2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 13
Presentation Time: 11:25 AM

STATE-OF-THE-PRACTICE IN CHARACTERIZATION OF LANDSLIDE PRONE TERRAIN USING REMOTE SENSING PLATFORMS, GLOBAL POSITIONING AND GEOGRAPHICAL INFORMATION SYSTEMS TECHNOLOGY


ROGERS, J. David and JAMES, Kevin L., Department of Geological Engineering, Univ of Missouri-Rolla, 129 McNutt Hall, 1870 Miner Circle, Rolla, MO 65409, rogersda@umr.edu

The identification of landslides and terrain evaluation in general is presently undergoing a radical shift: from the traditional employment of stereopair aerial photos to a wide variety of digital sensor platforms and techniques which are sensitive to many factors unknown to most end users. The USGS has shifted over to digital map products. Their principal photographic product are mosaic Digital Orthographic Quarter Quadrangles (DOQQs), which are orthorectified, so cannot be viewed in stereo.

Digital imagery is being marketed by commercial firms and governmental agencies. Many commercial operators are now capable of recording 1 m digital color imagery for inventory purposes, replacing stereopair aerial photos. NASA is selling their IKONOS 1 m resolution imagery with 4 spectral bands, while the USGS National Map program is imaging the nation’s largest cities using three bands (RGB), with close to 1 m resolution. The National Imagery and Mapping Agency is collecting digital imagery with three times the resolution of the NASA and USGS imagery, but with limited distribution.

The emerging collection technology includes Light Detection and Ranging (LiDAR), various bands of Synthetic Aperture Radar (SAR), Laser Detection and Ranging (LADAR), and multispectral scanners, which can sense up to 200 channels across the elctromagnetic spectrum. In February 2000 NASA launched the Shuttle Radar Topography Mission (SRTM), which used Interferometric SAR (INSAR) with 30-meter horizontal resolution to provide elevation control data for 80% of Earth. These sensory techniques compare favorably with photogrammetric-derived data, but topological conditions may be inadvertently corrupted by the 30m imaging process, creating “void areas”. Researchers need to compare data reliability versus distance versus pixel size and; hopefully, offer recommendations for how such data might be modified before distribution.