The data consists of a 3-dimensional point cloud of trees produced by a laser scanner, where each point is a sample of the tree’s surface. As for a single scanner location, a large part of the tree is occluded, multiple scans are commonly performed, and their data are co-registered (Li et al., 2020; Raumonen et al., 2015; Wan et al., 2019). It is important to not only place the points in a consistent coordinate scheme during co-registration, but also the scanner locations and to keep track of which points were produced by which scanner location for uncertainty propagation. To showcase our methodology, we applied it to terrestrial laser scanning data of 80-year-old oak trees of three different sizes (small, medium, and large). The raw point clouds were recorded at Alice Holt Forest, UK (51.1546°N, 0.8520°W) using a single-return phase shift Leica HDS- 6100 terrestrial laser scanner (Leica Geosystems, n.d.). The scans were conducted in March 2014 under dry conditions and low wind speeds (less than 2 m/s). Point clouds were acquired from six scan positions around each tree (azimuth angle: 0°, 60°, 120°, 180°, 240°, and 300°), each located 5 m from the tree base and 1.3 m above ground level. The TLS angular sampling resolution was 0.036° at each scan position, with a laser beam characterized by a 0.003 m spot size at exit, a divergence angle of about 0.013° (a 0.008 m spot size at 25 m, based on Gaussian definition). Detection and removal of scanner noise, multiple reflections, and ghost points were performed using a depth-discontinuity triangles-based method, which evaluates the angle between the local surface normal and the TLS viewing direction (Rombourg, 2019; Tang et al., 2007). The filtered point clouds were subsequently aligned within a common Cartesian coordinate system using Cyclone v9.0 (Leica Geosystems Ltd.), based on six ’6” tilt-and-turn’ reflective planar targets (Leica Geosystems Ltd.) positioned around each tree.