面雨量的不确定性对洪水期间径流模拟的影响  

The influence of the uncertainty of area rainfall on the runoff simulation during flood

作　　者：吴国芳 张世炜 何立群 WU Guofang;ZHANG Shiwei;HE Liqun(Shaoxing Water Conservancy and Hydropower Survey and Design Institute Co.,Ltd.,Shaoxing 312000,Zhejiang,China)

机构地区：[1]绍兴市水利水电勘测设计院有限公司,浙江绍兴312000

出　　处：《水利水电技术（中英文）》2023年第S02期141-151,共11页Water Resources and Hydropower Engineering

摘　　要：降雨是洪水过程模拟的重要输入,研究降雨输入的不确定性对洪水期间径流模拟的影响是优化洪水预报的重要手段。为研究不同降雨数据密度下面降雨量的特征以及其对径流模拟结果的精度影响,以曹娥江流域上浦闸以上为研究区域,2021年“烟花”洪水为研究案例,选取不同数量的实测降雨站点数量,使用泰森多边形插值法对研究区域内实测点降雨数据进行空间插值得到面降雨量。在高密度降雨数据情况下,使用自研发的率定软件对NAM水文模型和HD水动力的耦合模型进行模型参数的联合率定。确定模型参数后,模拟不同密度的面降雨输入下的径流结果,计算实测径流过程与模拟径流过程的误差,量化面雨量计算的不确定性对模拟精度的影响。结果显示:降雨数据密度由低至高,各站的模拟径流确定性系数分别为,嵊州新市站0.84、0.83、0.83、0.76;嵊州站0.97、0.96、0.95、0.92;黄泽站0.83、0.91、0.89、0.90;东山站0.97、0.97、0.98、0.98。结果表明:在降雨数据较少的情况下,面降雨数据易受到单个站点的数据影响,径流模拟结果的不确定性较大。整体而言,降雨数据的密度越大,面降雨量的不确定性越小,径流模拟结果的精度越高;在降雨数据达到一定密度后,降雨数据的增加对精度没有明显的提升效果。Rainfall is an important input for flood process simulation.It is a vital means to optimize flood forecasting by studying the influence of rainfall input uncertainty on runoff simulation during flood.In order to research the characteristics of rainfall under different rainfall data densities and its impact on the accuracy of runoff simulation result,taking the front basin of the Cao’e River as the research area,the"YanHua"flood in 2021 as the research case,selecting different numbers of measured rainfall data,using Thiessen polygon interpolation method to obtain the area rainfall under several densities of rainfall data.In the case of high-density rainfall data,using the self-developed calibration software to jointly calibrate the model parameters of the NAM hydrological model and the HD hydrodynamic coupling model.After determining the model parameters,simulating the runoff process under different densities of area rainfall input,calculating the error between the measured runoff process and the simulated runoff process,and quantify the influence of the uncertainty of area rainfall calculation on the simulation accuracy.The rainfall data density ranges from low to high,and the simulated runoff certainty coefficients for each station are 0.84,0.83,0.83,and 0.76 for Xinshi Station in Shengzhou,0.97,0.96,0.95,and 0.92 for Shengzhou Station,0.83,0.91,0.89,and 0.90 for Huangze Station,and 0.97,0.97,0.98,and 0.98 for Dongshan Station.The result show that area rainfall data are easily affected by the data of a single station in the case of less rainfall data,and the uncertainty of runoff simulation result is relatively large.Overall,the greater the density of rainfall data,the smaller the uncertainty of area rainfall and the higher the accuracy of runoff simulation result.However,after the rainfall data reaches a certain density,the increase of rainfall data does not significantly improve the accuracy.

关 键 词：面雨量 泰森多边形插值法 NAM水文模型 不确定性 降雨数据密度 

分 类 号：TV121.1[水利工程—水文学及水资源]