A 500-m aboveground biomass dataset for Eurasian grassland from 2000 to 2024

This dataset provides annual gridded products of above-ground biomass (AGB) and its predictive uncertainty (standard error, SE) for grasslands across the Eurasian continent from 2000 to 2024. The spatial resolution is 500 m, and values are reported in grams of dry weight per square meter (g/m²). As the largest contiguous grassland region, Eurasian grasslands play a critical role in the carbon cycle, forage supply, and ecosystem services. This dataset offers an important benchmark for quantifying grassland carbon stocks, monitoring ecosystem health, and supporting regional sustainable development. The products were generated using a large number of field-measured samples, integrating multi-source predictor variables including MODIS vegetation indices and reflectance, climate factors, soil properties, and topography, and applying an optimized stacked ensemble modeling framework. Independent validation indicates a mean per-pixel uncertainty of approximately 11% (relative uncertainty), with relative uncertainty below 20% across all grassland types, demonstrating reliable quality. This dataset fills a gap in long-term, high-resolution, grassland-specific biomass products for the region and can serve as a key benchmark for studies of the global carbon budget, assessments of climate-change responses, and the sustainable management of Eurasian grasslands. The dataset is provided as raster GeoTIFF (.tif) files with an annual temporal resolution and a spatial resolution of 500 m. File names follow the convention YYYY_AGB_500.tif, where YYYY means the year. All files are in the Mollweide projection. Each GeoTIFF contains two bands: AGB, representing grassland above-ground biomass, and SE, representing the uncertainty (standard error) of the product.