Base-age invariance properties of two techniques for estimating the parameters of site index models

Abstract

Using eight hypothetical data points (four growth series with two observations each) the difference between two site index curve parameter estimation methods, the base-age invariant stochastic regression approach (BAI) and the method of all base ages (BAA), is illustrated. The anamorphic form of the Schumacher log-transform height-age equation is used for the comparison. The two estimation techniques are equivalent for the linear case of the model but produce different parameter estimates for its nonlinear form. For both linear and nonlinear cases the BAI approach produced a lower root-mean-squared error than the BAA method. Only the BAI approach produced unique estimates for the site-specific parameters of both model forms and both growth series. Copyright © 2006 by the Society of American Foresters.