Author(s): Hongjoon; Shin; Kewtae; Kim; Chang-Sam; Jeong; Jun-Haeng; Heo
Linked Author(s): Kuk-jin KIM, Woo Chang Jeong, Sung-woo SHIN
Keywords: Generalized logistic distribution; Sampling variance; Maximum likelihood
Abstract: The generalized logistic (GL) distribution is recently recommended for flood frequency analysis in the UK instead of GEV model. However, there has been little study of the asymptotic sampling variance of the GL distribution using the method of maximum likelihood. In this paper, the estimation of quantile for the GL distribution is presented based on the method of maximum likelihood (ML) and the asymptotic variance of quantile estimator is derived as functions of the sample sizes, return periods, and parameters. Monte Carlo simulation experiments are also performed to verify the applicability of the derived sampling variances of quantile. As the results, the analytical expression is not suitable to apply except for small return period in case of small sample size. And the results of simulation close to analytical solution as sample size increases. The simulation can only be evaluated for 0. 5 0. 5 and the difference between analytical expression and results of simulation deteriorates regardless of sample size for shape parameter closes to 0. 5±. The difference increases especially for negative shape parameter. However, even for small sample size, the approximation perform relatively well in case of 0. 2 0. 2. β < + − < − < β < +
Year: 2007