The road network and defined bike route shape files where downloaded from the City of Vancouver’s open data catalog, along with the data regarding the volume of riders on each of the separated bikeway routes. This data was presented as number of riders per day, and volume amounts are available at the time of installation for each route. In the shapefile of bike routes, along with separated bikeways, there are other designated bike routes. These are designated as a bike route, but travel is with traffic on the roadway, or in a bike lane within traffic that doesn’t have a barrier. Because traffic isn’t separated on the other bike lanes or before the installation of barriers, it is not possible to differentiate types of traffic with the counting machines used to tally volume. Therefore there is no data of equivalent quality before a separated lane makes it possible to count ridership to this extent. The City of Vancouver reports that from 2008 to 2011, from before to after the installation of separated bikeways respectively, ridership increased by 40% in the city. This statistic was used to work out estimated amount of riders that would have been expected on each of the four separated routes before their installation.

 

The accident data was obtained from ICBC, as a data set of all accidents involving a bicycle in British Columbia from 2008 to 2012, with latitude and longitude locational references. The data also described the location of the accident more precisely, either with street intersection names, or a location on a block segment of a street.

 

The accident data was narrowed down to just represent accidents that occurred in downtown Vancouver, and organized by year. The accident data was projected and added as a vector layer of points, where each year that was used in the analysis was organized as a separate layer.

 

Considering my data constraints, I did my analysis using the year 2008 in comparison with two years after the complete installation of the separated lanes, so 2011 and 2012. I had limited information on bike ridership volumes, for more years or routes, past general trends of increases.

 

The bikeways shapefile was modified so that Burrard was split into separate sections for the street and the bridge to represent the location of the separated lanes. As well, Dunsmuir was also split to accommodate the two data sets of ridership for the routes of Dunsmuir Street and the Viaduct provided by the City of Vancouver.  The volume data was then joined to the appropriate bike lane.

 

SQL statements where then used to organized the accident data by which separated bike route they occurred on. This data was separated by location and year, and then summarized into separate tables to represent the sum of accidents per year on each of the routes. This data was then joined to the appropriate bike route, to create a complete chart of volume of riders and amount of accidents, by route for each of the 3 years of interest.

 

I then calculated the relative rate of accidents for each route by comparing the amount of cyclists and the amount of accidents. The results gave the percentage of riders who where in an accident with a vehicle, which varied between 0.000233 - 0.004451%.

 

The next step was to compare the change in rate of accidents from before the separated bikeways to after. Both 2011 and 2012 where compared with 2008 to show a percent change in the rate of accidents. The rate displayed the change of accidents in 2011 or 2012 as a percent of the 2008 figures.

 

The initial data set of accidents by year was converted to a raster so that intensity of accidents could be visualized by location. This was mapped along with the rate of change of accidents to show insights to the patterns of accidents.This has the ability to highlight hotspots for cyclist related accidents and visually compare rates of accidents and what locations may be causing a hight rate of accidents.