Abstract: In view of the superiority of the exponential Gamma distribution in hydrological frequency analysis, application of six common parameter estimation methods adopting this distribution is discussed. Annual maximum flood peak series with historical data taken into account are calculated for the flood series recorded at six hydrological stations: Ankang, Huazhu, Linjiang, Pengshui, Shanxi and Wusheng. And statistical performance and accuracy of the design values are analyzed using Monte Carlo simulations to compare six methods: optimized curve fitting method, moment method, mixed moment method, linear moment method, maximum likelihood method, and ordinary entropy method based on the maximum entropy theory. We discuss the fitting of the exponential Gamma distribution, and use the AIC criterion and RMSE calculations to evaluate the fitting effect. The results show that the optimal curve fitting method and linear moment method achieve a satisfactory fitting effect in flood frequency analysis, and that better results of the optimal curve fitting method can be obtained when the linear moment method is used as a reference method.

Key words: exponential Gamma distribution, moment method, linear moment method, maximum likelihood method, ordinary entropy method, historical flood