Premium
Comparison and analysis of design floods by the change in the order of LH‐moment methods
Author(s) -
Lee Soon H.,
Maeng Sung J.
Publication year - 2003
Publication title -
irrigation and drainage
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.421
H-Index - 38
eISSN - 1531-0361
pISSN - 1531-0353
DOI - 10.1002/ird.91
Subject(s) - l moment , generalized pareto distribution , moment (physics) , generalized extreme value distribution , statistics , mathematics , watershed , flood myth , outlier , order statistic , extreme value theory , geography , computer science , physics , archaeology , classical mechanics , machine learning
This research seeks to derive the design floods through the appropriate order of LH‐moments with the test of homogeneity, independence and outlier of data on annual maximum floods in six Korean watersheds. To select the appropriate probability distribution for annual maximum flood data according to watershed, we applied the generalized extreme value (GEV), generalized logistic (GLO) and generalized Pareto (GPA) distributions. We arranged L, L1, L2, L3 and L4‐moment method theories by the change in the order of LH‐moment (parameter estimation method) according to three probability distributions, and presented LH‐moment ratio diagrams, i.e. L, L1, L2, L3 and L4‐moment ratio diagrams. We judged the suitability of the applied GEV, GLO and GPA distributions through the LH‐moment ratio diagrams and the Kolmogorov–Smirnov (K–S) test. We derived the parameter and design flood of the appropriate distribution using L, L1, L2, L3 and L4‐moment methods, targeting the observed and the simulated annual maximum flood through Monte Carlo techniques. By comparing and analyzing design floods derived through LH‐moment methods and the appropriate distribution according to watershed, we presented the appropriate order of LH‐moments that can derive appropriate design floods. Copyright © 2003 John Wiley & Sons, Ltd.