Prediction of Wood Fiber Attributes from LiDAR-Derived Forest Canopy Indicators

We investigated the potential use of airborne light detection and ranging (LiDAR) data to predict key wood fiber properties from extrinsic indicators in lodgepole pine leading forest stands located in the foothills of central Alberta, Canada. Six wood fiber attributes (wood density, cell perimeter, cell coarseness, mature fiber length, microfibril angle, and modulus of elasticity) were measured at 21 plots, and with use of data reduction techniques, two components of wood properties were derived: wood strength, stiffness, and fiber yield and fiber strength and smoothness. These wood fiber components were then compared with extrinsic indicators of wood characteristic-derived LiDAR-estimated topographic morphology, tree height, and canopy light metrics. The first principal component indicating wood strength and stiffness was significantly correlated to the depth of different canopy zones (or light regimes; r(2) = 0.55, P < 0.05). The second component, related to fiber strength and smoothness, was significantly correlated to the height of the canopy and canopy thickness (r(2) = 0.65, P < 0.05). The results indicate that airborne LiDAR attributes can explain about half of the observed variance in intrinsic wood fiber attributes, which is approximately 5-10% less than that explained by growth-related field-measured variables such as diameter increment and height. This reduction in explained variance can be balanced by the opportunities for much broader spatial characterizations of wood quantity and quality at the stand and landscape levels. FOR. SCI. 59(2):231-242.