Predicting wood fiber attributes using local-scale metrics from terrestrial LiDAR data: A case study of Newfoundland conifer species

• We develop new structural metrics from terrestrial LiDAR data.
• We test the ability of these metrics to estimate wood fiber attributes.
• We identify explanatory variables using model selection and multimodel inference.
• Models estimate fibre attributes with R2 values up to 74%.
• Influential variables were species dependent.

Knowledge of wood fiber attributes (WFA) is important for evaluating forest resources and optimizing efficiency in the forest industry. To improve our ability to estimate WFA in the forest, we analyzed the relationships between structural metrics derived from terrestrial laser scanner (TLS) data and four key attributes of industrial significance: wood density, fiber length, microfibril angle, and coarseness. We developed a suite of structural metrics that relate to four aspects of the forest: canopy structure, competition, vegetation density, and local topography. We modeled WFA for sites dominated by black spruce (Picea mariana) and balsam fir (Abies balsamea) trees. For black spruce sites, R2 values ranged from 63% to 72%. Structural metrics that relate to competition were the strongest explanatory variables. For balsam fir sites, R2 ranged from 37% to 63% using structural metrics that relate mostly to canopy structure. Our results demonstrate that local structural variables are useful explanatory variables for predicting WFA of the dominant coniferous species in Newfoundland.