Characterization of Landslide Processes From Radar Remote Sensing and Hydromechanical Modeling

Landsides are a natural geomorphic process yet a dangerous hazard which annually causes thousands of casualties and billions of property loss in a global scale. Understanding landslide motion kinematics from early initiation to final deposition is critical for monitoring, assessing, and forecasting landslide movement in order to mitigate their hazards. Landslides occur under diverse environmental settings and appear in variable types; however, all types of landslides can be mechanically attributed to shearing failure at the basal surface due to stress regime shift contributed by internal and/or external forcing. Typical internal factors include soil/rock weathering, whereas typical external triggering forces encompass precipitation, groundwater, tectonic activity, landslide toe cutting, and landslide head loading. Physically, kinematics of most natural hillslope failures from instigation to cessation can be approached as a hydromechanical problem. Variable types of landslides were examined and characterized in this thesis from integrating satellite/airborne remote sensing and hydromechanical modeling, with the intention to generate insights for reducing landslide hazards globally. In particular, high-resolution satellite optical and radar images and airborne lidar Digital Elevation Models (DEMs) were extensively utilized through advanced quantitative techniques such as sub-pixel offset tracking and radar interferometry in order to capture landslide motion dynamics. Modeling efforts were incorporated to mechanically interpret the observed landslide kinematics and to further generate insights for evaluating and forecasting other similar landslides. The five case studies detailed in this thesis involve multiple distinct landslide behaviors and hydromechanical settings which are typical for many worldwide. Particularly, these investigations entail a consistently slow seasonal landslide, an alternately slow and rapid coastal landslide, a retrogressive and potentially catastrophic headscarp landslide, multiple irrigation-triggered slow and catastrophic landslides, and hundreds of other slow landslides near the U.S. west coast. Knowledge from hydrogeology, soil mechanics, grain-flow mechanics, and fluid mechanics was integrated to decipher and model the impacts on landslide dynamics from bedrock lithology, land uplift, precipitation-contributed pore pressure, groundwater, soil shearing dilation and contraction, and basal topography. From these five landslide case studies near the U.S. west coast, we were able to obtain enhanced understanding of the landslide processes and numerically characterize the key elements from failure instigation to movement evolution, interaction with waterbodies on the runout path, and final deposition. Our investigations particularly demonstrated the potential of integrating satellite observations and hydromechanical modeling to enhance our understanding of landslides and to reduce their hazards. Insights generated from our case studies are applicable to many similar landslides worldwide for their movement characterization and hazard assessment. Hence, this thesis was also aimed to motivate similar efforts globally for landslide hazard mitigation, especially in response to the projected increasingly frequent landslide events in the near future due to global climate change and expanding anthropogenic activities