Evaluation of Post-Wildfire Susceptibility to Landslide Hazards using Remote Sensing and GIS in the Salmon-Challis National Forest, Lemhi County, Idaho

摘要

Remote sensing and field-based research is used to investigate landslide hazards resulting from wildfire activity in central Idaho. This study focuses on three areas in the Salmon-Challis National Forest (SCNF). Field-based research is performed on the sites to test the applicability of remote sensing and geographic information systems (GIS) to hazard prediction. A relative landslide hazard map is developed for each of the test sites. The classification system for the hazard maps includes slope, aspect, burn severity, hydrology, bedrock geology, and soil characteristics. Each parameter is assigned a relative hazard ranking of ordinal numbers. Results indicate that digital elevation models are useful in providing comprehensive slope and aspect analysis. High-resolution multispectral data can be used successfully to map areas of high burn severity and unburned areas, although mapping areas of low to moderate burn severities proved more difficult. A comparison between field-generated hazard maps and digital-generated maps indicates that the greatest difference between two is scale. The digital method classifies the hazard maps in 4 to 10-meter pixels, whereas field mapping tends to classify larger areas based upon similar features. This study has produced a well-documented methodology of landslide hazard assessment and a qualification of the relative importance of land cover parameters in the SCNF.