产流模式空间分布对城市雨洪过程模拟的影响

Effect of spatial distribution of runoff generation patterns on urban storm flood process simulation

通过耦合网格产流计算模型(GRGM)和暴雨洪水管理模型(SWMM)汇流计算模块构建了GRGM-SWMM模型。以贾鲁河中牟水文站控制流域(包括郑州市中心城区)为研究区,基于18场实测暴雨洪水资料探究了产流模式空间分布对洪水过程模拟精度的影响。结果表明:研究区超渗产流模式空间分布面积占比高达72%,城市下垫面仍以超渗产流模式为主;相较于仅考虑超渗产流计算的SWMM,考虑超渗、蓄满和混合等不同产流模式空间分布的GRGM-SWMM模型场次洪水产流量、流量过程模拟精度明显提高,产流量模拟相对误差平均降低33.26%、决定系数平均提高0.089,流量过程模拟均方根误差平均降低5.13 m^(3)/s、纳什效率系数和决定系数分别平均提高0.227、0.207。

The GRGM-SWMM model was constructed by coupling the grid runoff generation model(GRGM)and the rainstorm flood management model(SWMM).Taking the control basin of Zhongmou Hydrological Station on the Jialu River(including the central urban area of Zhengzhou City)as the study area,based on 18 measured rainstorm and flood data,the impact of spatial distribution of runoff generation modes on the accuracy of flood process simulation was explored.The results show that the spatial distribution of the infiltration-excess runoff generation pattern in the study area accounts for up to 72%,and the urban underlying surface is still dominated by the infiltration-excess runoff generation pattern.Compared to the SWMM that only considers the infiltration-excess runoff generation for runoff calculations,the GRGM-SWMM model,which considers different runoff generation patterns such as infiltration,saturation,and mixing,significantly improves the accuracy of peak flood discharge and flow process simulation.The relative error of runoff generation decreased by an average of 33.26%,and the determination coefficient increased by an average of 0.089.The root mean square error of simulated discharge also decreased by an average of 5.13 m^(3)/s,while the Nash-Sutcliffe efficiency coefficient and the determination coefficient increased by an average of 0.227 and 0.207,respectively.