A Landsat-derived Burned Area Emergency Response (BAER) bi-temporal burn severity dataset for the conterminous United States
- Source
- Zenodo
- Published
- 2026
- Country
- United States
- Region
- North-America
- Format
- tif
- Access
- Open access
- License
- cc-by-4.0
- Themes
- Natural Disasters, Remote Sensing & Satellites, Geospatial & GIS Data
About this dataset
A bi-temporal (pre- and post-fire) burn severity dataset derived from USGS/USFS Burned Area Emergency Response (BAER) soil burn severity products and Landsat-8 OLI surface reflectance over the conterminous United States, using acquisitions from 2021 to 2025. It contains 3,156 spatially aligned 256 × 256 GeoTIFF tiles organized into pre/ and post/ (6-band, uint8; band order B2, B3, B4, B8A, B11, B12) and label/ (1-band, uint8). Labels follow a four-class soil burn severity scheme (0 = Unburned, 1 = Low, 2 = Moderate, 3 = High), with 255 marking NoData pixels outside the BAER grading extent.
How the data was collected
A bi-temporal (pre- and post-fire) burn severity dataset derived from USGS/USFS Burned Area Emergency Response (BAER) soil burn severity products and Landsat-8 OLI surface reflectance over the conterminous United States, using acquisitions from 2021 to 2025. It contains 3,156 spatially aligned 256 × 256 GeoTIFF tiles organized into pre/ and post/ (6-band, uint8; band order B2, B3, B4, B8A, B11, B12) and label/ (1-band, uint8). Labels follow a four-class soil burn severity scheme (0 = Unburned, 1 = Low, 2 = Moderate, 3 = High), with 255 marking NoData pixels outside the BAER grading extent.