考虑高水影响的洪水频率分布线型对比研究  

A comparison study on distribution for flood frequency analysis focus on the observed flow data in the high flow part

作　　者：叶长青[1,2] 陈晓宏[1,2] 邵全喜 张家鸣[1,2] 

机构地区：[1]中山大学水资源与环境研究中心,广东广州510275 [2]华南地区水循环与水安全广东省普通高校重点实验室,广东广州510275 [3]澳大利亚联邦科学与工业组织数学与信息科学研究所

出　　处：《水利学报》2013年第6期694-702,共9页Journal of Hydraulic Engineering

基　　金：国家自然科学基金重点项目(50839005);国家自然科学基金项目(51209095);国家重点基础研究发展计划(973)项目(2010CB428405);广东省科技厅项目(2010B050300010);广东省水利科技创新项目(2009-39);中英瑞气候变化适应项目(ACCC/20100705-1);广东省自然科学基金项目(S2011010001549)

摘　　要：洪水极值事件在全球范围频频发生,准确估计设计流量对降低洪水灾害十分重要。洪水频率分析已广泛应用于水文设计领域,有很多被推荐的频率分布,还包括很多选择分布的准则。在北纬23.5°—66.5°之间选取4条河流——泰晤士河、沃巴什河、北江和淮河,采用极大似然法估参,并以3种分布选择准则对8种频率分布进行拟合优选。为减少可能的分布线型数量,不仅总结对比100%实测年最大日流量与分布的拟合结果,而且对在洪水频率分析中起重要作用的50%高水端实测年最大日流量进行分布拟合对比。结果表明:(1)适合100%实测流量与50%高水实测流量的最优分布基本一致但存在差别,适合50%高水流量的分布与大洪水点据拟合得更好;(2)综合考虑100%与50%高水实测流量选择最优分布,泰晤士河为Gumbel,沃巴什河是GEV和P3,北江为Weibull和GEV,淮河为Weibull;(3)为满足工程设计洪水要求,选择最优频率分布时,应综合考虑既适合100%又适合50%高水实测序列的共同分布,以减少设计洪水的不确定性。Extreme flood events are becoming increasingly populated over the world. Better estimation of an- nual maximum daily flow is very important for flood hazard mitigation. Flood frequency analysis is widely used in hydrologic design. Many distributions have been proposed, several criteria have been used for select- ing the distributions. In this paper, by using four typical data sets from Thames River, Wabash River, Bei- jiang River and Huaihe River, which are all located within north latitude of 23.5-66.5 degrees, we exam- ined the performance of 3 distribution selection criteria for 8 popularly used distributions. In order to limit the number of potential distributions, we summarized the comparison results based on not only the overall fitting but also the upper 50% observed data which is trulyan interest in the flood frequency analysis. The results indicate that （1） the optimal frequency distributions for 50% high flow are basically the same as probability distributions selected by using 100% all the data in the region, but there are still differences, the fitting degree of optimal frequency distributions for the 50% high flow are better than 100% all the measured data; （2） probability distributions are characterized by Gumbel in Thames River, while GEV and P3 best fit the data at stations in Wabash River, Weibull and GEV in Beijiang River and Weibull in Huai- he River; （3） in order to meet the requirements of engineering design flood and reduce the uncertainty of design flood, the shared distribution should be considered in combination with 100% the measured data and 50% high flow observed data when choose the optimal distribution.

关 键 词：分布选择 洪水频率分析 拟合检验 高水尾端 准则 

分 类 号：P333[天文地球—水文科学]