A Comparative Study On The Rainstorm-standard Cohesion Relationship Between The Municipal Drainage And The Water Drainage With Different Sample Selection Methods

Affected by the extreme global climate change and the urbanization process,waterlogging disasters in our cities occur frequently.The urban waterlogging prevention and control system is usually composed of two parts: the municipal drainage system and the water conservancy drainage system,which are relationship of upstream and downstream in the process of rainwater drainage.So it will lead to the problem of the connection of the drainage capacity of the drainage system.Therefore,it is necessary to solve the problem of connection between the stormwater standards of municipal drainage and water conservancy drainage.The central urban area of Guangzhou was researched as a example in this paper to study the connection between the storm drain standard of municipal drainage and water conservancy drainage under annual maximum sampling method and over-threshold sampling method,and the impact of different sample selection methods on the storm standard between municipal drainage and water conservancy drainage was compared.The main contents and conclusions are as follows:(1)Analyze the elements of urban waterlogging,the relationship between the two-stage drainage system and the causes of waterlogging in the study area.Urban waterlogging disasters are formed by the combined action of three factors: disaster-causing factors,disaster-bearing bodies and disaster-preventing environments.However,the non-coupling of the municipal drainage system and water conservancy drainage system standards increases the risk of waterlogging.(2)The storm data series of four municipal drainage durations of 30-min,60-min,90-min and 120-min and a hydraulic drainage duration of 24 h in Guangzhou Meteorological Station from 1980 to 2013 were used for analysis and research.The univariate distribution law of the municipal storm drain and the water drainage drain storm under the annual maximum sample selection and the over-threshold sample selection are studied respectively,and the effects of the sample selection method on the design storm rainfall results are compared.Univariate frequency analysis based on the annual maximum sample selection shows that the P-? distribution and the generalized extreme value distribution fit the annual maximum sample selection rainstorm series very similar when the frequency is between 10% and 90%,but there are differences at both ends.Univariate frequency analysis research based on over-threshold sampling shows that:(1)The over-threshold distribution shows convergence when the threshold reaches a certain level;(2)The threshold is too small will affect the independence of the sample and increase the error of parameter estimation;(3)The overthreshold sampling method can ensure that the selection of the threshold within a certain range will not cause a large deviation in the design results of the rainstorm.Comparative analysis under the two sample selection methods shows that:(1)The super-threshold distribution model of the super-threshold sampling method is simple to calculate and has good stability,and the threshold has a wide selection range;(2)When it is not less than 5 years,the design values of the two sample selection methods are small,and the results in the small recurrence period(under 5 years)are quite different,especially under 2 years which up to 30 %;(3)The over-threshold sample selection makes full use of rainstorm information,and the design results are more objective and reasonable.(3)The empirical method and the Copula function method are used to study the connection relationship between municipal drainage storms and hydraulic drainage storms under annual maximum sampling and over-threshold sampling,respectively,and compare the impact of sampling selection methods on the results of convergence analysis.Research on empirical methods shows that the empirical connection relationship is susceptible to extreme rainstorm events in historical data,which leads to unreasonable connection relationships.The comparison of the stormwater standard connection between municipal drainage and waterlogging under the two sample selection methods based on the Copula function method shows that the combined risk fluctuation range based on the overthreshold sampling method is smaller than annual maximum method.For example,when 2-year municipal drainage rainstorm and 10-year water conservancy drainage rainstorm is combined,the combination of sample selection based on over-threshold and sample selection based on annual maximum value,the risk rates when municipal drainage is the main and water conservancy is corresponding are 0.5%?0.84% and 0.16%?1.16%,with water conservancy drainage mainly,municipal drainage correspondingly are 12.6%?14.9% and 20.58%?30.3% respectively,this is mainly because the annual maximum method selects some smaller rainstorm information and misses some larger rainstorm information,resulting in information deviation of the joint observation sample based on the annual maximum value selection is large.