Examination of uncertainty in per unit area estimates of aboveground biomass using terrestrial LiDAR and ground data

Abstract

In estimating aboveground forest biomass (AGB), three sources of error that interact and propagate include (i) measurement error, the quality of the tree-level measurement data used as inputs for the individual-tree equations; (ii) model error, the uncertainty about the equations of the individual trees; and (iii) sampling error, the uncertainty due to having obtained a probabilistic or purposive sample, rather than a census, of the trees on a given area of forest land. Monte Carlo simulations were used to examine measurement, model, and sampling errors and to compare total uncertainty between models and between a phase-based terrestrial laser scanner (TLS) and traditional forest inventory instruments. Input variables for the equations were diameter at breast height, total tree height (defined the height from the uphill side of the tree to the tree top), and height to crown base; these were extracted from the terrestrial LiDAR data. Relative contributions for measurement, model, and sampling errors were 5%, 70%, and 25%, respectively, when using TLS, and 11%, 66%, and 23%, respectively, when using the traditional inventory measurements as inputs into the models. We conclude that the use of TLS can reduce measurement errors of AGB compared with traditional inventory measurements.