Table 1. Research on Distributions and Estimation Methods (1991-1994) ------------------------------------------------------------------------------------- Distribution Estimation method Sampling statistics Reference -------------------------------------------------------------------------------------
P3 MOM/GS Var, Conf. Intv. Chowdhury and Stedinger (1991); Horlacher (1992)
EV1 MOM/HIS Bias, RMSE Guo and Cunnane (1991)
EV1, EV2, EV3 ML Bias Koch (1991)
EV1, LN REF, TEA, MOM -- Lind and Hong (1991)
P3 PWM/LB Bias, RMSE Wu et al. (1991)
GP, GP3 PWM Bias, Var, RMSE Wang (1991)
LN,LP3,GEV,GN ML, MOM, PWM Bias Durrans (1992a)
P3 MOM/LB, MOM Var Durrans (1992b)
LN, EV1, LEV1, G, N, GEV, P3, LP3, WAK ML, MOM, PWM Exceedance probability Gunasekara and Cunnane (1992)
GEV, EV1 PWM Bias, Var, RMSE Lu and Stedinger (1992a,b)
GP PWM, MOM RMSE Rosbjerg et al. (1992)
LN3 ML/HIS Bias, Var Stevens (1992)
LP3 MOM/T, SAM/T RMSE Ashkar et al (1993)
LN, LN3 EV1, LL, P3, LP3, GEV, WAK, LB ML, PWM, MOM Bias, Var Haktanir (1992); Haktanir and Horlacher (1993)
Halphen ML -- Bobée et al. (1994)
G, EV1, N/Box-Cox ML Var Martins and Clarke (1993)
GP ML, MOM, PWM, LS Bias, RMSE Moharram et al. (1993)
LP3 ML/HIS Bias, Var Pilon and Adamowski (1993)
LL ME, ML, PWM, MOM Bias, RMSE Singh et al. (1993)
P3 MOM/LB, ML/LB Bias, Var, RMSE Rasmussen et al. (1994) -------------------------------------------------------------------------------------------------
Distributions: G: Gamma; P3: Pearson type 3; LP3: Log-Pearson type 3; EV1: Extreme value type I; EV2: Extreme value type II; EV3: Extreme value type III; GEV: Generalized extreme value; LEV1: Log-Gumbel; GP: Generalized Pareto; GP3: 3- parameter generalized Pareto; N: Normal; N/Box-Cox: Box-Cox transformation to normality; LN: Lognormal; LN3: 3- parameter lognormal; GN: Generalized normal; WAK: Wakeby; LL: Log-logistic; LB: Log-Boughton
Estimation methods: MOM: method of moments; ML: Method of maximum likelihood; PWM: Method of probability weighted moments; SAM: Sundry averages method; ME: Method of maximum entropy; REF: Method of relative entropy with fractile constraints; TEA: Tail entropy approximation; LS: Least square; /GS: Generalized skew; /HIS: Historical information; /LB: lower bound estimated separately; /T: Trimmed or censored sample;