Simulation Of Rain And Floof In Fengxi Sponge City Based On Random Rain Pattern

Sponge city is one of the important measures to prevent urban floods and ensure water safety in China.The spatial and temporal heterogeneity of rainstorm in urban small watersheds is strong,and it is difficult for deterministic rain patterns to accurately describe their rainfall characteristics,and the uncertainty of rain and flood simulation results is large.Therefore,it is of great theoretical significance to study the simulation of rain and flood in sponge city generated by random rain patterns.In this paper,taking the experimental area of sponge city in Fengxi New Town,Xi ’an as the research object,the measured rainfall data of 120min rainfall duration are fitted in sections(10min)by using Gaussian,Weibull,Lognormal,Gamma and Generalized extreme distribution,and the fitting effect is evaluated according to the fitting degree R2,AIC criterion,RMSE evaluation and AD-P goodness of fit test,so as to determine the optimal distribution function of each period.Through the characteristic parameters of rain patterns(rain peak position coefficient r and peak times ratio bmax),the random rain patterns satisfying the optimal distribution and boundary conditions are selected.A total of 9 rain pattern scenarios were selected,including only changing the rain peak position,only changing the peak ratio,and changing both the rain peak position and peak ratio.The storm flood management model(SWMM)is used to simulate the process of rain and flood with rainfall return periods of one year,two years and five years,and the characteristics and law s of rain and flood in Sponge City are revealed.The research results are as follows:(1)By using the random rain pattern generation method based on variable parameter statistics,the typical rain patterns in the study area were randomly generated.The results showed that the distribution fitting results of each time period were not the same.The third,fifth and seventh time periods were optimally Gamma distribution;The best generalized extreme value distribution is in the 6th period and the 9th period;the Log Normal distribution is the best fit for the other periods,and the randomly generated rain patterns are abundant.(2)By constructing SWMM model in the study area,the process of rain and flood in different rainfall scenarios is simulated,and the process of flood production and outflow is analyzed,and the relationship between the characteristic value of rain pattern and the change of flood process is obtained.Both the process of rain production and outflow are sensitive to the characteristic value of rain pattern.(3)By calculating the four indexes of peak time,peak reduction rate,total outflow and comprehensive runoff coefficient in different recurrence periods,the law of rain and flood when the characteristics of different recurrence internal rain patterns change is obtained.It is found that the location coefficient of rain peak is the key to the peak time,and the ratio of peak times has a great influence on the other three indexes.The sponge city low-impact development system has poor ability to reduce rainfall in the high return period when the rain peak position is backward and the peak ratio is large.Based on the measured rainfall data,the process of rain and flood is simulated.The calculation results show that the random rain pattern can effectively simulate the uncertainty and rough result range caused by the randomness of rainfall under natural conditions,and more comprehensively include the rain flood law of Sponge City.The results can provide a reference for the storm and flood conditions involved in the construction and management of sponge cities.