Based off the analyses conducted, there is not a significant risk of landslides due to earthquake activity in the District of Squamish (the District), but there is an increasing risk due to rainfall.
Model Limitations
The maps produced are only as accurate as the data inputted. Recognizing that representative values were chosen for the majority of the parameters, these maps should only be taken as a rough guideline and should definitely not be considered definitive. For more detailed analysis, more work should be done looking into the ground parameters in the District of Squamish, as well as the effects of root cohesion on friction angles. Additionally, this analysis depended on the accuracy of the TIN upon which the DEM was based.
The empirical methods used were developed in California, which has different conditions than Squamish. It is closer to fault zones and the soils have a different depositional history and different geotechnical properties. There may be some differences in the models given the different conditions found in Squamish.
To assess slope stability due to rainfall we used a deterministic model whereby slopes were considered unstable if their driving forces were greater than their resistive forces. Reality is more complex – some slopes may remain stable if their Factor of Safety is less than one whilst others may fail if they have a Factor of Safety greater than one. To supplement our analysis we could have also used a probabilistic analysis which would produce results that give a slope’s probability of failure.
Lastly, this projected was limited by the amount the amount of time spent on the project. With additional group members, a more comprehensive analysis could have been conducted.
Further Work
As touched on when describing the limitations, this analysis could be improved with model calibration. This would work by performing an inventory of historical landslides in the area, then adjusting the empirical formulas to match the observed model.
Another important aspect that was not considered in this analysis was the modelling of debris flow paths. There are now some quite simple methods on how to do this. TauDEM is a free hydraulic modelling toolbox developed by the Hydrology Research Group at Utah State University. You can model the runout of source cells (in the case of this analysis, unstable cells) using a D-infinity flow model and finding the Downslope Influence (DI) of the cells.
Additionally, this model could be run under more severe rainfall conditions such as a 1:50 year or 1:100 year return period.
Conclusions
The maps produced can be used by the District of Squamish for land use planning and to prioritize slope stability work. The District may wish to develop a risk assessment classification scheme whereby they classify which risks are acceptable per different land uses. For example, an industrial area with few workers may be acceptable to place below a slope that is prone to failure with a 25mm/d rainfall, but an apartment building with lots of people would not be acceptable.
In addition, the District can use these maps to plan their slope stability work. If there are areas that are unconditionally unstable above high-risk areas, the District could perform slope stability work in those locations before moving to lower risk locations.
Lastly, I should reiterate that landslides are highly variable and dependent on their site conditions. As the site conditions change, so does the probability of a mass wasting event. Factors such as loading on the slopes and construction at the base of the slope also affect the slope stability. This analysis is only effective at the small scale, and the results should be interpreted as such.
I’d like to thank Dr. Scott McDougall, P.Eng., Dr. Brian Klinkenberg, and Kathryn DeRego for their guidance in this project and for their work throughout the GEOB 370 course.