FITTING DIAMETER DISTRIBUTIONS OF TROPICAL RAINFORESTS IN VIETNAM BY FIVE PROBABILITY FUNCTIONS

Abstract

This study aims to fit diameter distribution of tropical rainforests in Vietnam and to select suitable continuous distributions. To investigate the diameter distribution of trees in tropical rainforests in Vietnam, data from 20 one – hectare plots were used. The following functions were tested: Beta, Weibull (two- and three-parameter), Lognormal, and Gamma. Parameters of distribution functions were estimated using the maximum likelihood estimation method. The best fits were selected by the Kolmogorov–Smirnov test. The results showed that the diameter distribution of the tropical rainforest was best described by the three – parameter Weibull distribution, followed by Log – normal distribution, while the Weibull 2P model fails in every case to adequately describe these frequency distributions. Estimated parameters µ, β, and α of Weibull 3P distribution ranged from 5.697 – 10.155; 1.029 – 1.606, and 6.921 – 15.234, respectively, while estimated parameters µ and  of Log – normal distribution varied from 2.804 – 2.891 and 6.921 – 15.234, respectively. The studied forests showed diameter distributions with decreasing number of trees for larger trees.