10.1594/PANGAEA.898416Huo, Lian-ZhiLian-ZhiHuo0000-0001-6705-6453Boschetti, LuigiLuigiBoschetti0000-0001-6525-4413Sparks, Aaron MAaron MSparks0000-0003-1286-3770PANGAEA2019disturbanceFireForest lossland clearingstressWELDFile nameFile formatFile sizeUniform resource locator/link to fileSupplementary Dataset10.3390/rs1105047720 data pointstext/tab-separated-valuesCreative Commons Attribution 4.0 InternationalThis dataset consists of 30 m spatial resolution forest cover loss, classified by disturbance type, for the conterminous United States for the period 2003-2011. Disturbance types considered include: stem removal, fire, and stress (for further detail see referenced article). The dataset is distributed as GeoTIFF format files and is based upon data from the University of Maryland Landsat-based Global Forest Change Product and Web-Enabled Landsat Data (WELD). The dataset described here is split into four files representing general United States regions: 'West_US.zip', 'Central_US.zip', 'Southeast_US.zip', and 'Northeast_US.zip'. U.S. regions can be referenced with the file 'US_region_map.jpg'. Each 30 m pixel is classified as one of the following classes: 0 - No loss in the current period, 1 - Stem removal, 2 - Fire, 3 - Stress, and 255 - No data. Further information on classes can be found in the referenced article. Files are in GeoTIFF format using a geographic coordinate system based on the World Geodetic System (WGS84) reference ellipsoid and using the Albers Equal Area projection with standard parallels of 29.5 degrees N and 45.5 degrees N and a central Meridian of 96 degrees W. For further information on the dataset and its format see the referenced article.Supplement to: Huo, Lian-Zhi; Boschetti, Luigi; Sparks, Aaron M (2019): Object-Based Classification of Forest Disturbance Types in the Conterminous United States. Remote Sensing, 11(5), 477