Fire perimeter databases provide critical geospatial information that land managers use for a variety of purposes, including documenting fire occurrence, determining historical reference conditions and fire regimes, planning fuels treatments and fire management operations, and assessing potential vulnerability of infrastructure and resources. The Alaska Large Fire Database (ALFD) includes fire perimeters from 1940 to present based on the best-available data for delineation at the time of the fire. Inaccurate perimeters are confounding factors for researchers and practitioners, particularly when used to guide fire suppression tactics or inform natural resource decisions at the landscape level. Funding was received through the U.S. Fish and Wildlife Service (FWS)/ U.S. Geological Survey (USGS) Quick Response Program to develop and execute a workflow for updating fire perimeters on Alaska National Wildlife Refuges (NWR) using remotely sensed imagery and cloud computing.
The timely incorporation of fire perimeters into the ALFD has been an important element of every fire season in Alaska. However, a result of this process is that fire perimeters are generated using different methods, ranging from personnel in aircraft using paper maps to technicians interpreting remote sensing data. Alaska’s large and remote fires can also creep and smolder after fires have been called out. To address these inconsistencies, a retroactive analysis using Google Earth Engine was performed to efficiently calculate and depict differenced Normalized Burn Ratio (dNBR) on a near-annual basis. This standard index is routinely used to identify and characterize burned landscapes. By comparing fire perimeters in the ALFD to the corresponding dNBR footprints, researchers could determine the accuracy of perimeters and updated them as needed. Because the cloud-based archive of Landsat images dates back to the 1970s, most fire events within the ALFD for all refuge lands could be reexamined.
Fire perimeters were updated based on interpretation of annual dNBR mosaics or, when available, replaced by perimeter records available through the Monitoring Trends in Burn Severity program (MTBS). The updated perimeters are expected to provide a more consistent and accurate database of the location and extent of burned areas. The improved accuracy will also facilitate identification of short-interval reburns, a phenomenon with operational and ecological significance.
Northern portion of the 1990 BTT SW 53 fire perimeter as it appears in the Alaska Large Fire Database. Note the discrepancies of the perimeter (red) with the Monitoring Trends in Burn Severity differenced Normalized Burn Ratio (dNBR) mosaic which captures burn scar (black) as sensed by Landsat 5.
Disclaimer: Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.