Internal kinematics of the Slumgullion landslide (USA) from high-resolution UAVSAR InSAR data

Abstract

Landslides represent one of the most damaging natural hazards and often lead to unexpected casualties and property damage. They also continually modify our natural environment and landscapes. Knowledge of landslide systems is largely restricted by the stochastic nature, subjective interpretation and infrequent or spatially sparse surveying of landslides. Characterized by persistent daily movements of a couple of centimeters over multi-centennial timescales and a long narrow shape as long as ~4 km, the Slumgullion landslide in Colorado, USA represents an ideal natural laboratory to study slow-moving landslides. Here we demonstrate the capability of the highly accurate, spatially continuous airborne Synthetic Aperture Radar (SAR) system of the NASA Uninhabited Aerial Vehicle SAR (UAVSAR) to characterize the kinematic details of internal deformation of the Slumgullion landslide using SAR interferometry (InSAR). We develop a phase-based approach to automatically extract the boundaries of the mobile geological structures without unwrapping. Comparison with historic field observations from 1991 reveals the 40-m advance of the frontal toe and shift of an internal fault. The UAVSAR data also resolve an internal minislide (100 by 70 m), which moves more southerly than the main body at 5 mm/day in the lower part of the landslide. A Light Detection and Ranging (LiDAR) Digital Elevation Model (DEM) shows that the minislide is associated with the opening of a 30 by 10 m pull-apart basin and bounding strike-slip faults. These extensional structures, nearby incised streams, and steepened local slopes helped establish the kinematic environment for the formation of the secondary minislide. The disparity between the UAVSAR InSAR-derived horizontal moving directions and the LiDAR DEM-derived slope aspects suggest that while the surface topography governs the first-order orientation, the local kinematics is also subject to the variable nature of heterogeneous landslide materials and the irregular basal bedrock surface. The landslide velocity and precipitation show similar multi-annual variations. Our study demonstrates that the freely available, high-resolution UAVSAR data, have great potential for characterizing landslide kinematics and other small-scale geological and geomorphological processes.