Tree species identification is important for a variety of natural resource management and monitoring activities including riparian buffer characterization, wildfire risk assessment, biodiversity monitoring, and wildlife habitat improvement. Coordinate data from airborne laser scanners can be used to detect individual trees and characterize forest biophysical attributes. Metrics computed from LIDAR point data describe tree size and crown shape characteristics. The intensity data recorded for each laser point is related to the spectral reflectance of the target material and thus may be useful for differentiating materials and ultimately tree species. The aim of this study is to test if LIDAR intensity data and crown structure metrics can be used to differentiate tree species. Leaf-on and leaf-off LIDAR were obtained in the Washington Park Arboretum. Field work was conducted to measure tree locations, heights, crown base heights, and crown diameters for eight broadleaved species and seven conifers. LIDAR points from individual trees were identified using the field-measured tree location. Points from adjacent trees were excluded. We found that intensity values for different tree species varied depending on foliage characteristics, the presence or absence of foliage, and the position of the LIDAR return within the tree crown. In terms of the intensity analysis, the classification accuracy for broadleaved and coniferous species was better using leaf-off data than using leaf-on data while in terms of the structure analysis, the accuracy was better using leaf-on data than using leaf-off data. The stepwise cluster analysis was conducted to find similar groups of species at consecutive steps using k-medoid algorithm. When using both LIDAR datasets showed the most reasonable clustering result compared with the result using either one of the datasets.

The research presented in this dissertation provides a significant contribution to the understanding of how various tree species can be identified through the structural and spectral characteristics derived from LIDAR data.