HR: 0830h
AN: B41C-0892    [PDF]
TI: Increasing the accuracy of information from coarse-resolution satellite imagery
AU: * Hlavka, C A
EM: Christine.A.Hlavka@nasa.gov
AF: NASA Ames Research Center, MS 242-4, Moffett Field, CA 94035-1000 United States
AU: Dungan, J L
EM: Jennifer.L.Dungan@nasa.gov
AF: NASA Ames Research Center, MS 242-4, Moffett Field, CA 94035-1000 United States
AU: Roy, D P
EM: droy@kratmos.gsfc.nasa.gov
AF: NASA Goddard Space Flight Center, University of Maryland, Code 922, Greenbelt, MD 20771-0001 United States
AB: With the deployment of Earth Observing System (EOS) satellites that provide daily, global imagery, there is increasing interest in defining the limitations of the data and derived products due to its coarse spatial resolution, compared with higher spatial resolution but less frequently available data such as that provided by Landsat and airborne remote sensing, and the potential problems with combining EOS data with data from other sources. These issues are particularly relevant to studies of vegetation disturbance that require information on the amount and spatial pattern of fragmented types of land cover. For fragmented land cover, much of the detail, i.e. small fragments and notches in boundaries, is lost with coarse resolution imagery. As a result, not only do EOS image products look different than those generated with high spatial resolution imagery, but statistics such as land cover area and perimeter length are altered. Emerging methodologies based on fractal analysis and geostatistics can be used to adjust the information from satellite imagery to make them more consistent with finer resolution imagery and photography. These techniques make use of relationships between landscape measures at different spatial resolutions due to the self-critical processes that created them, and thus do not require a higher spatial resolution data set. Our adjusted estimates of wetland and burn scar area and forest perimeter from NOAA AVHRR imagery are much closer to values derived from Landsat than the usual products of pixel counts and pixel area. We also present adjusted estimates of burn scar area based on experimental EOS burn scar products and compare them with Landsat-based estimates.
DE: 0400 Biogeosciences
DE: 1640 Remote sensing
DE: 1694 Instruments and techniques
DE: 1890 Wetlands
SC: Biogeosciences [B]
MN: 2003 Fall Meeting