Seleção de equações para a estimativa de resíduos florestais para fins energéticos numa floresta tropical

Abstract

Regression equations were developed from independent variables selected among 94 tested, following the criteria of R > 0,7, which had been established variables groups for each one of the five dependent variables tested. The statistic results founded for the regression equation developed from the amount of data in 8 diametrics classes were superior to those examined when were considered the data following just one set of values. The two equation models were selected to estimate the volume of branches due to have presented the best statistics results. The models are the following: 1/VG = b0 + b1 (1/DAP2 . HF) 1/VG = b0 + b1 (1/DAP3) + b2 (l/DAP2 HF).