Résumé
A model comparison approach, based on the Akaike’s information criterion, was used to evaluate the contribution of terrestrial laser scanning (TLS) to the estimation of wood fiber attributes at the tree level for black spruce (Picea mariana (Mill.) B.S.P.) trees growing in Newfoundland, Canada. Substantial efforts were made to acquire, process, and develop accurate and detailed metrics of the tree, its crown, and its immediate environment. Based on the resulting data set, significant relationships were found, and models were successfully developed, using only TLS metrics, for predicting wood fiber attributes. The models accounted for 47%, 33%, 51%, 44%, and 52% of variance in wood density, coarseness, fiber length, microfibril angle, and modulus of elasticity respectively, with root mean square error values of 46 kg/m3, 37 µg/m, 0.20 mm, 3.5 º, 2.3 GPa. Our ability to estimate the wood fiber attributes was improved by combining TLS metrics with conventional field measurements. This study demonstrates that the use of TLS metrics improves the estimation of the wood fiber attributes at the tree level beyond that possible with conventional field measurements.