Characterization of diverse plant communities in Aspen Parkland rangeland using LiDAR data

Question: How effective is high-resolution airborne LiDAR technology for quantifying biophysical characteristics of multiple community types within diverse rangeland environments? Location: Native Aspen Parkland vegetation in central Alberta, Canada. Methods: Vegetation within 117 reference plots stratified across eight types, including forest, shrubland, upland grassland and lowland meadow communities, were assessed in 2001 for the height, cover and density of vegetation within various strata (herb, shrub and tree layers). Actual ground data were subsequently compared against modelled values for each community type and strata derived from the analysis of airborne LiDAR data obtained in 2000. Results: LiDAR data were effective for quantifying vegetation height, cover and density of the overstory within closed- and open Populus forest communities. However, LiDAR measurements typically underestimated the height and cover of shrublands, as well as most of the herbaceous communities. Analysis of LiDAR intensity data indicated reflectance generally decreased as LiDAR sampling points moved upwards from the ground to the vegetation canopy. Conclusions: While LiDAR technology is useful for characterizing deciduous forest properties, the quantification of understory vegetation characteristics, as well as those of individual shrublands and grasslands, was more limiting. Further refinements in analysis methods are necessary to increase the reliability of characterizing these communities.