论文部分内容阅读
洪水的统计分析是假设年最大洪峰流量为来自随机水文过程的随机变量,由于实测洪水系列仅是其总体的一个样本,因此其参数不一定等于总体的参数。本文将研究估算实测洪水样本系列参数的方法,并提出确定置信区间的方法。 如果将实测资料点绘在概率纸上,一定置信区间的上、下界线总是落在回归线的两边(图1略)。例如,置信度为90％,将意味着对一定的重现期(或概率)而言,回归线将各有5％的可能性超出上、下界线。
The statistical analysis of floods assumes that the annual maximum flood flux is a random variable from a random hydrological process. Since the measured flood series is only one sample of its population, its parameters are not necessarily equal to the overall parameters. This article will study methods for estimating series of measured flood samples and propose methods to determine confidence intervals. If the measured data points are plotted on probability paper, the upper and lower bounds of the confidence interval always fall on either side of the regression line (Figure 1 omitted). For example, a confidence level of 90% would mean that for certain recurrences (or probabilities), the likelihood that the regression line will have 5% each goes beyond the upper and lower bounds.