Shallow Landslide Susceptibility Mapping Using Stereo Air Photos and Thematic Maps

Abstract

Overlay (or heuristic) models have been shown to be a reasonable method to generate landslide susceptibility maps. However, overlay models are subjective and rely on expert judgment. The method developed here employs an overlay model but it tries to estimate certain parameters, such as slope angle or distance to streams, in a more quantitative manner. Each individual factor believed to have an influence on mass wasting processes is scaled to a value between 0 and 1. All different factors are then multiplied, producing the degree of stability, Df, which lies on a ratio scale, rather than a nominal or interval scale. Thus, a pseudo probability of failure can be obtained. The data required are derived partly from thematic maps and partly from stereo air photos. The air photo stereo pair is used to automatically derive a digital elevation model which is then used to create an orthoimage of an area. The air photos are also used to develop a land use map, an important component in estimating landslide susceptibility. This methodology was tested on a section of the Santa Monica Mountains near Malibu, California. The resulting map both visually and statistically appears to successfully identify problematic areas. Although the methodology was primarily developed to identify shallow slides, the crest of a large, deep-seated landslide was also recognized. The general methodology should be easily adaptable to other regions or could be applied with a different set of factors than those considered in this specific case.