ACCURACY OF TAPER EQUATIONS TO REPRESENT STEM PROFILES OF Pinus elliottii

Abstract

This study had the objective of evaluating the  accuracy of volume ratios, polynomial models and cubic spline functions for estimating diameter at different heights along the tree stem, and for different diameter classes. The data came from Pinus elliottii plantations in the northeastern region of Paraná State. The accuracy of the five tested models was evaluated through the following statistics: coefficient of determination; standard error of the estimate; diameter deviations at all “i” positions along the stem; standard deviation of the diameter differences; sum of squares of relative residuals; residual percentage of the diameters. Basing on these statistics, a ranking was prepared for each relative position of diameter measurement along the stem and by diameter class, aiming to express, in a summarized way, the accuracy of the five tested models. As main results it can be pointed out that: the cubic spline functions, and the  Clutter taper equation were not recommended for estimating diameters along the stem of Pinus elliottii  in this region, because they propitiate innacurate estimates; the Amateis and Burkhart taper equation, and the polynominal functions presented accurate diameter estimates along the tree stem, from the first standard log (2,2 or 2,4 m). If more uniform diameter estimates along the stem are desired, associated to accuracy of the estimates, the use of the Amateis and Burkhart equations, is recommended followed by the fifth degree polynomial, and by the fractionaire power polynomial. If the desired equation have to propitiate a greatter number of cases with accurate diameter estimate, but without uniformity along the stem, then the fractionnaire power polynomial is recommended, followed by the Amateis and Burkhart taper equation, and by the fifth degree polynomial.