论文部分内容阅读
极端洪水事件的频率分析往往局限于单个站点,当研究区域内包含多个水文站点时,单变量频率分析方法,会导致低估或高估洪灾风险率。因此,需要进行区域频率分析。传统区域重现期计算方法,同一重现期对应多种设计洪水组合,而基于Kendall分布函数的重现期计算方法(KRP)有效的解决了这一问题。故本文引入三维非对称Copula分布函数拟合区域内各个站点年最大流量的相关关系,利用半参数法估计Copula函数的参数,并采用KRP推求区域洪水发生的重现期。结果表明:区域发生T年一遇的洪水概率远远大于单个站点发生T年一遇的洪水概率;KRP克服了实测序列较短的问题,且能准确估算洪水重现期。本研究为防洪部门制定防洪措施提供一定的科学依据。
Frequency analyzes of extreme flood events are often confined to a single site. Univariate frequency analysis methods can lead to underestimation or overestimation of flood risk rates when multiple hydrological stations are included in the study area. Therefore, the need for regional frequency analysis. The calculation method of traditional regional recurrence period corresponds to multiple design flood combinations in the same recurrence period, and the KRP based on Kendall distribution function effectively solves this problem. In this paper, a three-dimensional non-symmetric Copula distribution function is introduced to fit the annual maximum flux of each site in the region. The semi-parametric method is used to estimate the Copula function parameters, and the KRP is used to infer the recurrence of regional floods. The results show that the probability of flooding once a year in a region is far greater than that of a single site in year T; KRP overcomes the problem of short measuring sequence and can accurately estimate the flood recurrence period. This study provides some scientific evidence for flood control department to make flood control measures.