Probabilistic seismic landslide mapping for western Metro Vancouver, British Columbia
Abstract
The annual direct and indirect costs related to landslides in Canada are estimated to be $200 million. The west coast of British Columbia (BC) has experienced the most landslide-related fatalities within Canada considering its mountainous terrain and unique physiography. In the urbanized Lower Mainland of BC, many residential areas extend to the edges and bases of escarpments. Even small landslides in these locations can damage houses, roads, and other structures. Combined with the high seismic hazard of the region, seismic slope instability becomes a significant geotechnical hazard in the Lower Mainland.
Seismic slope failures are predicted in practice using seismic displacement prediction models (SDPMs) based on a Newmark sliding block analogy. Seismically induced permanent displacements are calculated using earthquake hazard and soil strength parameters represented by yield acceleration of the slope (ky). This thesis presents a probabilistic solution for the seismic sliding displacement of slopes for the Metro Vancouver region considering the multiple seismicity sources and the latest updates in SDPMs. The uncertainties in input seismic parameters and SDPMs are both taken into account, and probabilistic displacements are determined for different values of ky and the initial predominant frequency of the sliding mass (fs). Regression analysis is performed to develop a regional predictive model to estimate probabilistic displacement values at a 2% probability of exceedance in 50 years hazard level for different slope conditions (i.e., ky and fs values) across Metro Vancouver. High-resolution topography data is used to construct semi-automated polygons to capture slope geometries. The regional displacement models are employed to assign the corresponding seismic displacement values to slopes, and the first probabilistic seismic landslide hazard map for Metro Vancouver is generated.