Assessing the Three-dimensional Frequency Distribution of Airborne and Ground-based Lidar Data for Red Pine and Mixed Deciduous Forest Plots

The following study reports on the preliminary results of a novel processing technique for characterizing the three-dimensional distribution of airborne and ground-based LiDAR data for conifer and mixed deciduous forest plots. The frequency of laser returns within 1m x 1m x 1m cubes (voxels) and the plotting of percentile distributions within each column (e.g. 1m x 1m x 25m, x, y, z) may benefit radiative transfer modelling for meteorological and carbon flux monitoring purposes. This study examines the voxel column percentile distributions at the 1m and 5m scales and at the plot level (three-dimensional percentile distributions) for both airborne and ground-based LiDAR systems. Here it is demonstrated that for airborne LiDAR, a high percentage of laser pulses intercept the top of the canopy with fewer returns from within the canopy and understory. Similarly, for ground-based LiDAR, a high percentage of laser pulses intercept the understory, stems, and lower canopy, with a lower percentage of pulses intercepting the upper canopy. AL excludes an average of approximately 4.0m from the centre (17m a.g.l.) to the base of the canopy (at 15m a.g.l.), whereas GBL excludes the top 21m to 24.5m of the canopy. This illustrates that substantial parts of the understory and canopy are excluded from airborne and ground-based LiDAR data, respectively. Suggestions are provided to best model the percentile distribution of laser pulses for both airborne and ground-based systems based on the scanner field of view, scan angles, and the distance from the centre of individual voxels.