Research On The Formula Of Rainstorm Intensity Of China Grid Based On Remote Sensing

Under the dual effects of global climate change and the accelerating process of urbanization in China,urban extreme precipitation leads to frequent urban waterlogging disasters.As one of the basic basis of urban rainwater drainage system planning and design,rainstorm intensity formula directly affects the investment of drainage engineering and urban safety.However,at present,in many cities in China,the rainstorm intensity formula used is old,and in many small cities,the local rainstorm intensity formula has not been worked out,only the formula of the adjacent areas can be used,which is extremely unfavorable to the urban infrastructure construction.The traditional formula of rainstorm intensity is based on the precipitation data of meteorological stations,which has the problems of uneven spatial distribution and difficult data acquisition.In order to solve the above problems,this paper uses the CMORPH satellite precipitation data with high spatial resolution(8km)to compile the rainstorm intensity formula at grid scale in China.The monthly precipitation data of 31 provincial capital cities were selected to evaluate the accuracy of CMORPH precipitation data.Taking Beijing Meteorological Station as an example,the method of fitting frequency distribution curve and rainstorm intensity formula parameters is selected.Finally,the frequency distribution curve and the formula parameter fitting method are applied to the derivation of the rainstorm intensity formula in the mainland of China,and finally the rainstorm intensity formula in the grid scale of the whole mainland of China is obtained.The main results are as follows:(1)The precision of CMORPH precipitation data is evaluated by the measured monthly precipitation of 31 provincial capital city meteorological stations.It is found that the correlation coefficient is 0.8013,the determination coefficient R~2 is 0.75,the RMSE value is 48.210mm/month,the MAE value is 29.012mm/month,and the relative error is-6.8%.It shows that although the capture of CMORPH precipitation data to precipitation is underestimated,the expression and control of CMORPH precipitation data to precipitation are obvious the overall system capacity is good.(2)It is found that the fitting effect of Pearson distribution is the best,Gumbel distribution is the second,and the Weibull distribution is the worst.Using Mcquardt method,quasi Newton method and genetic algorithm to fit the formula parameters of frequency analysis results,it is concluded that the fitting formula parameters error of quasi Newton method is the smallest,followed by genetic algorithm,and Mcquardt method is the largest.(3)According to Pearson's three type distribution and quasi Newton's method,the rainstorm intensity formula on the scale of grid in China's mainland is derived.More than 80%of the grids can meet the requirements of absolute mean square deviation(no more than 0.05mm/min)and relative mean square deviation(no more than 5%)in the standard,and most of the grids that fail to meet the requirements are concentrated in Xinjiang,Qinghai,northwest and northeast regions with small precipitation.(4)The spatial distribution of the parameters of the rainstorm intensity formula derived from the analysis shows that parameter a is generally small in the West and northwest,and the absolute mean square deviation of the whole eastern region and the Qinghai Tibet Plateau is relatively large;parameter C is mainly increasing from the southeast coast to the northwest inland,but the increasing range is not obvious;parameter B has no obvious spatial distribution characteristics;parameter n is more than the northwest inland It is large,and its value is evenly distributed in the middle,East and Southeast,but generally small,.(5)The correlation analysis between the existing rainstorm intensity formula in31 provincial capitals and the rainstorm intensity formula based on CMORPH satellite precipitation data shows that the correlation between the two is greater than 0.9 in 31cities,Haikou City has the highest correlation and Urumqi city has the lowest correlation.Comparing the root mean square error and the mean absolute error of the two indexes in 31 cities,Urumqi is the smallest,and Nanjing is the largest.Analysis of the relative error,except for Zhengzhou,Nanning and Haikou,the relative error of the three cities is negative,the other cities are positive,which shows that in most cities,the rainfall intensity calculated by CMORPH formula is higher than that calculated by the existing formula.When the rainfall duration is less than 120min in all provincial capitals,the rainfall intensity of CMORPH is generally higher than the rainfall intensity calculated by the existing rainstorm intensity formula based on the station,while when the rainfall duration is greater than 120min,it is the opposite.The analysis of the above results is mainly due to the difference between the two Rainstorm Intensity Formulas in the selection of rainfall data year,sample selection method,rainfall duration selection,frequency curve selection and parameter fitting method.The results of this paper can provide a scientific basis and reference for the design of urban rainwater pipe network,and provide a choice for the old city which lacks the rainstorm intensity formula and the existing rainstorm intensity formula.