Abstract This dataset provides an early access version of the European tree genus map at 10 m resolution for the year 2020, derived from Sentinel-1 and Sentinel-2 satellite data. The map distinguishes eight classes (Larix, Picea, Pinus, Fagus, Quercus, other needleleaf, other broadleaf, and no trees) and is distributed as Cloud Optimized GeoTIFFs (COGs) over a 100 km grid in EPSG:3035 (ETRS89 / LAEA Europe). The map was generated using a CatBoost model trained on forest plot inventories, citizen science observations, orthophoto interpretation, and LUCAS data, with additional features from DEM and climate datasets. Labels were filtered and aggregated to genus level to reduce noise. Early access notice This release is provided as an early access version. The map is still undergoing validation and fine-tuning, and a formal publication is planned. Updates and improvements may therefore be made in future releases. We welcome feedback and contributions of additional training data to further improve the map. Dataset description Resolution: 10m Format: Cloud Optimized GeoTIFFs (COGs) Tiling: 100km grid Coordinate reference system: EPSG: 3035 (ETRS89 / LAEA Europe) Legend 0 – Larix 1 – Picea 2 – Pinus 3 – Fagus 4 – Quercus 5 – Other needleleaf 6 – Other broadleaf 7 – No trees Methodology summary The classification was performed using a CatBoost model trained on diverse reference sources [1-10]: - National and regional plot inventories - Citizen science observations - Orthophoto interpretation - LUCAS data Training labels were filtered to reduce noise and aggregated to genus level. Predictor variables include annual statistics from Sentinel-1 and Sentinel-2, combined with auxiliary datasets on altitude (DEM) and climate. Further details on the methodology will be made available in the product publication, which will follow this early access release. Usage Notes CRS: EPSG:3035 (ETRS89 / LAEA Europe). Reprojection may be required for use with other datasets. Tiling scheme: Provided as 100 km × 100 km COG tiles. Users may mosaic tiles if needed. Classes: See legend above. Class 7 (“No trees”) includes cropland, grassland, built-up, and other non-tree areas. Early access status: Not yet fully validated. Regional inconsistencies and misclassifications may be present. Feedback & contributions: We invite users to share validation results and contribute additional reference data to improve future releases. How to cite If you use this dataset, please cite as: De Keersmaecker, W., Zanaga, D., Senf, C., Viana-Soto, A., Klapper, J., Blickensdörfer, L., Govaere, L., Lerink, B., Leyman, A., Schelhaas, M.-J., Teeuwen, S., Verkerk, P. J., & Van De Kerchove, R. (2025). European Tree Genus Map 2020 (Early Access Release) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.13341104 BibTeX @dataset{dekeersmaecker2025_treegenus, author = {De Keersmaecker, Wanda and Zanaga, Daniele and Senf, Cornelius and Viana-Soto, Alba and Klapper, Johanna and Blickensdörfer, Lukas and Govaere, Leen and Lerink, Bas and Leyman, Anja and Schelhaas, Mart-Jan and Teeuwen, Sander and Verkerk, Pieter Johannes and Van De Kerchove, Ruben}, title = {European Tree Genus Map 2020 (Early Access Release)}, year = {2025}, publisher = {Zenodo}, version = {early-access}, doi = {10.5281/zenodo.13341104}, url = {https://doi.org/10.5281/zenodo.13341104} } References [1] Alberdi, I., Bombín, R. V., González, J. G. Á., Ruiz, S. C., Ferreiro, E. G., García, S. G., Mateo, L. H., Jáuregui, M. M., Pita, F. M., & de Oliveira Rodríguez, N. (2017). The multi-objective Spanish national forest inventory. Forest systems, 26(2), 14. [2] Álvarez-González, J. G., Canellas, I., Alberdi, I., Gadow, K. V., & Ruiz-González, A. (2014). National Forest Inventory and forest observational studies in Spain: Applications to forest modeling. Forest Ecology and Management, 316, 54-64. [3] Finnish Forest Centre (Metsäkeskus). (2025). Forest resource lattice data (Hila-aineisto) [2019–2021]. Retrieved from https://www.metsakeskus.fi. [4] Fridman, J., Holm, S., Nilsson, M., Nilsson, P., Ringvall, A. H., & Ståhl, G. (2014). Adapting National Forest Inventories to changing requirements–the case of the Swedish National Forest Inventory at the turn of the 20th century. Silva Fennica, 48(3). [5] Govaere L. & Leyman A. (2023). Vlaamse bosinventarisatie Agentschap Natuur en Bos (VBI1: 1997-1999; VBI2: 2009-2018; VBI3: 2019-2021, v2023-03-17). [6] Heisig, J., & Hengl, T. (2020). Harmonized Tree Species Occurrence Points for Europe (0.2). https://doi.org/https://doi.org/10.5281/zenodo.5524611 [7] IGN. (2016). BD Forêt Version 2.0. January 2016 [8] Riedel T., Hennig P., Kroiher F., Polley H., Schmitz F., Schwitzgebel F. (2017): Die dritteBundeswaldinventur (BWI 2012). Inventur- und Auswertemethoden, 124 S. [9] Schelhaas MJ, Teeuwen S, Oldenburger J, Beerkens G, Velema G, Kremers J, Lerink B, Paulo MJ, Schoonderwoerd H, Daamen W, Dolstra F, Lusink M, van Tongeren K, Scholten T, Pruijsten L, Voncken F, Clerkx APPM (2022). Zevende Nederlandse Bosinventarisatie; Methoden en resultaten. Wettelijke Onderzoekstaken Natuur & Milieu, WOt-rapport 142. https://edepot.wur.nl/571720 [10] Villaescusa, R. & Díaz, R. (1998) Segundo inventario forestal nacional (1986–1996). Ministerio de Medio Ambiente, ICONA, Madrid. Acknowledgements We are very grateful for access to the forest plot inventories. We thank the Ministerio para la Transición Ecológica y Reto Demográfico (MITECO) for open access of the Spanish Forest Inventory (https://www.miteco.gob.es/). Finally, we would like to acknowledge the ForestPaths project (Co-designing Holistic Forest-based Policy Pathways for Climate Change Mitigation), that receives funding from the European Union's Horizon Europe Research and Innovation Programme (ID No 101056755), as well as from the United Kingdom Research and Innovation Council (UKRI).