To locate the ideal location for the starting point of my bus route or rather the location of the bus terminal, I conducted an MCE analysis considering the factors of land use and population density. The subheadings below detail this process:
a) Assigning Land Use Costs
In order to use the specific land use classifications in the land use layer as part of my consideration for the location of the bus terminal, I chose to assign costs to each classification based on my research and intuition. Highest values were assigned to those land uses where a bus terminal would not be ideal and low values were assigned to those land uses where a bus terminal would be ideal. Below you will find a table with all 22 of these land uses and their associated costs.
Land Use Classification Cost Agriculture 15 Industrial - Extractive 5 Recreation, open space & protected areas 2 Cemetery 10 Residential mobile home park 10 Lakes, large rivers & water 200 Residential single detached duplex 70 Residential rural 40 Residential low rise apartment 90 Residential townhouse 75 Residential institutional/non-market homes 80 Residential high rises 90 Commercial 80 Mixed res/comm - low rises 95 Mixed res/comm - high rises 95 Industrial 60 Institutional 80 Road - right of way 70 Port Metro Vancouver 5 Airport, airstrip 2 Rail, rapid transport & other transport 80 Undeveloped & unclassified 10
Table 1. Land use costs for Surrey.
b) Polygon to Raster
I began by converting the polygon layers of both population density and land use to rasters in order for me to normalize them and conduct the weighted analysis. I did this by using the polygon to raster conversion tool and selecting cost for the land use layer and population density for the dissemination areas layer as the input features.
c) Normalization
Next I used the fuzzy membership tool to normalize both the population density and the land use rasters in order for them to encompass only values that range from 0-1. Here I used the linear membership tool and assigned the appropriate minimum and maximum values.
d) AHP
Upon normalizing both the land use and population density factors, I was able to determine the weights for them using the analytical hierarchical approach using the website www.123.ahp.com. Below you will find an image of the weights I retrieved from this analysis.
e) Weighted Linear Overlay
Using the weights from the AHP, I then used the weighted sum tool and filled in these weights for my two factors. The resulting raster provided weights (or values) from 0-1 indicating the feasibility of areas within Surrey for locating the bus terminal.
f) Sensitivity Analysis
In order to assess how a different normalization approach would affect my results I conducted a sensitivity analysis. I did this using the equal weights model we used in Lab 2. Below you will find an image of the same model adjusted for my analysis.
g) Isochrone maps
In order to determine which location I should choose for locating the bus terminal from the results of the weighted analysis, I chose to use an isochrone map generator by the name of Travel Time Platform which generates maps to show areas that are reachable within 30 minutes by transit (in my case). Doing this allowed me to understand which location was the most reachable for commuters in other regions and therefore would the most beneficial. Since I was unable to figure out how to translate this map to ArcGIS, I chose to use the maps generated on the site itself.
