多项式正态变换水文频率分析方法的稳健性研究

Robustness of polynomial normal transforms in flood frequency analysis

采用Monte Carlo模拟分析技术,研究了多项式正态变换(PNT)法的稳健性能,并与线性矩(L M)法进行了比较.结果表明:在污染样本情况下,PNT法无偏性优于L M效性略差于L M法;当污染样本比例较大时,与L M法相比,PNT法更具稳健性.因此,当样本组成复杂或总体分布难以确定时,采用PNT法估计水文变量的设计值,有望提高估计精度.

The statistical robustness of the polynomial normal transforms (PNT) was investigated by use of the Monte-Carlo simulation technique, and the comparison between the linear moment (L???M) method and the PNT method was presented. The results indicate that the PNT method is superior to the L???M method in unbiasness, but slightly inferior to the L???M method in effectiveness for contaminated samples, and that the advantage of the PNT method is more obvious to samples with larger contaminated ratio.Therefore, if samples are contaminated or population distribution is unknown, the PNT method is more likely to provide a better estimation.