Northeastern Section - 48th Annual Meeting (1820 March 2013)

53
USE OF LIDAR IN A LANDSLIDE INVENTORY PROTOCOL IN VERMONT

Paper No. 53-4
Presentation Time: 3:10 PM

USE OF LIDAR IN A LANDSLIDE INVENTORY PROTOCOL IN VERMONT


CLIFT, Anne E., 53 Pinehurst Drive, Jericho, VT 05465, SPRINGSTON, George E., Geology and Environmental Science, Norwich University, 158 Harmon Drive, Northfield, VT 05663, gsprings@norwich.edu, and BECKER, Laurence, Vermont Geological Survey, 103 South Main St., Logue Cottage, Waterbury, VT 05671
Lidar is a key part of a new landslide inventory protocol being developed for the Vermont Geological Survey. The protocol is intended to identify existing active and inactive landslides and areas that are susceptible to landslides in non-bedrock terrain. It has been implemented at six test blocks in Chittenden County.

The protocol includes five phases: 1) Initial data collection on existing landslides; photo interpretation of orthophotos and leaves-off aerial photos; and field reconnaissance of a sampling of landslides. 2) Terrain analysis using GIS based on a bare-earth Lidar digital elevation model (DEM). The derived parameters include slope, roughness (standard deviation of slope), distance to stream, and topographic wetness index. 3) Frequency ratio analysis - Each parameter (slope, roughness, distance to stream, and topographic wetness) is divided into classes (e.g. slope <10o, 10-20o, 20-30o, 30-40o, 40-50o, >50o). The number of landslide pixels in each class is compared to the total number of pixels in that class and a frequency ratio is calculated. The frequency ratios for the most influential parameters are then added together, resulting in a preliminary map of potentially unstable areas. 4) Resulting maps are compared to the surficial geology, bedrock outcrops, soil survey data, topographic contours, and profile curvature. Sites are field checked to verify the results. 5) Maps showing potentially unstable areas are prepared.

The resulting maps are only as good as the data input into the analysis. The 3.2 meter lidar-derived DEM that is available for the study area has proved invaluable in assessing slope, roughness, profile curvature, topographic wetness, and producing contours at 2-meter intervals.