Standardized terrestrial laser scanning dataset of forests: 3D point clouds and tree-level attributes from 121 plots across five European regions

Forest structure strongly influences ecological processes, drives biodiversity patterns and habitat availability, and is directly linked to carbon storage. Terrestrial Laser Scanning (TLS) enables highly detailed three-dimensional characterization of forest ecosystems at the plot scale, capturing both canopy structure and understory vegetation with high spatial precision. However, many publicly available TLS datasets are limited in spatial extent or employ various acquisition protocols, complicating cross-site comparisons. This article introduces a harmonized TLS dataset comprising 121 forest plots (50 m × 50 m each) sampled across five measurement campaigns in Germany, Czechia, Spain (Galicia and La Palma), and Norway. All campaigns used identical TLS acquisition protocols, the same scanner (RIEGL VZ-400i), and a fully standardized data-processing workflow to achieve maximum comparability. For each plot, the collection provides a registered and filtered 1 cm resolution 3D point cloud (.laz), along with comprehensive tree-level structural and compositional data (including diameter at breast height, tree height, and identity as deciduous or coniferous), standardized shapefiles (plot extent, scan positions, and tree positions), and a preview image for rapid visualization. Additionally, plot-level descriptors for terrain, climate, and soil are included. The combination of a high number of plots, diverse environmental conditions, and strict methodological standardization enables robust comparative analyses of forest structural variability and cross-ecosystem comparisons of forest 3D structure. With additional reference data (e.g., labeling of the point clouds), the point clouds could be used for cross-site benchmarking of segmentation, inventory, and ecological modeling methods. It supports research on the relationships between forest structure and the environment.