人们在进行洪水频率计算时,人为地割裂了洪水过程。一般地,都是以每年的最大洪峰流量对应的洪水作为该年的洪水过程,因此会出现洪水过程遗漏现象。洪水频率结果为防洪工程的建立提供依据,而洪水场次的界定是洪水频率计算的基础。文中研究采用两种方式界定洪水场次,利用Gumbel-Logistic模型分别计算界定的洪水场次的频率结果,并与传统方式下获取的洪水场次的频率结果进行比较。结果表明:3种情形下的洪水频率结果存在较大的差异。其中,以P-III型分布得出6年重现期的洪峰流量为洪水场次选取的门限值获得的洪水场次,即利用本研究中的界定方式二,得到的最大洪峰流量对应的频率值偏低,依此设计防洪工程,相比采用另外两种洪水场次选取方式进行防洪工程的设计偏于安全。
Generally in calculating the flood frequency,the flood processes often have been separated.The definition of the flood process is the basis for the calculation of flood frequency,which provides evidence for the flood control works.The flood frequencies have been calculated by Gumbel-Logistic model for the different flood processes in this study.The research results showed the results of the flood frequencies existed great difference under the three conditions.We depicted the second method to define the peak flood threshold(Q= 3536.67m 3 / s),which corresponding to six years of the return period by P-III type distribution.According to the second method,56 floods have been obtained.The lower frequency values of the method is more safety for design of flood control works.