USING HIGH-RESOLUTION DEMs TO IDENTIFY SPATIAL CORRELATIONS BETWEEN TOPOGRAPHIC FORM AND HILLSLOPE FAILURE MECHANISMS, SOUTHERN CALIFORNIA

A large number of hillslope failures have been observed within a section of the Transverse Ranges north of the city of Los Angeles. The abundance and seemingly random distribution of these landslides poses a threat to infrastructure and utility lines vital to the needs of the Los Angeles area. The types of hillslope failures found in the area are highly diverse and include small debris flows, planar-block slides, rotational slides, wedge-block slides, rotational slides, and large earth flows. The spatial factors controlling the distribution of landslide location and types in this area remains poorly understood. Previous investigations of hillslope instability in the area are limited to individual landslides that are directly affecting utility lines or roads. While these types of studies are useful for conducting repairs and remediation efforts, they lack an understanding of the regional scale conditions and do not support the development of proactive approaches to landslide monitoring and mitigation in the area. In this study, we attempt to use topographic information derived from high-resolution digital elevation models in order to identify key spatial relationships that can be used to explain and predict the distribution of landslide locations within the study area. These high-resolution DEMs were created using airborne LiDAR collected in fall of 2007. Landslides were identified and mapped in a GIS using the high-resolution DEMs and low-altitude aerial photography. The high resolution DEM was then used to derive topographic measurements that include slope, aspect, curvature, landslide area, and hydrologic area for each landslide. These topographic factors are compared to other non-topographic spatial variables including soil type, lithology, and geologic structure of the bedrock to identify key spatial relationships. These spatial relationships are then used to develop predictive models of hillslope stability for the study area. The use of LiDAR technology for this study allowed these landslides to be mapped quickly and accurately over a relatively large area. In addition, key topographic statistics for each landslide were easily quantified and compared with the forcing factors. This suggests that the use of LiDAR is an efficient and effective tool for such regional-scale landslide investigations.