Terrascope users have shown great interest in global land cover maps in the past. The ESA WorldCover data products were downloaded over one million times in the past five years. Now, the production of yearly global land cover maps at 10 m resolution (and year-to-year change maps) is part of the Copernicus Land Monitoring Service. We are pleased to announce that Terrascope now also offers this information product (LCM-10). You can discover the map in the Terrascope viewer, download it via the Terrascope STAC browser and work with the data via the Terrascope openEO backend.
Why is it important?
The key added value is that the LCM-10 land cover map for 2020 is now usable in cloud-based analysis workflows. Through STAC, the dataset becomes discoverable and interoperable, and through openEO it can be processed directly close to the data.
This enables practical use cases such as extracting land cover statistics for countries, NUTS regions, or custom areas of interest. In other words, users can move from viewing a 10 m land cover map to deriving reproducible indicators, such as the area or percentage of tree cover, cropland, built-up area, wetlands, water, or other LCM-10 classes.
This is particularly relevant in the context of (official) statistics, where the challenge is often not the availability of EO data, but the “last mile”: turning pixels into comparable, reportable statistics. The example notebook also accounts for the fact that LCM-10 is provided in WGS 84 geographic coordinates, where pixel area varies with latitude, by applying an area correction before aggregating statistics.
How can you access the data?
The data can be accessed in several ways:
First, users can explore the LCM-10 layer visually through the Terrascope viewer.
Second, developers can discover and access the collection through the Terrascope STAC catalogue: https://stac.terrascope.be/collections/lcfm-lcm-10
Third, users can process the data through the Terrascope openEO backend, using their CDSE login. This allows scalable analysis without downloading the full dataset locally.
As a practical starting point, we prepared a notebook for extracting land cover statistics with openEO: https://github.com/Open-EO/openeo-community-examples/blob/main/python/LandCoverStatistics/land_cover_country_lcfm.ipynb
You can run it directly on Terrascope's JupyterHub: https://notebooks.terrascope.be/user-redirect/lab/tree/notebook-samples/openeo-samples/python/LandCoverStatistics/land_cover_country_lcfm.ipynb
The notebook is processes LCFM’s LCM-10 map and includes country-based geometry retrieval, multi-country processing, area correction for WGS 84 and integrated visualization. It can use Eurostat GISCO geometries, including NUTS regions, to produce statistics for administrative reporting units.
Land Cover distribution per province in Belgium, based on LCFM's LCM-10 land cover map. The ‘Other’ category compounds all other land cover classes than the top 4, but is mainly ‘Water’ for Belgium.
Generated using European Union's Copernicus Land Monitoring Service information.
What more can we expect in the future?
The current LCM-10 availability is an important first step. The LCFM service is developing annual LCM-10 products for 2021–2025, together with corresponding 10 m Land Cover Change Maps, LCCM-10. These products are expected to be released later in 2026. Then you will have 6 annual maps, and 5 change maps to work with.
For more information on the product, visit our blog.
