The final component of my results is the least-cost path I generated from Surrey to UBC, that is attached below.
Figure 9. The least-cost path generated from Surrey to UBC.
As can be seen by the map, the route begins at the bus terminal in Surrey that was located using the MCE analysis. The route then travels across the Patullo Bridge through New Westminister and through Vancouver’s neighborhoods of Renfrew-Collingwood, East Vancouver, Mount Pleasant, Kitsilano and finally to UBC. The entirety of this route spans a distance of 37 km. It is clear, by the land use designations here that this route (as a result of the cost surface I generated) prioritizes residential areas.
Figure 10. A map of the route followed by the now shutdown Surrey to UBC shuttle service (Translink – Retrieved from Daily Hive, 2018).
The route of the now shutdown shuttle service highlighted in Figure 10 on the other hand, moves through agricultural areas of Richmond and avoids a lot of the neighborhoods of Vancouver where a lot of UBC students live. Given that ridership was a huge barrier to success for this shuttle service, the route that I generated might be more useful as it covers a larger amount of areas that people reside in. Unfortunately I was unable to translate the shuttle service route into ArcGIS, but a future iteration of this project could involve this to produce a side by side comparison of these two routes against land use designations. Since this shuttle service has now shutdown, there was very little information that I could find on the route in general.
In terms of the potential limitations of this analysis, I recognize that land use is not the only consideration for a bus route. There are far more factors (age groups, location of UBC commuters, areas of high ridership) to consider and therefore this route is by no means the most perfect route. Another aspect to consider here is that I did not implement any bus stops or consider the factor of time. Thus, if this route was being considered it would be important to include these considerations potentially through the Network Analyst tool.
In regards to potential errors of my analysis, it is likely that through the polygon to raster conversion there was a loss of resolution, that the data I used may not reflect current conditions and therefore may render my results unreliable, and that despite the implementation of the cost surface, the route may still overlap areas that were assigned a higher cost to ensure avoidance. Regardless of the aforementioned limitations and errors, I feel that my analysis is a good starting point for discovering a bus route from Surrey to UBC.