PREDIÇÃO DOS PARÂMETROS COMO ALTERNATIVA PARA OBTENÇÃO DE VOLUME DE MADEIRA EM FLORESTAS PLANTADAS

Abstract

The study aimed to evaluate the application of the parameter prediction technique in the 'Schöpfer' tapering function, using dendrometric data from sample units as input variables. The study region is located in the “Planalto Norte Catarinense, Santa Catarina” region, Brazil. Data from the continuous inventory carried out in 2018 were used, covering forest stands with ages ranging from 5 to 16 years and information on the destructive cubage of 130 Pinus taeda L. trees, sampled using the Hohenadl methodology, stratified by diameter class, age, density, and forest site. The sampling units used in the adjustment were selected by Euclidean distance, and the dendrometric variables of these units served as inputs variables in the model1 suggested by the author. For comparative purposes, the tapering function was adjusted in four different age classes, with an amplitude of 2.5 years, in this work denominated as conventional adjustment. For the development of the Parameter Prediction technique, 29 sample units were used, with reference to cubed trees. The adjusted models were statistically evaluated by the standard error of the estimate (Syx%), adjusted termination coefficient (R²ajt), and residual dispersion (Res%), in addition to quantitative comparisons with the observed values of volume, diameter along the bole, and form factor, as well as graphical assessments of tree shape. The results indicate that both the conventional fit and the parameter prediction technique are adequate and accurate for estimating diameters, volume, and quantifying forest assortment. The proposed model has limitations regarding the use of the stepwise technique, but it is capable of generating a thinning equation per tree and precisely describing the thinning of trees by calculating coefficients using dendrometric variables.