RAP 10m Annual Vegetation Cover

RAP 10m Annual Vegetation Cover and Production

Datasets:

Description:

RAP 10m Vegetation Cover: RAP Vegetation Cover 10m provides gridded annual estimates of plant functional group cover classes across rangelands in the western United States. Estimates span 2018–present and are produced at a 10-meter spatial resolution using top-of-atmosphere Sentinel-2 reflectance and a one-dimensional convolutional neural network (1D CNN). The model was trained on 47,833 field plots collected via the Natural Resource Conservation Services’ Natural Resources Inventory (NRI), Bureau of Land Management’s Assessment, Inventory, and Monitoring (AIM) program, National Park Service, and other contributors. The dataset includes fractional cover for 10 plant functional groups and land cover types—Annual Forbs and Grasses, Perennial Forbs and Grasses, Shrubs, Trees, Litter, Bare Ground, Invasive Annual Grasses, Sagebrush, Pinyon-Juniper. Predictions were made using temporally segmented and log-normalized Sentinel-2 reflectance, spatial coordinates, and derived vegetation indices (NDVI, NBR2). The model demonstrated strong predictive performance and showed slight improvements when compared with 30-m Landsat-based vegetation cover estimates by capturing finer-scale spatiotemporal heterogeneity critical to modeling rangeland ecosystems. Output is provided as annual 10 m GeoTIFFs available and as Google Earth Engine assets.

Organization: Maintained and produced by USDA Agricultural Research Service

Website: RAP Website

Google Earth Engine Catalog:

Vegetation cover: projects/rap-data-365417/assets/vegetation-cover-10m
Invasive annual grasses cover: projects/rap-data-365417/assets/invasive-annual-grass-cover-10m
Sagebrush cover: projects/rap-data-365417/assets/sagebrush-cover-10m
Pinyon juniper cover: projects/rap-data-365417/assets/pj-cover-10m
More information on Earth Engine assets through the Earth Engine Community Catalog.

Spatial resolution: 10-meter

Time Span:

Variables:

Terms of Use: