Comparing trigonometric variableform taper function with generalized Brink stem profile function

Using sample taper data collected from Pinus radiata plantations at different sites of New South Wales State, Australia, this paper compares trigonometric variableform taper function (Bi function) with generalized Brink stem profile function (GB function) when estimating diameter under bark at any height of stem. Nonlinear least square regression was used to estimate parameters of both functions. The residual errors, mean square error of prediction, mean absolute difference of prediction and determination coefficient were chosen as the criteria for evaluation of the fitting effective. Results showed that after the parameters in the two taper functions were obtained by regression, Bi function performed more accurately than GB function in predicting diameter under bark at any height of stems; the Bi function was superior to GB function on bottom stem fitting effect, and the Bi function was overall stabler than GB function. In addition, based on parameter estimating results, the Bi function is simplified and the fitting effect of the simplified taper function has been validated, which verifies the deformation and flexibility of Bi function.