Using QuickBird and Landsat Imagery to
Assess Burn Severity and Monitor Vegetation Recovery

An urgent task after a large wildfire is to map burn severity to indicate fire effects on soils and vegetation. The Normalized Difference Vegetation Index (NDVI) is indicative of green vegetation; differencing pre- and post-fire imagery (i.e., dNDVI) is indicative of vegetation change. We compared dNDVI derived from high-resolution QuickBird and moderate-resolution Landsat imagery for mapping burn severity after the 2005 School Fire. High resolution QuickBird imagery allows users to distinguish fine-scale burn severity characteristics (e.g., individual trees, exposed soil patches, and small canopy openings). Image data were compared to soil and vegetation field data collected immediately, one, and two years post fire. Preliminary results suggest QuickBird imagery more accurately indicates unburned (green) and charred (black) areas, as compared to Landsat imagery. We also found dNDVI may be a reasonable substitute for the more commonly used differenced Normalized Burn Ratio (dNBR) when related to post-fire soil and vegetation measurements.