Tuesday, 18 November 2003: 2:30 PM
Data acquisition timing for burned area remote sensing and relationships to measures of burn severity
A U.S. Department of Interior program to use Landsat to spatially quantify burn severity, using the differenced Normalized Burn Ratio (dNBR or delta NBR), has prompted examination of a number of burns where both Initial and Extended Assessments were completed. The Initial Assessment typically is done with post-fire Landsat data acquired near the end, or very shortly after active fire. The Extended Assessment incorporates a post-fire Landsat scene from the growth period after fire, which may vary in time-since-burn from 8-12 months at more northerly latitudes having clear seasonality, to a month or so in sub-tropical regions where vegetation grows more-or-less continually throughout the year. Such timings affect the information content of the data, in regards to which aspects of burn severity are being captured. Therefore, they influence the types of applications appropriate for the data. Initial Assessment can provide good delineation of burned area, and preliminary estimates of severity. Vegetative regeneration will likely be missing, however, which may lead to overestimating severity. Such timing typically occurs at the end of the fire season when natural unburned vegetation is cured out, and contrast within burned areas may be diminished. Late-season initial assessment, however, may be the only timing available for some emergency response applications. By sampling the growth period after fire, Extended Assessment captures vegetation that has had a while to respond to fire. It can demonstrate both the survivorship potential and the delayed mortality of plants, each of which contributes to the ecological measure of burn severity. Results may be most useful for final portrayal and statistics of severity. They would be suited for projects that depend on accurate delineation of burn heterogeneity, like those comparing multiple burns over space and time, or in research and development of new models and methods. They also might better address long-term ecological consequences, such as impacts to sensitive species or communities, as well as certain risk factors, such as future fire potential.
Supplementary URL: