USING AIRBORNE LIDAR TO CHARACTERIZE LANDSLIDE MORPHOLOGY AND DYNAMICS

Landslide mapping efforts have traditionally focused on delineating the margins of active slides, thus neglecting the wealth of information contained within landslide interiors. Although helpful with the identification of landslide boundaries, airborne LiDAR topographic data also enables detailed (~1m scale) characterization of internal landslide surfaces across extensive areas. The form of deformation features, such as compressional folds, back-rotated scarps, levees, broken ground, and shear fractures, reveals insights on the style and history of slope instability. Here, we present results from sites throughout the Western US where airborne LiDAR has fueled our investigation into short- and long-term dynamics of landslide processes. Using 2-D Fourier- and wavelet-based spectral filters to transform LiDAR DEMs, we observed that landslide-prone terrain is rougher than unfailed terrain at the scale of 10-40m. Using LiDAR data in the Portland Hills, OR, and Puget Sound, WA, we used this spectral-based diagnostic indicator to generate automated landslides maps for comparison with independently mapped landslides and achieved a greater than 85% success rate. In the Eel River catchment, CA, DEMs derived from LiDAR suggest that more than 95% of the terrain has experienced previous slope instability although only a fraction of the landslides are active. After ortho-rectifying sequential air photos from 1944 to 2006, we tracked the position of trees and large bushes on several active landslides at an approximately 10-yr interval. Velocities vary in accordance with mapped deformation features and some slides exhibit increased movement during the 1960-1970s, coincident with several intense hydrologic events. More recent deformation (2007 and onward) at the Eel River sites is revealed through processing of space-based ALOS InSAR data. Areas of rapid movement measured between February 2007 and February 2008 on the >5km long Boulder Creek earthflow correspond with a region of dense gully network development mapped with the LiDAR DEM, suggesting a process coupling between vigorous slide activity and overland flow erosion. Our results highlight the fundamental role that LiDAR plays in both identifying landslides and characterizing their behavior.