National Sentinel-2 Grassland Mowing Products for Slovenia (2017–2025) with Copernicus GRAME Comparison Layers (2017–2024)

National Sentinel-2 Grassland Mowing Products for Slovenia (2017–2025) This dataset provides national-scale raster products describing the number and estimated timing of grassland mowing events in Slovenia derived from Sentinel-2 satellite time series for 2017–2025. The package also includes spatially aligned Copernicus Land Monitoring Service (CLMS) Grassland Mowing Events (GRAME) comparison layers for 2017–2024. The national Sentinel-2 products were generated using a pixel-based mowing detection approach that identifies characteristic decreases in the Normalized Difference Vegetation Index (NDVI), followed by vegetation regrowth. Candidate mowing events are further evaluated using temporal consistency, spatial support, observation quality and additional spectral information from NDII and shortwave infrared (SWIR) observations. Dataset contents The dataset is organized into annual ZIP archives containing GeoTIFF raster layers. For each year, the national Sentinel-2 products include: Original detected mowing count (SI_MOWING_COUNT). Mowing count after the SEIVE date-consistency filter (SI_MOWING_COUNT_SEIVE). Spatially harmonized mowing count (SI_MOWING_COUNT_SPATIAL_HARMONIZED). Estimated day of year for up to six detected mowing events (SI_MOWING_DOY_1 to SI_MOWING_DOY_6). An evidence-based mowing confidence score (SI_MOWING_CONFIDENCE). Number of valid Sentinel-2 observations (SI_VALID_OBSERVATIONS). Percentage of observations excluded by the Sentinel-2 Scene Classification Layer (SI_SCL_MASKED_PERCENT). For 2017–2024, the annual archives additionally include aligned CLMS GRAME mowing count, first-event date and confidence layers. These layers are provided for comparison with the national Sentinel-2 results. No GRAME comparison layers are included for 2025. Spatial and temporal characteristics The raster products cover grassland areas across Slovenia on a common 10 m grid in the WGS 84 / UTM zone 33N coordinate reference system (EPSG:32633). The national mowing detection analysis covers the growing seasons of 2017–2025. Interpretation and limitations The products represent satellite-derived estimates of mowing activity rather than direct field observations. The estimated dates indicate the timing of detected vegetation changes and may differ from the actual mowing dates. Detection reliability depends on the availability and temporal distribution of cloud-free observations, vegetation conditions and local land management practices. Confidence scores describe the strength and consistency of the available evidence and should not be interpreted as calibrated probabilities of correct detection. Confidence values from the national Sentinel-2 and CLMS GRAME products are based on different methodologies and are not directly interchangeable. The spatially aligned GRAME products are intended for comparative analysis and should not be considered independent ground-truth observations. Provenance and documentation The national Sentinel-2 products were developed by the University of Ljubljana, Faculty of Civil and Geodetic Engineering, as part of a project commissioned by the Institute of the Republic of Slovenia for Nature Conservation (ZRSVN). The accompanying README.md provides further information on file organization, raster characteristics, data encoding and recommended use. Copernicus data acknowledgement The aligned GRAME comparison layers contain European Union's Copernicus Land Monitoring Service information. The source products have been spatially aligned to the national Slovenian raster grid for comparative analysis. Source: Copernicus Land Monitoring Service – Grassland Mowing Events. The dataset was produced within the project "Analiza modela košenj za potrebe razvoja novih ukrepov za varstvo travišč" (Contract No. 4303-0033/2026-10), commissioned by the Institute of the Republic of Slovenia for Nature Conservation (ZRSVN) and carried out by the University of Ljubljana, Faculty of Civil and Geodetic Engineering.