Mapping burn severity in southern California using spectral mixture analysis

John Rogan, Janet Franklin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Scopus citations

Abstract

Several remote sensing techniques have been used successfully to map the areas of wildfire burn scars. Burn severity mapping, however, presents a suite of problems, caused by spectral confusion between vegetation affected by surface fire and unburned vegetation, between moderately burned vegetation and sparse vegetation, and between burned shaded and unburned shaded vegetation. A single date Landsat-7 Enhanced Thematic Mapper image was used to map five burn severity classes in two areas affected by wildfire in southern California in june 1999. Spectral mixture analysis (SMA), using four reference endmembers (vegetation, soil, shade, non-photosynthetic vegetation) and a single (charcoal-ash) image endmember, was used to enhance the image prior to supervised classification of burn severity. SMA provided a robust technique for mapping fire-affected areas due to its ability to extract subpixel information and minimize the effects of topography on single date satellite data. Overall kappa classification accuracy was high (0.81 and 0.72, respectively) for the burned areas, using five burn severity classes. Individual severity class accuracies ranged from 0.53 to 0.94.

Original languageEnglish (US)
Title of host publicationInternational Geoscience and Remote Sensing Symposium (IGARSS)
Pages1681-1683
Number of pages3
Volume4
StatePublished - 2001
Externally publishedYes
Event2001 International Geoscience and Remote Sensing Symposium (IGARSS 2001) - Sydney, NSW, Australia
Duration: Jul 9 2001Jul 13 2001

Other

Other2001 International Geoscience and Remote Sensing Symposium (IGARSS 2001)
Country/TerritoryAustralia
CitySydney, NSW
Period7/9/017/13/01

ASJC Scopus subject areas

  • Software
  • Geology

Fingerprint

Dive into the research topics of 'Mapping burn severity in southern California using spectral mixture analysis'. Together they form a unique fingerprint.

Cite this