Country-wide data of ecosystem structure from the third Dutch airborne laser scanning survey

• DOI: 10.1016/j.dib.2022.108798
• Publication Date: 2022-12-05T06:37:03Z
• Authors: W. Daniel Kissling; Yifang Shi; Zsófia Koma; Christiaan Meijer; Ou Ku; Francesco Nattino; Arie C. Seijmonsb...; Meiert W. Grootes
• Keywords: Ranging; Laser Scanning

The third Dutch national airborne laser scanning flight campaign (AHN3, Actueel Hoogtebestand Nederland) conducted between 2014 and 2019 during the leaf-off season (October-April) across whole Netherlands provides a free open-access, country-wide dataset with ∼700 billion points point density of ∼10(-20) points/m2. AHN3 cloud was obtained Light Detection And Ranging (LiDAR) technology contains for each x, y, z coordinates additional characteristics (e.g. return number, intensity value, scan angle rank GPS time). Moreover, has been pre-processed by 'Rijkswaterstraat' (the executive agency Ministry Infrastructure Water Management), comes Digital Terrain Model (DTM) Surface (DSM), is delivered pre-classification into one six classes (0: Never Classified, 1: Unclassified, 2: Ground, 6: Building, 9: Water, 26: Reserved [bridges etc.]). However, no detailed information on vegetation structure available from cloud. We processed (∼16 TB uncompressed data volume) 10 m resolution raster layers ecosystem at extent, using novel high-throughput workflow called 'Laserfarm' cluster virtual machines fast central processing units, high memory nodes associated big storage managing large amount files. (available as GeoTIFF files) capture 25 LiDAR metrics structure, including height 95th percentiles normalized z), cover pulse penetration ratio, canopy cover, within defined layers), structural complexity skewness variability vertical distribution). make use projected coordinate system (EPSG:28992 Amersfoort / RD New), are ∼1 GB in size, can be readily used ecologists geographic (GIS) or analytical open-source software such R Python. Even though class '1: Unclassified' mainly includes points, other objects cars, fences, boats also present this class, introducing potential biases derived products. therefore validated >180,000 hand-labelled 100 randomly selected sample plots (10 × each) Netherlands. Besides vegetation, boats, cars were identified sampled plots. misclassification rate (i.e. non-vegetation that assumed to vegetation) low (∼0.05) accuracy (∼90%). To minimize existing inaccuracies product ships water bodies, chimneys roofs, roads might incorrectly points), we provide an mask captures buildings generated cadaster dataset. This newly new opportunities ecology biodiversity science, e.g. mapping 3D variety ecosystems modelling biodiversity, species distributions, abundance ecological niches animals their habitats.