非参数核估计方法在洪水频率分析中的应用

NONPARAMETRIC KERNEL ESTIMATION OF FLOOD FREQUENCIES

目前,在洪水频率分析中,一般事先假定实测洪水系列服从某一线型,例如,服从P-Ⅲ型分布,再用参数统计方法分析洪水频率。然而,这种假定在多数情况下是不合理的,甚至会产生严重的后果。本文用非参数核估计方法对洪水系列分布的概率密度进行了估计,结果显示,非参数核估计法不仅具有事先不需要假定总体的线型的优点,而且估计的结果与参数统计法相比具有较高的精度。

A current used approach to flood frequency analysis is based on the concept of parametric statistical inference. In this analysis the assumption is mad that the distribution function describing flood date is known, for example a Pearson type - Ⅲ distribution. However, such an assumption is not always justified and often lead to other difficulties . A new method is developed in this article based on the nonparametric procedure for estimating probability distribution function. The results indicate that the nonparametric kernel estimation of flood frequeruies is a viable alternative with the advantage of not requiring a distributional assumption , and also has the better result.