Figure 9: Final Suitability Raster from Multi-Criteria Evaluation
Table 4: Zonal Statistics for each of the proposed extensions.
The three lines proposed by Bernardo, Kiralyfalvy and Smith (2015) included an Eastern Scarborough extension, a Humber extension, and an Etobicoke extension. Since there is a limited amount of funding allocated towards transit each year, building three lines is very improbable in the short term (Kyte, 2011). Therefore, research was conducted to determine which of these three lines is best suited for the City of Toronto. An Multi-Criteria Evaluation was conducted using factors which were deemed important to subway line siting by relevant literature, in order to determine the most beneficial subway extension.
Figures 3 to 8 found under Objective 3 represent the reclassified factors associated with determining a new subway extension, and display areas of higher suitability in dark orange. Figure 3 represents population by ward. The higher the population the greater need there is for a new subway line. Therefore these higher population areas are the areas of higher suitability (Toronto Transit Infrastructure Limited, 2012; Burda & Haines, 2011). Income by Toronto ward is represented by figure 4. In this figure the areas of higher suitability are correlated with areas of lower income, as areas of lower income would benefit the most greatly from the implementation of a new subway line (Kyte, 2011; Burda & Haines, 2011). The cost of building each extension is depicted in figure 5. Strictly with regards to cost, the more expensive the line the least suitable it is to build. This is due to the limited amount of government subsidies allocated to transit (Toronto Transit Infrastructure Limited, 2012; Burda & Haines, 2011). Proximity to existing subway lines is represented in figure 6. Areas in closer proximity to existing transit were deemed less suitable due to the fact that transit is already readily available to these areas (Calvo et al., 2013; Burda & Haines, 2011). Proximity to points of interest is represented in figure 7. Places of interest in Toronto attract many tourists from around the world. By providing transit to these places, ridership is expected to increase (Wall & Sinnott, 1980). As such, areas closest to the points of interest are more suitable than those which are further away. Lastly, proximity to major intersections is represented in figure 8. Higher suitability was awarded to areas closest to busy intersections. The implementation of a subway line along these areas would reduce the foot traffic at these intersections (Mackett & Edwards, 1998).
The final suitability took five of the six factors into consideration. Population, income, proximity to subway lines, proximity to major intersections and proximity to points of interest were considered. The only factor not considered in the final suitability raster was cost. However, cost was considered during the final stage of selecting the most suitable line. Since the cost factor was simply a per kilometer cost, it was evident that the longest line would be the most expensive and the shortest line would be the least expensive.
A multi-criteria evaluation was used to analyze the data. The main factors associated with building a new subway line which were identified through literature were included. Using the Raster Calculator and equation 3, all the criteria and weightings of importance were considered. Figure 9 above illustrates the output of the raster calculator which is the final suitability raster. From this output, the three areas were compared visually. With red areas on the final suitability raster representing more suitable areas to build a subway, and green representing less suitable areas. The upper right where the Eastern Scarborough extension was proposed is mainly red. The Etobicoke extension is a mix of yellows and reds. Lastly, where the Humber extension is proposed is mostly yellow and green. Therefore, visually the Eastern Scarborough extension is clearly the most suitable extension.
To validate this decision, the Zonal Statistics function was performed within ArcGIS, the results of which are displayed in Table 4. The results indicate that the Eastern Scarborough extension had a mean value of roughly 3.9, followed by the Etobicoke extension at roughly 3.7, and finally the Humber extension at roughly 3.1 (Table 4). The mean values represent the average suitability score (on a scale of 1 - 5) allocated to each line in terms of the cells that these lines interact with. By this logic, an area with a higher mean value also has a higher suitability score. The sum value associated with each line represents the sum of the suitability scores of each intersecting cell. The highest value is approximately 463 for the Humber Extension, followed by 417 for the Eastern Scarborough extension and 351 for the Etobicoke extension (Table 4). The sum values are strongly correlated to the length of each line, as the longer the line is the more cells that it intersects with. Therefore, despite the Humber extension having the highest sum, it is the longest line, which also results in it having the the highest cost of construction. The mean of this extension was also less than the Eastern Scarborough extension. The Etobicoke extension had the second highest mean suitability score, and was also the least expensive option. Therefore, in a situation where cost is considered more important than achieving the highest suitability, the development of the Etobicoke extension would be a logical choice. However, given the statistics and visual interpretation, the Eastern Scarborough extension was deemed to be the most beneficial extension overall, as it scored the highest overall with regards to suitability, and is also not the most expensive option.