Search
US Forest Service Research & Development
Contact Information
  • US Forest Service Research & Development
  • 1400 Independence Ave., SW
  • Washington, D.C. 20250-0003
  • 800-832-1355
You are here: Home / Research Topics / Research Highlights / Individual Highlight

Research Highlights

Individual Highlight

Forest Land Estimates Improved by Novel Automated Mapping Technique Using Winter Satellite Imagery

Comparison of vegetation change tracker output products without (left) and with (right) snow-covered winter satellite imagery. Kirk Stueve, Forest ServiceSnapshot : Most automated satellite-based approaches for mapping forest lands rely on summer satellite imagery and are usually inconsistent with FIA plot-based estimates. Incorporating winter imagery in the mapping approach helps reduce the abundant false positives for forest and forest disturbance that frequently occur during the growing season, especially where forest is intermixed with wetland and agricultural landscapes. Reliable estimates of forest lands between FIA field plots allow customers the opportunity to explore forest dynamics more confidently beyond the grid of FIA plots.

Principal Investigators(s) :
Zimmerman, Patrick L.Nelson, Mark D.
Perry, Charles H. (Hobie)Gormanson, Dale D.
Research Location : Western Great Lakes area
Research Station : Northern Research Station (NRS)
Year : 2011
Highlight ID : 318

Summary

Forest Inventory and Analysis (FIA) plot data provide invaluable information about the distribution and health of our nation's forests to scientists and the public alike. Forest Service scientists found that winter satellite imagery with the vegetation change tracker (VCT) could generate more reliable estimates of forest lands in the western Great Lakes area. These VCT data were consistent with those from FIA plots. The VCT is an automated forest mapping algorithm that exploits the Landsat archive to produce comprehensive maps of forest changes and is well-suited for filling in data gaps between FIA plots. Using winter imagery exploited sharp seasonal contrasts of forested and nonforested areas and enabled the removal of most false positives, providing an efficient and reliable option for filling in gaps between FIA plots. False positives for forest and forest disturbance are a serious problem for the gaps in many FIA data grids, especially in intermixed forest and wetland and agricultural landscapes.

Forest Service Partners

External Partners

 
  • Chengquan Huang, University of Maryland, Department of Geography