Variable suction and its effect on stability at the Ripley Landslide near Ashcroft, British Columbia

The high concentration of landslides south of Ashcroft in the Thompson River valley of British Columbia, Canada has periodically affected operation and maintenance of track infrastructure for both of Canada’s primary railway operators as far back as written historical records exist. Historical site investigation and study of the landslides in the region previously identified the primary controlling factors for landslide displacement. However, the arid climate of the region causes a large portion of the landslide head scarps to be unsaturated and minimal study has focused on the consequences of climatic variation that drive soil water content changes in the vadose zone. The research program focused on a heavily instrumented and recently active landslide in the Thompson River valley, known as the Ripley Landslide. The study collected soil samples for unsaturated material characterization, instrumented and monitored the near-surface soil water content changes, compared these changes to annual displacement trends, and developed a 3D modelling framework to establish the impact of variable soil suction on stability at the Ripley Landslide. The research methodology began with a field program to install matric suction sensors, including low-cost dataloggers. Soil samples recovered from the borehole investigation were used to determine the unsaturated material properties. Collection of meteorological observations over several years were interpreted in relation to displacement rates. Records from historical borehole investigations, geophysical surveys, and instrumentation monitoring were incorporated into a 3D model of the Ripley Landslide. Investigation during the research program provided a detailed description of the upper till unit present at the Ripley Landslide. Soil classification and behaviour estimates provided inputs that were vital to the stability model. Interpretations from matric suction monitoring and climatic variables documented their influence on historical displacement rates. New investigation techniques (such as ERT, SMD, and stable water isotopes) provided further evidence for increased soil water content in the head scarp tension cracks that contributed to deeper infiltration. 3D limit equilibrium and 3D finite element models estimated groundwater movement, within a set of known criteria, and determined that matric suction contributed at least 4% to the overall factor of safety. Meanwhile, the river buttressing effect increased the factor of safety by at least 11%. As a result, a loss of suction, coinciding with low river level, was found to be a significant destabilizing factor at the Ripley Landslide. Research contributions from the study improved our understanding of the interrelated factors driving landslide displacement rates and are generally applicable to other landslides in the Thompson River valley. Impacted railway operators may use the knowledge presented in this thesis to identify hazardous conditions leading to increased maintenance and landslide risk throughout the Thompson River Valley rail corridor