A test of alternative estimators for volume at time 1 from remeasured point samples
Abstract
Two estimators for volume at time 1 for use with permanent horizontal point samples are evaluated. One estimator, used traditionally, uses only the trees sampled at time 1, while the second estimator, originally presented by Roesch and coauthors (F.A. Roesch, Jr., E.J. Green, and C.T. Scott. 1989. For. Sci. 35(2):281-293). takes advantage of additional sample information that becomes available at time 2. In this test the Roesch et al. estimator was always lower in sum of squared differences and sum of absolute differences for board foot and cubic foot volume than the traditional estimator. In addition a simulation revealed no discernible bias for the Roescher et al. estimator over the traditional estimator.