| The rainstorm intensity formula is an important basis for the design of rainwater drainage and flood prevention engineering in cities, and it directly affects the function of disaster prevention and the environment of cities. With the development of society and change of environment, rainstorm intensity formulas in many large cities have become unsuitable. Furthermore, many middle and small cities don't have the formulas of their own. Just because of the latter problem, the development of Mianyang city is constrained, and new suitable formulas is urgently needed.Based on the original rainstorm data in Mianyang city, the project compares four samplings, analyzes three frequency models, gets 12 groups of i~t~T data, and then calculates the formulas of 12 models through a calculating software applying Marqurlt method. With the calculated formulas' errors as the standard, the project optimizes the best sampling method and frequency model which suits Mianyang best, and determines the rainstorm formulas with high precision. In this project, the conclusions are as follows:(1) Through the comparison of four sorts of rainstorm data, the annual several samples has the best representation, but the method' independency is poor and the sampling work is large, and it suits the short rainstorm data better. Super-standard amount method will not be considered because it is the same as annual several samples method and annual super-large value method respectively due to the different selected standards. The annual super-large value method and the annual maximum value method are simple but have less representation. To the rainstorm data with long measure period and good representation, the annual super-large method or the annual maximum value method could be selected.(2) Through the analysis of the errors of 12 groups of data with three frequency modeling, it is concluded that the P-III distributed has the smallest error; the exponential distributed is on the second place, and the Gumbel distributed has the biggest error. Sampling with the annual several samples method and adjustment with P-III model has the smallest error, and next is sampling with the annual several samples method and adjustment with exponential and Gumbel model.(3) Through the errors comparison after the determination of parameters in rainstorm intensity formula, the error which use annual several samples method and exponential distribution is found to be the smallest, which is different from the results from frequency analysis. Thus, when calculating formula of urban rainstorm intensity, the best mode is to determine the formula with considering the development requirement of city,using range,characteristics of rainstorm data,frequency distribution and errors of rainstorm formula.(4) Through the analysis of the curves of 12 modes double logarithmic i~t in different recurrence period, it is found that the conclusion is in accordance with the results analyzed by rainstorm intensity formula. In other words, through the i~t~T data' double logarithmic curves, the best sample selection and frequency model can be got before calculating storm equation, so the work optimizing storm formula parameter can be reduced greatly.(5) High precision and applicable rainstorm intensity general formula and single recurrence period part formula in Mianyang city are optimized. |