Detection of Earthflow Using a GPS and LiDAR Integrated Survey: A Case Study from the Slumgullion Landslide, Lake City, Colorado

The Slumgullion landslide in the San Juan Mountains near Lake City, Colorado has been a natural laboratory for landslide and environmental studies since the early 1900s. The landslide site covers 4.6 square kilometers and consists of an active part which has been moving continuously for about 300 years over an older, much larger, inactive part. We conducted an integrated GPS and LiDAR survey at the landslide site in one-week period from July 3rd to July 10th, 2015, with the primary purpose of delineating short-term ground deformation associated with the earthflow using advanced GPS and LiDAR techniques. A GPS network with twelve semi-permanent stations was set up, including seven stations on the sliding mass and five stations outside the sliding mass. A RIEGL VZ-2000 terrestrial laser scanner was used to collect data in the field. Airborne laser scanning data were collected by the National Center for Airborne Laser Mapping. We compared different registration methods for datasets acquired by the terrestrial laser scanner. A rapid workflow for field surveying and data processing was developed to generate high-resolution digital terrain models. The movement of the Slumgullion landslide was derived from semi-permanent GPS observations, and two repeated terrestrial laser scanning surveys conducted during the one-week period. A 1.47 cm horizontal daily movement was detected from the GPS observations. We compared different change detection strategies for the LiDAR point clouds measurements. Lateral landslide movements were detected from cloud-to-cloud comparison using the data from terrestrial laser scanning; the accumulated motion ranged from 3 cm to 10 cm during the survey week. The movement measurements derived from GPS and the terrestrial laser scanner agreed well. Our study demonstrates a method of identifying slow earth mass movement using the integration of GPS, terrestrial, and airborne laser scanning datasets. We developed a workflow for terrestrial laser-scanning data processing. Our method could be applied to study landslides in other regions. It is expected that our results will promote the application of GPS and LiDAR techniques in the practice of landslide hazards mitigation.