{"name": "Wildfire Analysis Toolkit", "package_name": "wildfire_analysis_toolkit", "description": "Wildfire burn severity from satellite imagery to MTBS-style classes, area summaries, maps and reports, from a fire perimeter or just an MTBS fire's name: NASA HLS, Landsat and Sentinel-2, calibrated, with dNBR, RdNBR and RBR.", "about": "Processing tools for wildfire burn severity, modeled on the MTBS QGIS Fire Mapping Tool. They appear in the Processing Toolbox under Wildfire Analysis Toolkit.\r\n\r\nBurn severity (whole pipeline) goes from a fire's perimeter and dates to an image pair, pre- and post-fire NBR, dNBR with the unburned offset, MTBS severity classes, area summaries per sub-area and an HTML report. The fire name and dates can be read from a WFIGS perimeter record. Severity map (PDF) makes a print-ready map.\r\n\r\nImagery: Suggest image pairs picks pre- and post-fire dates by the MTBS rules. Sources are NASA HLS v2.0 (needs a free NASA Earthdata login in a .netrc file), Landsat 4-9 Collection 2 and Sentinel-2 L2A from Microsoft Planetary Computer (no account), or scenes you downloaded. Images are calibrated from their metadata, cloud-masked, and read only for the area of interest.\r\n\r\nSeverity: Build dNBR (dNBR, RdNBR, RBR and the offset), Draft burned-area perimeter, Classify severity (MTBS classes, with thresholds given or estimated from nearby MTBS fires), and Burn severity summary (percent, hectares and acres per class for a fire and its sub-areas).\r\n\r\nMTBS fires: give an MTBS fire's id or name and year, and the plugin fetches its perimeter and MTBS record (ignition date, MTBS's own image pair, thresholds) itself, then compares the result with MTBS's published classes. A perimeter you give is matched to its MTBS record too: a WFIGS perimeter by its IRWIN id, any other by an MTBS fire inside it that ignited between its dates. Find MTBS fires looks up any of the 31,000+ fires MTBS has mapped since 1984, by name, id, years or area. No manual download: the USGS MTBS records are fetched and cached automatically, and the plugin works offline from the cache. The newest fires' perimeters need a one-time 420 MB USGS download, made only when you allow it, then cached. When NASA HLS lacks MTBS's scenes, both come from Planetary Computer, so the pair is processed alike; Sentinel-2C scenes are supported.\r\n\r\nValidated against 17 published MTBS fires in Arizona, New Mexico, Colorado and Utah, 2022-2024: with MTBS's dates, perimeters and thresholds, the classes agree at a mean kappa of 0.775 (0.711 to 0.864). Uses only QGIS, GDAL and NumPy, which ship with QGIS.\r\n\r\nThe README in the code repository has every parameter, the method, troubleshooting and the validation results.", "homepage": "https://conklingeospatial.com/projects/qgis-wildfire-analysis-toolkit/", "repository": "https://github.com/stephenconklin/QGIS_Wildfire_Analysis_Toolkit", "tracker": "https://github.com/stephenconklin/QGIS_Wildfire_Analysis_Toolkit/issues", "author": "Stephen Conklin", "tags": ["wildfire", "fire", "burn severity", "mtbs", "dnbr", "rdnbr", "rbr", "nbr", "hls", "landsat", "sentinel-2", "planetary computer", "earthdata", "stac", "remote sensing", "zonal statistics", "processing", "burned area", "fire perimeter", "nifc", "wfigs"], "downloads": 9, "latest_version": "2.1.2", "versions": [{"version": "2.1.2", "experimental": false, "qgis_min": "3.44.0", "qgis_max": "4.99.0", "downloads": 9, "uploaded_by": "stephenconklin", "upload_datetime": "2026-11-08T15:27:57.403089"}]}