Parameter Sensitivity Analysis And Calibration Of Urban Rainfall Runoff Model

With the rapid development of the society,the urbanization process,people's living standard improved,the quality of life have been greatly improved,but at the same time all kinds of natural disasters occurrence frequency increases,especially urban waterlogging,not only caused great economic losses to the society,but also cause serious influence to people's life and even threaten the safety of life.The original reason is the rapid development of urban construction,and the low-standard drainage system has been unable to develop synchronously with it.But fundamentally speaking,it is the development of the city that improves the hardening rate of the ground,resulting in the increase of surface runoff,which causes a huge burden on the underground drainage system and further increases the frequency of urban waterlogging.The establishment of urban storm flood model can help us to strengthen the understanding of urban storm flood problems.The simulation analysis of urban storm flood model can not only study the causes of urban storm flood,but also test the feasibility of alleviating urban flood measures through model simulation,which has certain reference value.In this paper,FT district of SZ city was taken as the research area.On the basis of the collected data,GIS platform and SWMM software were used to build the urban storm flood model.The parameter sensitivity analysis and calibration of the model were carried out.1.The status quo of domestic and foreign studies on urban storm flood models and the sensitivity analysis and calibration of common parameters are introduced.2.Construction of urban storm flood model based on SWMM.By using the powerful spatial data processing function of GIS to extract the relevant hydrological data,the rainfall runoff model is constructed.Build the non-point source pollution model and LID model according to relevant data.3.Sensitivity analysis of rainfall runoff model parameters.Morris method was used to analyze the parameter sensitivity of rainfall-runoff model,and the sensitivity degree of parameters was determined according to the classification of parameter sensitivity.4.BP neural network was used to calibrate the parameters of the rainfall runoffmodel.A three-layer BP neural network was built to calibrate the parameters of the model and further optimize the model,so that the error between the simulated results and the measured data of the calibrated parameters could be controlled within 5%.5.LID model reduction effect analysis.Non-point source pollution model and LID model are added into the rainfall runoff model.Through comparative analysis of the reduction effect of LID measures on surface runoff and surface pollutants,it is shown that LID development can effectively alleviate urban waterlogging and provide certain reference value for the construction of sponge city.