Datasets for testing the robustness of LiDAR vegetation metrics to varying point densities

The calculation of vegetation metrics from LiDAR point clouds might be affected by the available point density of a dataset. Testing how the same LiDAR vegetation metrics differ with different point densities can therefore inform about their robustness for upscaling metrics to other areas or other LiDAR point clouds. The datasets made available here were generated to test the robustness of LiDAR vegetation metrics to varying point densities and spatial resolutions (i.e., plots of 1 × 1 m, 2 × 2 m, 5 × 5 m and 10 × 10 m size). A total of 25 LiDAR vegetation metrics representing different aspects of vegetation height, vegetation cover and structural complexity were tested (see metric definition in Kissling et al. 2023, https://doi.org/10.1016/j.dib.2022.108798). The metric calculation was similar to the metric calculation in the Laserchicken software (Meijer et al. 2020, https://doi.org/10.1016/j.softx.2020.100626) and the Laserfarm workflow (Kissling et al. 2022, https://doi.org/10.1016/j.ecoinf.2022.101836). The Dutch AHN4 dataset from the years 2020–2022 with a point density of 20–30 points/m² was used. Initially, 100 plots (i.e., squared polygons around centre points) were randomly placed across the Netherlands in Dutch Natura 2000 sites that predominantly contain woodland habitats (using shapefiles from the European Environmental Agency). For each centre point, square polygons of the desired resolutions (i.e., 1 × 1 m, 2 × 2 m, 5 × 5 m or 10 × 10 m plot size) were generated. The square polygons were subsequently used to clip the LiDAR point clouds from the Dutch AHN4 point cloud dataset. Since not all locations of the 100 randomly placed plots contained points, the actual sample sizes were slightly smaller than 100, i.e., 94 plots for the 1 × 1 m, 2 × 2 m and 5 × 5 m resolution and 95 plots for the 10 × 10 m resolution. Metrics were calculated with the original point density of the Dutch AHN4 dataset (20–30 points/m2) and with six systematically down-sampled point clouds for the same plots (i.e., keeping 5%, 10%, 20%, 40%, 60% and 80% of the points in the original point clouds). For each clipped point cloud of a plot at a given resolution, the points were first sorted according to their GPS acquisition time (from earliest to latest). Points were then systematically discarded and only 5%, 10%, 20%, 40%, 60% and 80% of the points in the original point clouds were kept. The kept points were used for calculating the 25 LiDAR vegetation metrics.