Sites for New & Improved Bike Lanes
Throughout the project, we had the hypothesis that Kelowna’s existing bike lanes would not accessible to everyone due to their existing placement. It was a pleasant surprise to find that a significant portion of existing bike lanes falls within areas identified in our analysis as areas that should be prioritized for new or improved bike lanes.
Total of existing bike infrastructure: 369,214m
Total bike infrastructure to be upgraded: 264,769m
Total of bike infrastructure to be created: 40,214m
Future Discussions
Similar analyses can be conducted in other cities for two purposes.
- Determine locations for new bike lanes, or the improvement of existing bike lanes.
- Conduct a social equity analysis on existing bike lanes.
An analysis of this nature may also be conducted on other public infrastructure types. Such as access to critical infrastructures such as hospitals, clean drinking water, and emergency services.
Next Steps
If the City of Kelowna decided to use this analysis in beginning the process of implementing new bike lanes (or improving existing ones), the city should then look at specific onsite conditions (see uncertainty).
Error and Uncertainty
Our methods of analysis are not perfect and are susceptible to uncertainty. While we have used data provided by the City of Kelowna and Statistics Canada to determine areas where bike lanes should be prioritized, we recognize that without looking at specific on-site conditions, bike lanes may not be built.
Point of Interest Data
For points of interest, there are some limitation for our selection of data as well. When we were looking for the shapefile of schools in Kelowna, only elementary, secondary and high schools are available. While universities are not included, our definition of students is limited, and the proposed bike lane does not benefit them as well.
Also, comparing to students from elementary and secondary school, it is likely that high school students can walk longer distance after biking, but our buffer distance is fixed at 400 meters so it may not accurately fit the behavior of everyone
For commercial areas in urban center, we only included the zoning code C4, but according to the Zoning Bylaw of Kelowna, there are many other zoning codes that are defined as commercial and could possibly benefit the general public living in Kelowna as well. However, considering the size of this project as constraint, we decided to limit our zoning code as C4 only, which limit the size of decision set while ensuring the need of general public is being considered.
Demographic Data
We made an assumption that everyone has the ability or means to commute by bicycle. Due to the censored nature of census data, it is impossible to identify who would be unable to determine commute time and method of those unable to use a bicycle, thus we make the assumption everyone can.
In our analysis of commute times we gave priority to areas with lower commute time and high rates of motor vehicle commuting. This helped us suggest bike infrastructure that could encourage the greatest amount of people to change their commute method. While we didn’t account for those who already commute by walking or biking, we assume that any upgrade to Kelowna’s bike lane network will benifit them as well.
For our analysis of potential social inequalities, we grouped all visible minorities into a group who are traditionally socially marginalized. We understand that not all visible minorities experience the same level of marginalization, if at any, but this is not something we can quantify and therefore we assumed that all visible minorities are evenly treated, but marginalized in some form.
Road & Base Data
Our analysis does not consider road width and road shoulder width, which are important considerations when determining feasible bike lanes. At the time this project was developed, a dataset containing this information was not publicly available. It would have also been good to consider traffic counts to ensure bike lanes are placed to maximize rider comfort.
Since we did not have access to proper BC TRIM data for elevation, we had to convert contours to TIN to DEM. The conversion process may introduce error which would have been carried forward to our final analysis.