Scott McDougall

A risk evaluation compares a risk with perceptions of tolerable risk, benefits of taking the risk, and available resources and options for managing the risk. It is part of a structured framework for guiding decisions. It can help answer questions like, is the community safe enough from landslides and, if not, how much should be invested to protect it?Since 2005, for residential developments in British Columbia, probability of death due to a landslide, which we call life-loss risk, has been compared with risk tolerance thresholds that are intended to differentiate risks that are too high to be accepted from those that are low enough to be lived with. Local governments have used these risk evaluations to request funding from provincial and federal governments for landslide protection. The risk assessments were intended to provide clear justification for the landslide mitigation investments. But, an analysis of case studies has shown that this intention has not been met. Instead, the resulting landslide mitigation designs have tended to be unaffordable, and the risk evaluations have failed to persuade funding authorities that proposed risk management solutions are a justifiable allocation of governments’ limited resources.Therefore, this research aims to review and improve the methods for quantitative life-loss risk evaluations used for landslide risk management in British Columbia. Its goal is to shape a process that leads to fair, efficient, and affordable landslide risk management solutions that can be adapted for implementation around the world.This research explores both sides of the risk evaluation comparison, including the risk estimate itself and the risk tolerance criteria. It provides a new method for improving the accuracy of life-loss risk estimates based on historical fatality data, and it explores the overlooked and misunderstood aspects of risk tolerance thresholds and conditions for tolerating risks. In conclusion, the research shows that a simple comparison of life-loss risk with a universally applied risk tolerance threshold is usually inadequate and inappropriate. Many other factors control decisions, like economic risks, available funding, and political priorities. It is now time to include these factors more explicitly in our landslide risk management decision process.

View record

Tailings dams are a fundamental component of mining infrastructure as they retain mine tailings, a complex material composed of finely ground rock, water and process effluent. Tailings dam breaches (TDBs) can cause catastrophic tailings flows that travel fast, cover large areas and cause widespread inundation. The ability to understand and predict the motion of tailings flows is a crucial step in protecting people, infrastructure and the environment. This thesis aims to improve predictive empirical and numerical models of potential tailings dam breaches and their downstream impacts to help practitioners develop more reliable inundation maps, dam classifications and emergency response and preparedness plans.To do so, a new tailings flow runout classification system was first developed. A comprehensive database of 33 TDBs was then compiled, and a new volume vs. inundation area relationship was developed for tailings flow runout prediction. Comparisons with similar relationships developed for other types of mass movements indicated that tailings flows are, on average, less mobile than lahars but more mobile than non-volcanic debris flows, rock avalanches, and waste dump failures. The adaptability of four numerical models to tailings flow runout modelling was also explored by conducting back-analyses of two well-described historical TDBs through a benchmarking exercise. The results showed that all four models are capable of reproducing the bulk characteristics of the real events. However, the study also highlighted challenges in the selection of appropriate model input parameter values and the need to develop better guidance on the use of these types of models for tailings flow runout prediction.To address these challenges through improved understanding of numerical model uncertainties and sensitivities, the First-Order Second-Moment (FOSM) methodology was applied to a sub-database of 11 back-analyzed historical tailings flows using the HEC-RAS numerical model. The results showed that the total released volume is among the top contributors to the sensitivity of modelled inundation area and maximum flow depth, while surface roughness is among the top contributors to the sensitivity of modelled maximum flow velocity and flow front arrival time. The FOSM methodology was also used to demonstrate a probabilistic approach to model-based tailings flow runout prediction.

View record

Rapid, flow-like landslides, such as rock avalanches and debris flows, cause human and economic losses around the world. The hazards associated with these events are partly related to the spatial runout extent of the flows, and their depth and velocity at elements at risk, which are highly uncertain. This work details the development of methods to estimate the variability of spatial impacts and the impact intensity with statistical models, and by examining the ranges of outcomes in numerical models.Descriptive attributes and mapping techniques were described, and two datasets of rock avalanches and sediment mass flows associated with rock avalanches were compiled. These data were used to develop statistical models to estimate probabilities for a range of potential impacts. These methods provide a new way to assess rock avalanche runout, and a first method for preliminary prediction of mobilized sediment runout.Large rock avalanches or flowslides can generate signals that can be detected by seismometers. Seismic data were used along with aerial imagery to constrain numerical simulations of rock avalanche events considering multi-stage initiations, examining how variability in the model parameters and initiation conditions affected variability in the runout and intensity. This work revealed the modelled seismic signature of a landslide was highly sensitive to the initiation conditions, showing the possibility of multi-stage initiation is an important consideration in understanding landslide dynamics. Observations of debris flows show complex flow patterns, with surges of high discharge followed by periods of low discharge. This surging behaviour is well known, but has not been incorporated into numerical landslide runout models before. By examining the interactions between topography, model parameters and inflow conditions, it was found that all three of these factors interact and influence the simulation results. Substantial variation in the model results could be achieved by varying the model inflow with all other model inputs held constant. The observed variability demonstrates the importance of considering surging behaviour when estimating debris flow impacts. This thesis demonstrates new methods to estimate landslide impacts considering sources of variability that are observed in nature, and provides a framework for both professional application and future research.

View record

Rapid landslides pose a significant hazard worldwide, and there is currently no routine way of predicting the impact area and velocities of these catastrophic events. Increased development in marginal areas is changing the landslide risk in many parts of the world. There is an urgent need for practical methods to predict the motion of these tragic events to cope with this changing risk. Practical methods currently in use rely on simplified landslide statistics that have a high degree of uncertainty, and are often unable to predict landslide velocities. The focus of this thesis is on developing practical methods to reliably predict the motion of rapid landslides so that public safety in landslide prone areas can be improved.This thesis makes extensive use of runout modelling in order to analyse the motion of rock avalanches, debris avalanches and flowslides. The work presented here can be broadly divided into two categories; the development of new tools and techniques to model flow-like landslide motion, and the compilation and analysis of a database of case histories. The new tools include: 1) A new rheology appropriate for the simulation of liquefied materials; 2) A new dynamic model to simulate the initially-coherent motion of some rock and debris avalanches; 3) Two new calibration methodologies. These techniques were then applied to a database of rock avalanches, debris avalanches and flowslide case histories in order to infer movement mechanisms and give guidance for forward prediction. The main findings include: 1) The character of the path materials is a plausible explanation for the mechanism governing rock avalanche motion. Based on this, a probabilistic framework to predict rock avalanche motion was suggested; 2) A back-analysis of a fatal debris avalanche that occurred in British Columbia in 2012 revealed that this flow was likely moving in an undrained condition, which had significant implications for the analysis of its motion; 3) It was found that flowslides can occur in fine grained colluvium, and this material should be recognized as potentially liquefiable.

View record