This dataset presents information about 48 sites in the boreal forests of Alaska and Northwest Canada. At each site, we established a UAV transect and collected LiDAR and RGB data with a Yellowscan Mapper mounted on a DJI M300. The data was collected in the summers of 2022 in Canada and 2023 and 2024 in Alaska. Each LiDAR transect was processed into a colored pointcloud, and individual trees were segmented. We computed tree characteristics: tree height, crown area, crown diameter, and estimated age (based on tree core data and a regression model). Using a random forest algorithm, we classified the individual trees into 2 categories: evergreen (Spruce, Pine, etc.) and deciduous (Betula, Aspen, etc.). The pointcloud were aggregated into 20 x 20 meters tree patches. We computed the abundance of evergreen and decidous categories in each patch per height category (below 5 meters, between 5 and 12, above 12 meters). This data is presented as individual table per site. This dataset supports the findings of Enguehard et al., 2025.