论文部分内容阅读
现行洪峰频率计算的统计理论基础在于洪峰流量为独立随机变量．然而目前出现一种议论，认为洪峰序列有些情况下虽然无短滞时相关结构，但却存在着长滞时相关结构．这种奇特的特性可称做统计混乱性．混乱性无疑增加了估计参数的不确定性，并严重影响洪水风险分析的可靠性．本文根据中国部分河流长期观测资料，分析了洪峰序列的统计变化特性．结果表明：洪峰序列短滞时相关结构特性不明显，统计混乱性也不明显，在目前情况下可不予考虑．
The statistical theory of the current peak frequency calculation is based on that the peak flow is an independent random variable. However, there is an argument that there is a long-lag correlation structure in some cases of the flood peak sequence, although there are no short-delay related structures. This peculiar feature can be called statistical confusion. Confusion undoubtedly increases the uncertainty of the estimated parameters and seriously affects the reliability of flood risk analysis. Based on the long-term observation data of some rivers in China, this paper analyzes the statistical variation characteristics of flood peak sequences. The results show that the structural characteristics of short-time correlations of flood peak sequences are not obvious, and the statistical confusion is insignificant, which can not be considered in the current situation.