Compatible Estimation of Components of Forest Growth from Remeasured Point Samples with Restricted Generalized Least Squares

Author: Lynch, Thomas B.

Source: Forest Science, Volume 41, Number 3, 1 August 1995 , pp. 611-628(18)

Publisher: Society of American Foresters

Buy & download fulltext article:

Price: $29.50 plus tax (Refund Policy)

Abstract:

Certain of the estimators which have been proposed for estimation of components of forest growth from remeasured horizontal point samples are not restricted to ensure that the initial volume estimate plus the growth estimate is equal to the final volume estimate. By viewing the initial volume, the final volume, and the components of growth as parameters in a linear model within the framework of restricted generalized least squares, estimates can be obtained which satisfy logical restrictions relating to forest growth. The model should be fitted to data obtained from a relatively large sample of points. Estimates of growth components are obtained on each point in the usual way. However, rather than simply averaging these estimates over all points, restricted generalized least squares can be used to estimate growth components in the linear model described above. Since the estimation technique can handle missing or extra points at the two measurement times, the technique can be applied to data obtained by sampling with partial replacement. For. Sci. 41(3):611-628.

Keywords: Sampling with partial replacement; permanent points; continuous forest inventory; seemingly unrelated regressions

Document Type: Journal article

Affiliations: 1: Department of Forestry, Oklahoma State University, Stillwater, OK 74078

Publication date: 1995-08-01

More about this publication?