全过程联合校正方法在饶河流域洪水预报中的应用研究  

Study on the Entire-Process Joint Correction Method for Flood Forecasting in Raohe Watershed

作　　者：王钦钊 黄一昕 梁忠民[2] 邓晓栋 WANG Qinzhao;HUANG Yixin;LIANG Zhongmin;DENG Xiaodong(Hydrological Monitoring Center of Jiangxi Province,Nanchang 330002,China;College of Hydrology and Water Resources,Hohai University,Nanjing 210098,China;Hydrology and Water Resources Monitoring Center of Xinjiang River and Raohe River of Jiangxi Province,Jingdezhen 333003,China)

机构地区：[1]江西省水文监测中心,江西南昌330002 [2]河海大学水文水资源学院,江苏南京210098 [3]江西省信江饶河水文水资源监测中心,江西景德镇333003

出　　处：《水文》2021年第2期45-50,共6页Journal of China Hydrology

基　　金：国家自然科学基金(41730750);国家重点研发计划(2018YFC0407206,2016YFC0402709);江西省水利科技项目(201921YBKT21)。

摘　　要：为提高饶河流域洪水预报精度,将全过程联合校正(EPJC)方法与三水源新安江(XAJ)模型结合,按一定比例划分洪水预报总误差为各过程误差,基于系统响应理论反演得到面雨量计算误差和模型参数误差,重新输入流域水文模型正演得出校正后的洪水过程。选取洪峰流量相对误差、峰现时间绝对误差、径流深相对误差和确定性系数作为模型评价指标,量化XAJ预报与EPJC校正结果,并与传统的动态系统响应曲线(DSRC)方法进行对比。结果表明,EPJC能够考虑全过程误差,与实际误差分布情况更相符,可提升饶河流域的洪水作业预报精度,值得进一步研究与推广。In order to improve the accuracy of flood forecasting in Raohe watershed,this paper combined entire-process joint correction(EPJC)method and XAJ model,and divided the total error of flood forecasting into each process error by proportion.The errors of areal rainfall and model parameter were estimated based on the inversion of the system response theory.Then the corrected flood forecasting process was obtained by the forward modeling of error.Four indexes,including the relative error of peak discharge,the absolute error of peak lag time,the relative error of runoff depth and the Nash-Sutcliffe coefficient were selected for evaluating the performance of XAJ model and EPJC method.Then,the EPJC method was compared with a traditional dynamic system response curve(DSRC)method.The results show that EPJC can consider the error of entire process,which is more consistent with the actual error distribution.

关 键 词：实时洪水预报 全过程联合校正 系统响应理论 误差修正 饶河流域 新安江模型 

分 类 号：P338[天文地球—水文科学]