LiDAR-derived dataset of extracted trees and shrubs from AHN4 airborne laser scanning in the Oostvaardersplassen wetland, the Netherlands

Open Access
Authors
Publication date August 2026
Journal Data in Brief
Article number 113169
Volume | Issue number 68
Number of pages 21
Organisations
  • Faculty of Science (FNWI) - Institute for Biodiversity and Ecosystem Dynamics (IBED)
Abstract
This article describes a geospatial dataset of trees and shrubs extracted from the fourth Dutch national airborne laser scanning (ALS) survey (AHN4) covering the Oostvaardersplassen wetland in the Netherlands. The dataset was generated in the context of the MAMBO project, which focuses on automated workflows for deriving habitat condition metrics from LiDAR data for European habitat monitoring. The input data consist of nationally acquired AHN4 point clouds (2020-2022). A dedicated processing workflow was applied to classify and extract woody vegetation points in marsh and reedbed environments, where trees and shrubs co-occur with dense herbaceous vegetation.
The workflow includes (i) pre-processing and segmentation of vegetation points, (ii) clustering of tree and shrub points using neighbourhood analysis and voxel-based aggregation, and (iii) delineation of individual woody elements through seed detection and connectivity analysis in three-dimensional space. Accuracy assessments were conducted using manually created ground truth datasets derived from ALS point clouds and by referencing 8 cm resolution aerial imagery for selected validation plots and patches. The dataset includes classified ALS point clouds (LAZ), extracted tree and shrub point clouds (LAZ), derived raster products (GeoTIFF), vector boundary files (ESRI Shapefile), processing scripts, and documentation. The dataset and associated processing workflow, source code, and metadata are publicly available through Zenodo, GitHub, and the LifeWatch Metadata Catalogue. The dataset can be reused for applications such as habitat structure mapping, woody vegetation monitoring in wetlands, LiDAR-based workflow benchmarking, training and validation of classification algorithms, and upscaling of tree extraction approaches to other airborne laser scanning datasets.
Document type Article
Language English
Published at
https://doi.org/10.1016/j.dib.2026.113169 (Final published version)
Downloads
Permalink to this page
Back