Estimates of Tree Biomass, and Its Uncertainties Through Mean-of-Ratios, Ratio-of-Means, and Regression Estimators in Double Sampling: A Comparative Study of Mecrusse Woodlands

Abstract

Frequently, biomass expansion factors (BEFs), the respective biomass densities, and their uncertainties are computed without taking into account the appropriate estimators. The objective of this study was to compare the estimates of BEF, BEF-based biomass densities, and their uncertainties using different estimators (mean-of-ratios, ratio-of-means, and regression estimators) in double sampling. Our results demonstrated that increased uncertainty is associated with regression-based biomass densities, and that the computation of BEF using merchantable timber volume should utilize regression estimators, not the usual ratio estimators, which preferably, should be avoided altogether, as they are found to be subjective and more susceptible to errors and personal judgment.