This paper presents a new method to determine parameters in the nonlinear mathematical model of the urban rainfall intensity formula. The method transforms the nonlinear equation into a linear one, and incorperates a ...This paper presents a new method to determine parameters in the nonlinear mathematical model of the urban rainfall intensity formula. The method transforms the nonlinear equation into a linear one, and incorperates a damping factor and a step factor to improve the precision of the calculated results. It works out the parameters based on given rainstorm intensity, period of the recurrence of rainstorm and lasting time of the rainfall. The results turns out to have the smallest residual error compared with those obtained by several other methods and satisfy the related standard. The method proves of fast convergence, desirable generality, stability and accuracy, which overcomes the defects of other existing numeral methods.展开更多
A total of 58 years self-recording rainfall records at Guilin National Weather Station from January 1957 to December 1957 were used. Using the annual maximum sampling method,11 short-duration maximum rainfall records ...A total of 58 years self-recording rainfall records at Guilin National Weather Station from January 1957 to December 1957 were used. Using the annual maximum sampling method,11 short-duration maximum rainfall records in every year were selected,and statistical samples of rainstorm intensity formula were obtained. Fitting of statistical samples was conducted by using three kinds of frequency distribution curves: Pearson-Ⅲdistribution,Gumbel distribution and exponential distribution. The formula of rainstorm intensity was derived by the least square method and Gauss Newton method,and then the accuracy of results was inspected. The results showed that the fitting effect based on the Pearson-Ⅲdistribution and Gumbel distribution was better than that of the exponential distribution;the accuracy of rainstorm intensity formula based on Pearson-Ⅲdistribution was better than that of Gumbel distribution and exponential distribution;the accuracy of rainstorm intensity formula derived by the least square method was better than that of Gauss Newton method;the accuracy of interval parameter formula of rainstorm intensity was better than that of rainstorm intensity formula.展开更多
文摘This paper presents a new method to determine parameters in the nonlinear mathematical model of the urban rainfall intensity formula. The method transforms the nonlinear equation into a linear one, and incorperates a damping factor and a step factor to improve the precision of the calculated results. It works out the parameters based on given rainstorm intensity, period of the recurrence of rainstorm and lasting time of the rainfall. The results turns out to have the smallest residual error compared with those obtained by several other methods and satisfy the related standard. The method proves of fast convergence, desirable generality, stability and accuracy, which overcomes the defects of other existing numeral methods.
基金Supported by Meteorological Scientific Research Plan Project of Guangxi Meteorological Bureau(2019M16)Independent Scientific Research Project of Guilin Meteorological Bureau(201904)。
文摘A total of 58 years self-recording rainfall records at Guilin National Weather Station from January 1957 to December 1957 were used. Using the annual maximum sampling method,11 short-duration maximum rainfall records in every year were selected,and statistical samples of rainstorm intensity formula were obtained. Fitting of statistical samples was conducted by using three kinds of frequency distribution curves: Pearson-Ⅲdistribution,Gumbel distribution and exponential distribution. The formula of rainstorm intensity was derived by the least square method and Gauss Newton method,and then the accuracy of results was inspected. The results showed that the fitting effect based on the Pearson-Ⅲdistribution and Gumbel distribution was better than that of the exponential distribution;the accuracy of rainstorm intensity formula based on Pearson-Ⅲdistribution was better than that of Gumbel distribution and exponential distribution;the accuracy of rainstorm intensity formula derived by the least square method was better than that of Gauss Newton method;the accuracy of interval parameter formula of rainstorm intensity was better than that of rainstorm intensity formula.
