Weibull function adjustment procedures for prognosing diameter distribution of eucalyptus plantations

Resumo

This study evaluated the prognosis of the diameter distribution of forest stands of Eucalyptus spp., using the Weibull probability density function, with two and three parameters and two systems of equations for forest prognosis (Systems I and II), the System I made by a larger number of stand variables. We used data of 791 permanent plots measured at ages 25-122 months. The two-parameter function was fitted by maximum likelihood methods and linear approach and the three parameters only by the maximum likelihood method. The quality of the settings of the Weibull function was assessed using the Kolmogorov-Smirnov test (α = 0.01). To adjust the equation systems were randomly selected 551 plots and to validate the data, the 240 remaining installments. The quality of the adjustments of system equations was evaluated from the distribution of residuals charts, correlation coefficient (r), square root of the mean error (RQEM), average percentage differences (MDP) and bias. The f.d.p. Weibull three parameters set by the maximum likelihood method making use of System II is the most suitable procedure for the prognosis of clonal stands of Eucalyptus urophylla and Eucalyptus grandis unthinned.