作者: 立华 熊
关键词: Econometrics 、 Computer science 、 Frequency analysis
摘要: 水文频率分析计算是水利工程规划设计、施工以及运行管理的基础工作,传统的水文频率分析计算的一个基本前提是水文序列满足一致性假设。近几十年来,受气候变化和人类活动影响,许多河流的径流序列存在非一致性,导致传统基于一致性假设的水文频率计算方法的适用性受到严峻挑战,因此研究非一致性条件下水文频率分析方法具有重要的意义。本文在总结了国内外最新的非一致水文序列频率分析研究成果的基础上,将该研究方向的研究重点、难点和热点归纳为如下四方面:1) 单变量水文序列的非一致性诊断;2) 单变量水文序列非一致性的数学描述与归因分析;3) 非一致性条件下的单变量随机事件重现期定义和估计;4) 多变量非一致水文序列的频率分析。最后,针对这些问题,对今后的研究进行了展望。 The assumption of stationarity is a basic premise behind conventional hydrological frequency analysis for design water resources projects. Under changing environments, series in many rivers have been found to exhibit nonstationarity. As result, the methods hy-drological based on may be invalid. In recent years, fre-quency nonstationary has attracted much attention. this paper, re-searches are briefly reviewed terms four aspects: 1) detection nonstationarity univariate series; 2) mathematical description and physical attribution 3) definition calculation return period events under conditions; 4) multivariate series. Finally, some perspectives presented further de-velopment improvement analysis.