Bayesian estimation of parameters for pearson III distribution

In this paper Bayesian method was used to estimate the parameters of Pearson III distribution, the posterior distribution was calculated by the Markov Chain Monte Carlo (MCMC) approach which avoiding the difficulty of boresome integration. Metropolis-Hasting (MH) algorithm, Delayed Rejection (DR) algorithm and Adaptive Metropolis (AM) algorithm are employed for parameter sampling at the site of interest, respectively, then EM (Ensemble Mean) and ES (Ensemble Spread) were applied to evaluate convergence of the parameter sampling. The results indicate that the AM sampling is much better than the other two algorithms in convergence rate.