Higher moments method for generalized Pareto distribution in flood frequency analysis

3rd International Conference on Water Resource and Environment

地球科学;生态环境科学

Zhou, C.R.^1 ; Chen, Y.F.^1 ; Huang, Q.^1 ; Gu, S.H.^2

College of Hydrology, Water Resources Hohai University, No. 1 Xikang Road, Gulou District in Nanjing, Jiangsu Province, China^1

Shanghai Hydrological General Station, No. 241 East Longhua Road, District Xuhui in Shanghai, China^2

关键词: Flood frequency analysis;    Generalized Pareto distribution;    Generalized Pareto Distributions;    Higher probability weighted moments;    Linear combinations;    Peak-over-threshold series;    Probability weighted moments;    Statistical experiments;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/82/1/012031/pdf
DOI  :  10.1088/1755-1315/82/1/012031

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

The generalized Pareto distribution (GPD) has proven to be the ideal distribution in fitting with the peak over threshold series in flood frequency analysis. Several moments-based estimators are applied to estimating the parameters of GPD. Higher linear moments (LH moments) and higher probability weighted moments (HPWM) are the linear combinations of Probability Weighted Moments (PWM). In this study, the relationship between them will be explored. A series of statistical experiments and a case study are used to compare their performances. The results show that if the same PWM are used in LH moments and HPWM methods, the parameter estimated by these two methods is unbiased. Particularly, when the same PWM are used, the PWM method (or the HPWM method when the order equals 0) shows identical results in parameter estimation with the linear Moments (L-Moments) method. Additionally, this phenomenon is significant when r ≥ 1 that the same order PWM are used in HPWM and LH moments method.