Methodology

The Methodology of this project is fairly simple in concept. First and foremost it is a Multi-Criteria Evaluation (MCE). Combining the stated factors (Proximity to public transit, motorvehicle thefts, proximity to parks and vehicle accidents) together, assigning each factor a different weight, should produce an outcome of areas which are desireable to drive and undesireable to drive.

 

Constraints and Factors

Because cars in the city are limited to paved roads the matter of constraints was fairly simple: The areas of study must be limited to roads ONLY. This was done by making the streets layer boolean (1 = roads 0 = Not road) and adding it as a constraint. How factors were processed and treated are shown belown:

 

Proximity to public transit:
For this criteria there were 3 relevant layers: Skytrain Stations, Bus Exchanges and Bus Routes. All three layers were made to be Boolean in nature where Skytrain stations, Bus Routes and Bus Exchanges were all assigned a value of 1 in their respective layers. In this manner we are assessing the level of unsuitability as opposed to suitability. Both bus exchanges and Skytrain stations don't exist ON roads. They do however have an affect on surrounding roads in the form of increased pedestrian traffic and for bus exchanges and some skytrain stations, increased vehicle traffic in the form of buses. Therefore it was neccessary to place a buffer around all these entities. A 250m buffer was put in place as the area where bus and pedestrian traffic is highest around the points. Bus routes needed no buffer because they covered the streets and the area that the project is constrained to.



Proximity to parks:
Parks in Vancouver were in the same situation as public transit points: they weren't ON the roads themselves but the activity that took place at the parks meant it had an affect on roads. This activity takes the form of increased pedestrian activity such as children playing, families, and residents with pets. A 150m buffer was placed around all parks.

 

Automobile Accidents and Motor Vehicle Thefts:
These were the only fuzzy factors in the project. A high number of accidents or thefts meant a higher degree of unsuitability (or lower degree of suitability).

At this point of the project we have assigned accident and theft statistics to neighborhoods and also reclassified a number of factors to be boolean. We have taken these files and standardized the coverage area by using the concatenate and window modules. The image below is a cartographic model that summarizes what we have thus far:

 

Putting it All Together:
Now that we have done all the prep-work for each factor and constraint we can now conduct a multi-criteria evaluation. This was done using the MCE Decision Wizard in IDRISI.

1. The lone objective was to locate the most unsuitable areas for driving and parking in Vancouver.
2. The objective was isolated to city streets and as a result the only constraint was a layer of Vancouver streets.
3. There were 6 factor files to input into the wizard.These factors were:

4. Next we move to factor standardization. The fuzzy factors of Accidents and Motorvehicle Thefts were both monotonically increasing. As shown in the graph on the right, as these numbers rise, the degree of unsuitability rises also.
5. Now we move on to weighting each factor. For this the AHP or Analytical Hierarchy Process was used.
Different configurations of weights were attempted for this process. When the Motorvehicle Thefts and Automobile Accidents were given a high weight there was almost 0 variation in the output layer. This meant that parks and transit could be interpreted as having no effect which in the real world is untrue. This will be discussed more indepth in the Discussion and Conclusion portion. In the mean time, to show variation, and the effects of parks and transit, these factors were given higher weights than the motor vehicle thefts and automobile accidents.

With the weights given and the Wizard complete we can now move on to the analysing the output of the project in the Spatial Analysis page or return to the Top of the Page.