皮尔逊型Ⅲ设计洪水的可靠性研究

Reliability Study on Design Floods Derived from thePearson Type Ⅲ Distribution

提出了计算皮尔逊 型设计洪水可靠性的计算公式,通过应用蒙特卡罗方法给出了数值计算结果,以此定量描述给定样本长度和设计频率下的设计洪水的可靠性。研究结果表明:统计参数的抽样误差越大,设计洪水可靠性就越低;设计洪水的重现期越大,其可靠性亦越低。因此,延长样本长度(考虑历史洪水)和区域频率分析,是提高设计洪水可靠性的两个主要途径。

The confidence interval is often used to assess the estimation error of the design floods; however, it cannot provide a clear value for defining the design floods' reliability. A formula is proposed to quantitatively calculate the reliability of design floods derived from the Pearson Type III distribution. The numerical results calculated by the Monte Carlo method are presented to assess the design flood's reliability under the given sample length and design frequency. The conclusions are as follows: the larger the sampling errors of the statistical parameters estimated from the limited sample are, the lower the reliability of design floods is; the larger the design flood's return period is, the lower its reliability is. The augmentation of sample length (i.e. including the historical floods) and the regional flood frequency analysis are considered to be the two main methods of increasing the reliability of the design floods. 