This study explores the application of interpolating and non-interpolating spatial prediction algorithms to interpreting shear surface geometries. A number of spatial prediction techniques have been tested, and the most appropriate algorithms for the Downie Slide dataset have been selected based on the root mean squared error (RMSE) determined from cross-validation. Visual assessment of reasonable spatial patterns has allowed for final selection of algorithms that produce geologically realistic results. Through this process, the performance of a number of interpolation algorithms has been tested in terms of accuracy and the development of reasonable spatial patterns. The goal of this study has been: (a) to develop a methodology for interpolating three-dimensional shear surface geometries and (b) to assess which interpolation methods are most appropriate for the interpretation of the Downie Slide basal slip surface geometry, based quantitatively on RMSE and qualitatively on the geological "trueness" of the geometric output.Keywords Shear surface geometry . Three-dimensional landslide model . Downie Slide
IntroductionThe stability analysis of massive landslides and large slope failures commonly use two-dimensional cross-sectional analyses or basic three-dimensional geometries. These simplified interpretations of the shear surface geometry do not adequately represent the true shear surface conditions in terms of shear zone location, thickness and shear strength parameters, making it nearly impossible to calibrate a three-dimensional analysis with respect to spatially distributed slope deformation measurements.Recent improvements to slope stability analysis methods allow analysis of more complex shear surface geometries through threedimensional numerical modeling. This has given rise to the need to develop a methodology for interpreting the true three-dimensional geometry of slip surfaces within large slopes. Interpreting subsurface geometries is difficult due to the inherent data limitations; without expensive and extensive excavation or drilling programs, the exact location of slip surfaces within a massive landslide is usually only defined by a sparse borehole drilling program and the location of the slip surface outcrop at the landslide boundary.
The Westwood mine is located approximately 40 km east of the town of Rouyn-Noranda and 80 km west of the town of Val-d'Or in Quebec, Canada. Operations at the Westwood mine were halted by three large-magnitude seismic events which occurred over two days in May, 2015. This paper provides an overview of the seismic events and summarises the numerical back-analyses completed using global mine-scale and local drift-scale models, to gain an understanding of the mechanisms that fundamentally contributed to what was considered, by a review panel of experts, to be an unforeseen series of events. A global mine model was required to provide a reasonable estimate of the induced stresses on nearfield pillars and excavations and a detailed small-scale model was necessary to adequately capture the failure mechanisms in drift ribs, which played a role in sublevel pillar performance. The global model did not provide sufficient resolution to capture drift-scale mechanisms. While the local model was insufficient for calibration of the mine system boundary conditions, the combination of global and local models is valuable in defining mine-scale stress conditions and their impact on local, drift-scale failure mechanisms.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.