This page details the methods this project used during our analysis.
Point of Interest Data (POI):
First, we checked the attribute table of shapefiles to ensure the data is correct. Then we used the tool “select by attribute” to locate the areas with zoning code C4 in the shapefile “Zoning” in order to find the Urban Centre Commercial.
Next, we buffer the shapefiles of POI. According to Mehaffy, Porta and Romice (2014), there is a common unit used in urban planning called neighborhood unit, which is approximately 400 meters. Therefore, we set our buffer distance as 400 meters as well. Then we rasterized the buffers with 100 cells as cell size to execute the next step, Multi Criteria Evaluation.
For POI, our team focuses on achieving social equity, so our selection of places are based on fitting the needs of most people.
For general residents, we chose commercial areas in the urban center. Because the urban center is where most people will go when they need something, from grocery shopping to meeting friends at café. Therefore we selected areas with zoning code C4, which is Urban Centre Commercial, hoping the bike lane brings a more convenient route for the general public.
For students, we picked the high schools as that is the place where they visit the most throughout the year. Thus, there is a high demand for better bike lanes that provide them a more convenient trip.
For tourists, we selected parks as the representation. Because most tourists do not come to Kelowna for shopping, but for its natural scenery. Therefore, we included a bike lane as consideration to provide a bike lane that provides better access to those attractions.
Demographic Data:
Social inequality data
To examine potential social inequality, we took census data for the percentage of people within a census dissemination area who spend 30% or more of their income on housing or rent. We also retrieved data on visible minorities in each dissemination area. These data sets would allow us to establish if any dissemination areas might be more socially unequal than others. To do this we ran a multi-criteria evaluation (MCE), with weights of 40% for rent, 40% for housing, and 20% for visible minorities. We put a larger emphasis on cost of living as this can be a direct measurement of inequality and poverty. We looked for areas with higher rates of people spending 30% or more of income and housing, and higher rates of visible minorities. The resulting MCE was factored in our final evaluation.
Commuting data
We wanted to maximize the impact of Kelowna’s bike lanes on commuters. To do this we took census data on commute time and commute method. This data was parsed into shorter commute times of 15 minutes or less, and people who commute by some form of motor vehicle (Car, ride sharing, motorized public transit). People with shorter commute times are more likely to take a bicycle than those who need to commute farther distances, and people who commute by motor vehicle within these areas would be encouraged to bike if they had access to better cycling infrastructure. We ran an MCE to determine places with high rates of short commute times and high rates of motorized transit. The MCE was weighted 50% for each factor while giving priority to the aforementioned commuting rates. This gave us an overview of dissemination areas where a bike lane could have the most impact by turning more motor commuters into cyclists.
Road & Base Data:
Not all roads are suitable for bike lane construction. Established practices in bike lane development call for a maximum road gradient of 3% to suit ensure that cycling paths can be used for All-Ages-and-Abilities (AAA rating). Gradients above 8% would cause most cyclists to dismount and walk. Keeping this information in mind, roads were classified by slope as follows.
| Ranking | Gradient | Suitable for |
|---|---|---|
| AAA | < 3% | All Ages & Abilities |
| AA | < 7% | Experienced Riders |
| A | < 10.5% | Advanced riders only |
| Unsuitable | > 10.5% | Unsuitable for most cycling |
Acquiring Elevation Data
Elevation data was acquired using 1m contour data provided by OpenData Kelowna. The contour data was provided in a DWG file containing both points and polylines. Contours are not useful for the purposes of this analysis. For this project, a DEM is preferred. As a result, contour points were converted to a TIN, which was then converted into a DEM.
Assigning Slope Values
Road data was available both as point and polyline data. Polylines indicated road segments, while the points indicated the start and end of a segment. Each start and end point was assigned an elevation value based on the data from the DEM. Slope values were then assigned to each road segment using the formula below. Since a single road segment may have a positive gradient going one way, and a negative gradient going the other way, the absolute value of the road is taken for classification purposes.
slope = ((end_elevation) - (start_elevation)/(segment_length))*100
Only public paved roads were used for our analysis. Not included in this analysis are lanes and sidewalks. Lanes were excluded due to their primary function to serve utility vehicles and it is prohibited to bike on sidewalks unless marked otherwise.
Final Suitability: Areas of Priority
Once the final demographic and point of interest (POI) data was converted into raster form, a final multi-criteria evaluation was conducted on the data to determine the locations in Kelowna where bike lanes should be considered. Since the objective of this analysis is to ensure access to Kelowna’s bike lane infrastructure is socially equitable, we decided to weight the factors as follows:
| 40% | Demographic Data 20% Commute Distance multi-criteria evaluation 20% Marginalization multi-criteria evaluation |
| 60% | Point of Interest Data 20% Buffer of existing schools 20% Buffer or Urban Commercial Areas 20% Buffer of existing parks |
Each factor was transformed using its default settings.
The MCE yielded a raster with two values. One indicating areas where bike lanes are to be prioritized (based on the aforementioned criteria), the second being areas that should not be prioritized. It should be noted that road conditions and road gradients were not considered as part of this MCE analysis.
Determining New Bike Lanes
To add road data, a data set of road gradients was overlaid on top of the aforementioned MCE analysis. Since the MCE analysis narrowed down the locations where new bike lanes should be built, the project team manually determined the best road segments for bike lane implementation based on proximity and connectivity to the existing bike lane infrastructure.