This study aimed to test several traditional and generic hypsometric models, analyzing the adjustments and behavior in different groupings of independent variables that characterize a forest stand. Traditional and generic hypsometric models were determined, the most exact equation was selected through the determination coefficient and residual standard error. To verify whether equations selected for each situation are different, a complete randomized experimental design in split plot arrangement was adopted. The Scott & Knott mean test was applied for the cases where it was detected significant difference in the variance analysis were detected. It was verified that the adjustment per plot using traditional models is the ideal procedure to estimate tree height. However, the adjustment of the generic model gave good estimates, indicating the possibility of its use in substitution to the traditional models.
Regression models; height; eucalyptus
