Urban Flood Model Based On Local Inertial Approximate Shallow-water Equations And Its Application

Under the background of global climate change and urbanization,urban flood has become one of the key factors restricting social and economic development,and seriously threatens the safety of people’s lives and property.It has become the focus of society at present.With the rapid development of numerical computing technology,the numerical simulation technology,as an important tool to study urban flood disaster management,has been widely concerned by the academic circles and has been developed by leaps and bounds.Based on the previous research,this paper summarizes the existing shortcomings and new ideas worth trying in urban flood modeling.A finite volume solution of a simplified scheme of shallow water equations-local inertial approximate equations is proposed.The urban flood modelling technology based on the local inertial approximate equation was studied.The main work of the thesis includes:(1)A finite volume method based on unstructured grid was proposed to solve the local inertial approximate equation.The numerical flux at the mesh interface was evaluated by the Roe Riemann solver.A new Surface Reconstruction Method and an implicit discrete scheme of friction source term were applied to improve the stability and computational efficiency of the numerical scheme.Finally,a two-dimensional hydrodynamic numerical model(LIM)based on the local inertial approximate equations was constructed.Four different benchmark cases showed that calculation efficiency of the LIM model can be improved by 160% ～ 180% with high accuracy.(2)Based on the research of the calculation principle of SWMM model,the onedimensional and two-dimensional coupling methods in urban flood model were discussed in detail.According to their different water exchange modes,three coupling methods were adopted to calculate the exchange water.The results of three theoretical cases showed that the coupling models have high accuracy and stability.The construction methods of the integrated hydrology and hydrodynamics model and the full-hydrodynamic model were discussed in detail,and the characteristics of the two construction methods were compared.(3)Based on the research on construction methods of urban flood model,an integrated hydrology and hydrodynamics model and a full-hydrodynamic model based on the LIM were constructed respectively.Taking two regions with high urbanization characteristics as research cases,the urban stormwater flood process was simulated,and the urban flood model based on shallow water equation was used as the benchmark.The performance of two low-complexity urban flood models constructed in this paper was analyzed and compared.The results showed that the two low-complexity models can improve the calculation efficiency by more than 140%,and can accurately describe the flood characteristics in subcritical flow regime(the relative error is less than 10%),but can not correctly describe a few local supercritical flows(the relative error is more than 50%).Overall,the LIM’s performance in urban flood model is satisfactory,which can be an alternative to construct the urban flood model.(4)A numerical solution scheme of the LIM-SWM hybrid mode was proposed,and an urban flood model based on the LIM-SWM hybrid model was constructed.The benchmark case test results showed that the proposed solution scheme to the hybrid mode can greatly improve the numerical accuracy of LIM model in local areas,and only has a slight impact on the computational efficiency.The high urbanization case results showed that under the appropriate Froude number factor,the hybrid model can correctly describe the flood characteristics in supercritical flow regime on the basis of slightly affecting the calculation efficiency,and it also has a positive impact on in nearby areas,so that the model accuracy has been improved to a certain extent in the whole research area.At the same time,this study found that the Froude number factor is the key parameter for the hybrid model to realize high-precision and highefficiency simulation,and 0.5～0.8 is the value range of the factor recommended by this study.