Paper No. 4-15
Presentation Time: 8:30 AM-4:30 PM
DETECTION OF EARTHFLOW CREEP FROM TOPOGRAPHIC DIFFERENCING OF AIRBORNE LIDAR AND SUAS – DERIVED HIGH RESOLUTION TOPOGRAPHY, SHURTZ LAKE, UTAH, USA
The Shurtz Lake earthflow, located near Thistle, Utah, was first observed in May, 1997 after damaging power lines and later railway lines (Ashland, 1997). It is apparently developed in the North Horn and Ankerah formations (similar to the nearby Thistle landslide), and is ~1200 m long, up to 315 m wide, and traverses ~250 m in relief. We measured ongoing motion of the earthflow by differencing high resolution topography (HRT) created using a small uncrewed aerial system (sUAS) in November, 2019 against airborne LiDAR HRT data collected in 2018 (~8 points/m2; Utah AGRC), and in 2017 (~11.8 pts/m2; Jones, 2018, OpenTopography). The sUAS-derived HRT (~273 pts/m2) was made with structure-from-motion (SfM) processing of 720 photos collected with a Sensefly eBee Plus and georeferenced using onboard dual-frequency differential global navigation satellite system (dGNSS) and 29 ground check points. Vertical precision of the SfM data relative to the checkpoints is 3.7 cm RMSerror. Co-registration of the three HRT data sets was improved using iterative-closest-point to quantify and remove misfit of the datasets in stable areas off the earthflow. Digital elevation models from 2017 (LiDAR), 2018 (LiDAR), and 2019 (SfM) were differenced, and the point clouds from 2017 (LiDAR) and 2019 (SfM) were also differenced. Differencing reveals localized creep, primarily at the fronts of lobes within the earthflow. Locations from where material was displaced are also evident. We measured downslope lobe-front advancement of ~.015 to ~30 m, with ~.010-.015 m being the approximate lower limit of detection with the available data. The results show that differencing HRT, sUAS – derived data in particular, is a viable and efficient method to detect earthflow motion. In addition, we anticipate that future differencing using only high point density sUAS-SfM HRT will allow the detection of smaller amounts of movement at a significantly better spatial resolution.