methods
data aquisition and representation
I began the project by aquiring the necessary base layers, an outline of the city of Vancouver and UBC, and the Vancouver road network, from the UBC Geography server.I then downloaded a bicycle route layer from the City of Vancouver Data Catalogue. This layer, however, had no connectivity to the Vancouver road network layer, and did not distinguish between types of bicycle routes (shared lane, separated pathway, etc.). As such, this layer was essentially useless for network analysis. In order to create a bicycle route layer with the proper topology, I added a field called BIKE_ROUTE to the road layer and classed each segment into the appropriate category of bicycle route (determined by a higher detail PDF cycling route map). Sections of the bicycle route network that do not follow the road network had to be hand digitized.
The businesses and locations I chose to use as destinations for this project were, for the most part, arbitrarily selected. Maps intended for actual distribution would likely use sposnor businesses as the destinations. The layers Parks and Olympic Sites were downloaded from the Vancouver Data Catalogue. In order to map business locations, I created an address locator with the roads layer. Business addresses were obtained from Google Maps, and location on the map was confirmed with the address locator. Points without addresses or known locations were simply hand digitized using other maps as guidelines.
network analysis
Once I had a full network of bike routes and all of the destinations in place, I then used ArcMap's Network Analyst function to perform a best route analysis. For the purposes of this project, I simply used distance as the cost, and one-way sections of roads as the only restriction. I originally intended to weight the routes according to the type of bicycle route (higher cost on shared lanes and lower cost on separated lanes), but decided this would favour certain corridors, and therefore favour certain businesses. I used the routes determined by the network analysis as the final routes for my output, although upon visual review, they may not make the most sense logically.
uncertainty
There is a fair bit of uncertainty in the data used for this project. Most of the uncertainty stems from the fact that different data sources were used leading to slight discrepancies between layers, seen in the image below.
The digitizing of bike routes and some of the locations on the maps could also be slightly imprecise.
output
The intended output for this project is a set of aesthetically pleasing, user-friendly maps, meant for basic navigation. While accuracy is essential in this application, precision is not quite as vital. As such, the uncertainty in the data should not greatly affect the integrity of the output, but it may lead to slight visual inconsistencies.