1-km aboveground forest biomass across the western United States, 1984–2025

The aboveground forest biomass dataset is produced by integrating the Landsat time series (1984-2025) with GEDI (Global Ecosystem Dynamics Investigation) biomass data. The Landsat time series was processed using LandTrendr (Kennedy et al., 2010), and four spectral indices were used to characterize vegetation dynamics: Normalized Difference Vegetation Index (NDVI), Normalized Difference Moisture Index (NDMI), Normalized Burn Ratio (NBR), and NBR2. Topographic variables (elevation, slope, and aspect), mean climate variables (precipitation, temperature and solar radiation over 1980-2025) from GRIDMET (Abatzoglou, 2013), and geographic coordinates (latitude and longitude) were included as additional predictors, with GEDI biomass as the response variable. All data were aggregated to a 1-km pixel resolution at an annual time step. Using GEDI training data from the five available years (2019-2022 and 2024) and the corresponding predictors, we trained a machine learning model based on LightGBM (Light Gradient Boosting Machine). We then estimated aboveground forest biomass for each year from 1984 to 2025. The resulting dataset was validated against FIA (Forest Inventory and Analysis) inventory data (R2 = 0.9), airborne lidar-based biomass estimates (R2 = 0.7), and aboveground carbon emissions from fire-related forest loss (van Wees et al., 2022) (R2 = 0.9). Note: This is an initial version of the dataset and may be revised in future releases. Data Characteristics: Spatial Coverage: Western United States Projected Coordinate System: NAD 1983 Contiguous USA Albers (EPSG: 5070) Spatial Resolution: 1000 m × 1000 m Temporal Resolution: Annual Temporal Coverage: 1984-2025 Units: Mg/ha (megagrams per hectare) NoData Value: -9999 Data Type: Float64 File Formats: GeoTIFF (.tif) File Naming Convention: WUS1kmAGBD_YYYY.tif, YYYY = 1984-2025 References: Abatzoglou, J. T. (2013). Development of gridded surface meteorological data for ecological applications and modelling. International Journal of Climatology, 33(1), 121–131. Kennedy, R. E., Yang, Z., & Cohen, W. B. (2010). Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr — Temporal segmentation algorithms. Remote Sensing of Environment, 114(12), 2897–2910. https://doi.org/10.1016/j.rse.2010.07.008 van Wees, D., van der Werf, G. R., Randerson, J. T., Rogers, B. M., Chen, Y., Veraverbeke, S., et al. (2022). Global biomass burning fuel consumption and emissions at 500 m spatial resolution based on the Global Fire Emissions Database (GFED). Geoscientific Model Development, 15(22), 8411–8437. https://doi.org/10.5194/gmd-15-8411-2022