Research On Robustness Analysis Of Rainfall Water Quality Process Simulation And Optimization Design For Sponge Facility

In the context of global climate change,with the rapid urbanization process in China,there are still huge challenges in urban rainwater management.The concept of Sponge City construction put forward new ideas for rainwater management in China.However,the current research on water quality models that guide the planning and design of sponge city is still insufficient,and the rationality of its application in rainwater management design cannot be guaranteed.In addition,in the planning and design of sponge city,there is usually a lack of consideration of multiple cost-effect indicators and detailed design.In order to ensure the effect of sponge city construction,this research is carried out around urban rainwater management.Based on the field monitoring data in sponge city demonstration area and the automatic calibration framework,the water quality process of the urban rainwater model SWMM was studied,and the suggestion for water quality model improvement and model-based LID design were given.At the same time,a new multi-level(LID facility selection,connection and implementation scale)optimization design framework for sponge city construction considering multi-objectives(runoff reduction,flood peak reduction,pollution reduction,cost,etc.)was proposed and the method was successfully verified.Based on theoretical research,we managed to establish a localized sponge facility optimization design platform,which can be applied to the planning and design of sponge city construction in China.Based on the packages of R language and optimization algorithm,an automatic calibration framework for the SWMM model was constructed.Based on the objective automatic calibration framework and on-site water quality monitoring of natural rainfall events,using the Nash-Sutcliffe Efficiency Coefficient(NSE),a new concept of model robustness of water quality process and its analysis method were proposed,to evaluate the pollutant control performance of LID infrastructure.In a sponge city demonstration area in Shenzhen,China,a total of 166 calibrations were completed.The analysis of the calibration results showed the heterogeneity of the robustness indicator among different facilities and water quality variables.Analysis revealed that the rainfall characteristics are related to the indicator,namely both the rainfall intensity and the antecedent dry day are beneficial to the model robustness.K-means clustering shows that the model robustness increases with the the pollutant concentration within a certain range(<1 mg/L).The robustness analysis of the water quality model suggests us rethink whether existing rainwater management measures and future-oriented LID design can effectively cope with rapid changes in external conditions,especially in the context of ecological city construction and climate change.On the basis of water quality model research,a multi-objective and multi-level sponge facility optimization design research method based on genetic algorithm and multi-criteria analysis was established,which is used for the optimization of the type selection,connection selection and implementation scale of LID facility layout in sponge city construction,considering environmental benefits and costs.Collected and processed various geological and meteorological data in the study area,SUSTAIN model was used to analyze the scope of the planned LID facility to be deployed in the study area,and to generate the SWMM model of the study area.The monitoring data were used to calibrate and validate the SWMM model,and the NSEs are 0.77 and 0.52,respectively.A multiobjective and multi-level sponge facility optimization design framework was proposed:runoff reduction rate,flood peak reduction rate,TSS reduction rate and construction cost were selected as optimization objectives and evaluation criteria,and their quantitative methods were determined as well.The "swmmr" package based on R language was implemented to coding of LID selection,connection mode and implementation scale in the catchment,as optimization variables.The suitable plot and its SWMM model used to test the feasibility of the optimization design framework were selected and set.The optimization algorithm was debugged to determine the optimal configuration,and the optimization framework was run to obtain a Pareto front surface composed of a total of94 non-dominated solutions;the TOPSIS algorithm was used to optimize the 94 nondominated solutions: the entropy weight method was used to determine the criterion weights,then the optimal solution with the largest degree of closeness and the worst solution with the smallest degree of closeness are obtained and the scheme analysis is carried out.Fixing the weight of cost and equally dividing the weights of environmental criteria,the influence of cost weight changes on the selection of schemes was studied.Based on the research of water quality model and the multi-level and multi-objective optimization design framework of sponge facility,the existing advanced urban hydrological model SUSTAIN was re-developed,and a set of localized sponge facility optimization design platform was constructed,which on a mature visualization platform can be used to realize the multi-level and multi-objective optimization design of sponge facility,and serves the planning and design of sponge city construction in China.Through the analysis of the original project files,the development environment was established and the project was successfully modified and compiled.According to the actual development requirements and development difficulties,some development tasks and part of the localization of the original SUSTAIN were completed,which layed a good foundation for the subsequent improvement and development of the platform.