Flash Flood Modul Smiulation Stytem has good applicability in the upper Baihe River basin.3)SCE-UA algorithm,AWDO algorithm,SSA algorithm and GWO algorithm were used to carry out the test function simulation experiment,and SCE-UA algorithm and AWDO algorithm were used to calibrate the parameters of the Flash Flood Modul Smiulation Stytem respectively.The results showed: According to the simulation results,the comprehensive performance of AWDO algorithm is the best.The SCE-UA algorithm and AWDO algorithm were applied to the parameter calibration of the Flash Flood Modul Smiulation Stytem.From the two indexes of Nash coefficient and the percentage of flood peak relative error,it can be seen that AWDO algorithm is better than SCE-UA algorithm in the optimization result.4)Xin ’anjiang model was established to simulate 22 floods,and SCE-UA algorithm and AWDO algorithm were used to calibrate model parameters respectively.The results show that the average Nash efficiency coefficient of 22 floods is 0.847,and the average flood peak relative error percentage is 17.65%.The Xin ’anjiang model also has good applicability in the upper reaches of Baihe River basin.And AWDO algorithm is also superior to SCE-UA algorithm in parameter calibration results.5)The simulated flood results were compared between the Flash Flood Modul Smiulation Stytem and the Xin ’anjiang model.The results show that the Nash-Efficiency coefficient of the Flash Flood Modul Smiulation Stytem is similar to that of the Xin ’anjiang model,but the percentage of flood peak error is better than that of the Xin ’anjiang model.Therefore,the Flash Flood Modul Smiulation Stytem is more suitable for the upper reaches of Baihe River basin,and is more suitable for the flood prediction and early warning of small watershed in mountainous areas.According to the above content,this paper explores the applicability of the Flash Flood Modul Smiulation Stytem in the upper reaches of Baihe River basin,and uses the SCE-UA algorithm and AWDO algorithm to calibrate the model parameters,in order to provide a reference for the flood prediction of small watershed in mountainous area and the parameter calibration of the model. |
PDF Full Text Request