Forcings and kinematics of slow-moving landslides revealed by airborne and spaceborne SAR data

Abstract:

The hazards that landslides present and their impact on landscapes primarily depend on their volume and the rate at which they move. Landslide rheology governs the deformation and flow behavior of sliding masses. However, quantifying landslide dynamics is challenging due to the stochastic nature of the properties and forcings of the system. Although slow-moving landslides rarely lead to casualties, they may affect nearby lifelines and infrastructures and they may transform to instant failures during or after intense precipitation or an earthquake. Here we focus on two natural laboratories – the Slumgullion landslide, Colorado, which has moved at tens of millimeters per day for centuries, and the East Bay Hills landslides near Berkeley, California, which overlap with the creeping Hayward fault. We apply airborne UAVSAR and satellite Sentinel-1 and ALOS-2 data to solve for the complete three-dimensional (3D) landslide movements from multiple trajectories and imaging geometries. The high-resolution UAVSAR data and appreciable velocity at Slumgullion allow us to extract motion from the azimuth direction of the flight lines by tracking the pixel offsets, in addition to the InSAR range measurements. We investigate the formation of a distinct minislide against the margins of Slumgullion, which is associated with the opening of a 30 by 10 m pull-apart basin and bounding strike-slip faults. We consider the non-Newtonian Bingham model and the power-law model to quantify the intrinsic viscosity of the landslide body, and to explicitly describe the basal geometry and the subsurface mass flux. In the East Bay Hills, we are the first to present the 3D surface kinematics of the slow-moving landslides and creeping faults. The surface velocities change from right-lateral slip of <~5 mm/yr along the Hayward fault to the slope aspect almost normal to the faults at <~40 mm/yr. Discontinuous bands of horizontal extension and contraction suggests potential landslide nucleation zones for more rapid slope failures. For both landslides, we characterize their responsiveness to hydroclimatic variability from rain and/or snow. Our results demonstrate that a combination of interdisciplinary observations, methods and models advance our knowledge of the forcings and kinematics of slow-moving landslide systems.