Northern Rocky Mountain Science Center (NOROCK)
Home | About Us | Science | Product Library | News & Events | Staff | Students | Partners | Contact Us
Post-fire Burn Assessment by Remote Sensing on Federal Lands
USGS-National Park Service Burn Severity Mapping - http://burnseverity.cr.usgs.gov/
Remote sensing and field sampling techniques to support national assessment of burned areas was under development from the mid-1990's by Carl Key and Nate Benson, USGS and National Park Service, respectively. At the time, there was no standard approach to broadly quantify and spatially represent burn severity or the ecological effects of fire within the NPS. The quality and extent of such data varied greatly among NPS units; there was no routine process to continually generate new information as fires occurred, while no post-fire information existed for many past fires. We pursued a number of strategies to initiate a national program for burn severity mapping and monitoring within NPS and USGS. First was research to develop effective remote sensing from available technology that was affordable, relatively simple to process, provided national coverage, and yielded reasonably consistent results across ecosystems. It was also important have accessible archived data, so that past fires could be represented in the same database by consistent means. This led to a Landsat-based change detection approach and formulation of the differenced Normalized Burn Ratio (dNBR), referencing the first growth period after fire as the key time to gauge severity. Second was research to conceptually define burn severity ecologically and advance methods to measure it in the field, in contexts appropriate for the scale and detection capabilities of the sensor and dNBR. This led to developing the Composite Burn Index (CBI), used to summarize the fire effects on a site for validation, calibration and understanding of remotely sensed data. Both dNBR and CBI methods were documented in FIREMON), a 2001 Joint Fire Sciences Program (JFSP) funded project to create and disseminate a broad suite of standard fire effects monitoring protocols. We also presented FIREMON classroom and field training for interagency personnel. Third was enlisting support of the NPS and USGS Center for EROS to implement the program on a national scale beginning in 2001. Only by the size and existing resources of EROS was such a program made feasible. Fourth was to engage in education and field sampling missions principally with NPS fire personnel stationed around the country, but also including individuals from other agencies and universities. Knowledge of fire effects and remote sensing was disseminated, while our understanding was expanded over a wide range of fire situations and effects. The network of trained teams enabled consistent and extensive data collection in numerous burns and diverse ecosystems. Widely distributed CBI data afforded feasibility and sufficient coverage for national validation and calibration of dNBR, which became the subject of another JFSP project 01-1-4-12.
Around 2001, methods were also transferred to the US Forest Service where, along with the US Department of Interior, remote sensing techniques were adopted for use in Burned Area Emergency Rehabilitation (BAER) assessments, depending on availability of suitable Landsat data. Such assessments are different in character than the extended assessments, given they must be completed very soon after significant burning, are limited to selected fires, and target soil effects or impacts of greatest priority for emergency treatment.
With NPS support, EROS staff and production capabilities currently generate burn severity spatial data for all NPS fires since 2000, and maintain web-based information and data distribution. Fires since 1983 also have been completed for some areas, while NPS units are beginning to use CBI data to provide a baseline for subsequent recovery and monitoring. We continue to be actively engaged in technical support and trouble shooting for production activities, such as in scene selection and quality control. We also continue to carry out training and field sampling, and continue with remote sensing and fire effects research in partnership with NPS. Current emphasis is to complete publication on evaluation results across the U.S., to demonstrate useful applications of the extensive data holdings, and to derive new information on trends and the landscape ecology of fire.
Monitoring Trends in Burn Severity (MTBS) - http://www.mtbs.gov
In many ways, MTBS is a natural extension of National Park Service-USGS Burn Severity Mapping, with methodology that links back to earlier work by Key and Benson at Glacier National Park. In June 2004, the Government Accountability Office (GAO) recommended that all federal land management agencies adopt the USGS-NPS burn severity mapping protocol as a monitoring mechanism to assess fire effects and broader wildland fire patterns and trends (GAO. 2004. WILDLAND FIRES: Forest Service and BLM Need Better Information and a Systematic Approach for Assessing the Risks of Environmental Effects. (http://www.gao.gov/new.items/d04705.pdf). Soon after, the Wildland Fire Leadership Council (WFLC) requested and later supported such a program under the National Fire Plan, which established MTBS in late 2005.
MTBS covers all fires over a minimum size on Federal or State lands throughout the U.S. The USGS Center for EROS and the US Forest Service Remote Sensing Application Center share production activities. Current and future fires are included, as well as past fires back to about 1982. The main goal of MTBS is to generate standard, comparable data to monitor fire trends and management effectiveness on a national level, in contrast to direct emergency response applications. MTBS annually reports to WFLC on the area burned by severity class and other stratifications, and provides technology transfer, remote sensing data, and derivative products on fire extent and severity to the general fire science community. Current products are based on Landsat imagery and the dNBR obtained during the growing season after fire for an extended assessment on most burns. In ecotypes that recover quickly, such as grasslands, results may be based on an initial assessment acquired as soon as suitable data is available after fire.
Processes to help and improve MTBS continue to be explored and evaluated, along with expanded understanding of how various ecosystems respond to fire. There is interest, for example, in how a relative version of dNBR contributes to the definition and mapping of severity in some systems (Miller and Thode 2007). Moreover, investigations that use the burn severity time-series in landscape and fire ecology are of interest, including relationships to climate, trends and modeling of fire behavior, and responses of vegetation and wildlife. Current contributions are also made to MTBS in a science advisory capacity. That involves development of instructional materials and guidance, for example on scene selection, severity classification and thresholding, image processing, and dNBR normalization . . . training and assistance in field sampling using CBI . . . developing a set of reference fires to serve in calibration across MTBS analysts . . . quality review of products . . . and creating a photo series of the burn severity continuum in key ecosystems of the U.S.
