Detection of sub-canopy forest structure using airborne LiDAR

Knowledge on forest structure is vital for sustainable forest management decisions. Currently, Airborne Laser Scanning (ALS) has been well established as an effective tool to delineate and characterize the structure of canopies across a range of forested biomes. However, the use of ALS to provide information on sub-canopy structure is less well developed. Sub-canopy structure consists of suppressed mature trees, regenerating tree saplings, shrubs, herbs, snags and coarse-woody-debris. With the increasing density of ALS point clouds, new opportunities exist to describe these sub-canopy structural components in forests that were previously difficult to detect using passive remote sensing technologies. In this research we use discrete return ALS data acquired at a density of 23 points x m(2) to estimate sub-canopy forest structure for 48,000 ha of conifer dominated forest in central British Columbia, Canada. We first segmented the forest vertical structure into canopy and sub-canopy based on Lorey's mean height (HL). HL, which favours larger trees as a baseline for canopy height, provides separation between the taller trees that dominate the canopy and smaller trees that represent the sub-canopy. We defined sub-canopy trees as those < 70% of HL. Both ground-truthed forest inventory data and the ALS point cloud were then segmented into canopy and sub-canopy components. A mixture of standard height-based and density-based ALS metrics were then computed to develop predictive models of sub-canopy component of the stands. Models were calibrated with 28 ground plots and developed using stepwise regression with the strongest predictors being a combination of height, structure and cover-based metrics. Two model sets were developed, one for the entire point cloud, and another for an isolated sub-canopy point cloud defined by HL. The isolated sub-canopy set of models resulted in stronger cross-validated R-squared values of 0.88, 0.68, and 0.55 for tree volume, basal area and number of sub-canopy trees, respectively. We then applied these models over the entire study area to characterize the sub-canopy structure, resulting inventory can be used by land managers for a number of purposes including selecting candidate locations for selective logging to preserve mid-term timber opportunities, fire susceptibility and carbon sequestration modelling, and wildlife habitat values.