Adjustment of mathematical models for estimating above-graound biomassin in a Dense Ombrophilous Forest

Abstract

The aim of this work is to test mathematical models, 10 arithimatic and 9 logarithimic, intending to estimate individual above ground biomass in Dense Ombrophilous Forest. After it, and based on the most commonly used statistical criteria in forest science, the best model was chosen. The studied area is located in the municipalities of President Ramos and Apuína Nereu, in the State of Santa Catarina. The database for this study originally came from a research concerning the forest inventory that has as one of its aims the quantification of the biomass in the properties of the Hemmer Industry and Trade Company. From this inventory, 213 trees were sampled by the destructive method. Also, their diameters at breast height and total height were measured; the total biomass above ground was determined. Based on a correlation matrix, it was made a selection of the variables with the best correlation to be used in the tested models. The coefficient of adjusted determination R²aj, standard error of estimate (Syx%), graphic analysis of the residual waste and residual amplitude (RA) represented the criteria used to evaluate the tested models. Generally, all the tested models presented satisfactory adjustments, but the logarithimic model, ln BT = b₀ + b₁ ln dap + b₂ ln ht , with R²aj of 0,91, Syx% of 26,37, AR of 78,25 and based on the graphic analysis of the residuals, was selected to estimate total biomass.