ACCURACY ASSESSMENT OF A LANDSAT-ASSISTED VEGETATION MAP OF THE COASTAL PLAIN OF THE ARCTIC NATIONAL WILDLIFE REFUGE

An accuracy asessment of a vegetation map based on Landsat multispectral scanner data was conducted on the coastal plain of the Arctic National Wildlife Refuge. The overall map accuracy was 37 percent (SD=4 percent) based on ground data collected at 126 sites. The majority of errors (27 percent of the total observations) occurred between closely related land-cover classes. Classes were related to each other along gradients of moisture, shrub cover, or total plant cover. Misclassification errors, where the plant community observed on the ground clearly did not agree with the map classification, occurred at 22 percent of the polygons sampled. The most common misclassifications occurred for braided river floodplain communities which were classified as very wet, wet, or moist/wet graminoid tundra. Description errors occurred for 4 percent of the polygons indicating that some of the land-cover class descriptions need additions or modifications to allow the user to accurately identify the class. In conclusion, the Landsat MSS map provides information on the distribution of general vegetation types across the coastal plain, but does not provide site-specific vegetation data.

  • Availability:
  • Corporate Authors:

    American Society for Photogrammetry and Remote Sensing

    5410 Grosvenor Lane, Suite 210
    Bethesda, MD  United States  20814-2160
  • Authors:
    • Felix, N A
    • Binney, D L
  • Publication Date: 1989-4

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00485400
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jul 31 2003 12:00AM