REMOTE-SENSING OF FOREST-FIRE SEVERITY AND VEGETATION RECOVERY

Burned forested areas have patterns of varying burn severity as a cons equence of various topographic, vegetation, and meteorological factors . These patterns are detected and mapped using satellite data. Other e cological information can be abstracted from satellite data regarding rates of recovery of vegetation foliage and variation of burn severity on different vegetation types. Middle infrared wavelengths are useful for burn severity mapping because the land cover changes associated w ith burning increase reflectance in this part of the electromagnetic s pectrum. Simple stratification of Landsat Thematic Mapper data define varying classes of burn severity because of changes in canopy cover, b iomass removal, and soil chemical composition. Reasonable maps of burn severity are produced when the class limits of burn severity reflecta nce are applied to the entire satellite data. Changes in satellite ref lectance over multiple years reveal the dynamics of vegetation and fir e severity as low burn areas have lower changes in reflectance relativ e to high burn areas. This results as a consequence of how much the si te was altered due to the burn and how much space is available for veg etation recovery. Analysis of change in reflectance across steppe, rip arian, and forested vegetation types indicate that fires potentially i ncrease biomass in steppe areas, while riparian and forested areas are slower to regrow to pre-fire conditions. This satellite-based technol ogy is useful for mapping severely burned areas by exploring the ecolo gical manifestations before and after fire.