城市路网汇流线性效应与Clark分布式单位线适用性检验
Assessment on the Linear Effect of Urban Road Network Confluence and Applicability of Clark Unit Hydrograph Model
作者：申红彬(华北水利水电大学 河南省水利水运工程技术研究中心, 河南 郑州 450045;北京师范大学 水科学研究院 城市水循环与海绵城市技术北京市重点实验室, 北京 100875)；徐宗学(北京师范大学 水科学研究院 城市水循环与海绵城市技术北京市重点实验室, 北京 100875)；王海周(华北水利水电大学 河南省水利水运工程技术研究中心, 河南 郑州 450045)；张书函(北京市水科学技术研究院 北京市非常规水资源开发利用与节水工程技术研究中心, 北京 100048)
Author：SHEN Hongbin(Henan Eng. Research Center of Hydraulic and Waterway, North China Univ. of Water Resources and Electric Power, Zhengzhou 450045, China;Beijing Key Lab. of Urban Hydrological Cycle and Sponge City Technol., College of Water Sci., Beijing Normal Univ., Beijing 100875, China)；XU Zongxue(Beijing Key Lab. of Urban Hydrological Cycle and Sponge City Technol., College of Water Sci., Beijing Normal Univ., Beijing 100875, China)；WANG Haizhou(Henan Eng. Research Center of Hydraulic and Waterway, North China Univ. of Water Resources and Electric Power, Zhengzhou 450045, China)；ZHANG Shuhan(Beijing Eng. Research Center for Non-Conventional Water Resources Utilization and Water Saving, Beijing Water Sci. and Technol. Inst., Beijing 100048, China)
收稿日期：2019-05-24 年卷（期）页码：2020,52(3):62-69
期刊名称：工程科学与技术
Journal Name：Advanced Engineering Sciences
关键字：水文模型;城市路网;汇流;Clark单位线;线性效应
Key words：hydrological model;urban road network;confluence;Clark unit hydrograph;linear effect
基金项目：城市水循环与海绵城市技术北京市重点实验室开放基金（HYD2019OF02）；国家科技重大科技专项课题资助项目（2013ZX07304-001）
中文摘要
模型适用性是水文模型应用需要考虑的一个重要问题。针对Clark分布式单位线模型，结合北京未来科技城北区城市路网降雨径流监测数据，通过模型模拟与效果评价（包括确定性系数R2、Nash-Sutcliffe效率系数NSE和总径流量相对误差RE），综合检验路网汇流的线性效应与Clark分布式单位线法的适用性。结果表明：径流流量模拟值与实测值变化趋势基本符合，确定性系数R2值在0.40～0.89之间，平均值为0.73；Nash-Sutcliffe效率系数NSE值在-0.77～0.87之间，平均值为0.26；总径流量相对误差RE值在-0.77～0.36之间，平均值为-0.18；总体模拟效果可以接受。其中，不同场次降雨径流模型评价指标变化与降雨强度密切相关，确定性系数R2值与Nash-Sutcliffe效率系数NSE值均随雨强的增大而逐步趋近于1，总径流量相对误差RE值则随雨强的增大而振荡趋近于0，说明模型效果随雨强增大而趋于优化、适用性程度不断提高，并间接反映出路网汇流的线性效应不断增强。
英文摘要
The applicability of a hydrological model is an important issue for its application. On the basis of Clark linear distributed unit hydrograph model, the urban road network confluence in the North region of Future Science & Technology Park located in Beijing was selected for case study, in order to comprehensively validate the linear effect of road network confluence and applicability of Clark unit hydrograph model. The model performance was tested out by using determination coefficient (R2), Nash-Sutcliffe efficiency (NSE) and relative error of total runoff (RE). The results showed that the average value of determination coefficient is 0.73 which varies from 0.4 to 0.89, the average value of Nash-Sutcliffe efficiency is 0.26 which varies from -0.77 to 0.87, the average value of relative error of total runoff is -0.18 which varies from -0.77 to 0.36. The variation rules of different indexes are closely related to the rainfall intensity for individual rainfall, in which with the increase of rainfall intensity, the determination coefficient and Nash-Sutcliffe efficiency increase toward 1.0, with the relative error of total runoff approaches zer. This indicated that the simulation effect of Clark unit hydrograph model tends to be optimal as the linear effect of road network confluence process increases continuously with the increase of rainfall intensity.
【关闭】
