Measurement of walkability variables
Measurement of Residential
Density
Residential
land-use per CT is selected from the land-use database and the number
of dwelling units per CT is isolated from the Census 2006 data. Net
residential density is then calculated by dividing the number of dwellings
per CT by the residential area of the CT. Map: Residential Density
Measurement of Street
Connectivity
Areas with higher densities of intersections are assumed to offer more
destinations for pedestrians (i.e. shops, restaurants, parks, schools)
within walking distance from home (Ackerson 2005). Additionally, the hierarchy of
intersections (i.e. dead-ends, 3-way/ 4-way intersections) within a study area
is a proxy for the variety of route choices available to pedestrians in a given
area. As a lot of dead-ends require people to take longer routes to their
destinations, they reduce the pedestrian through-traffic and are less walkable.
3-way intersections are an indication of ‘moderate’ walkability and four-way
intersections are an indicator of ‘good’ walkability (Ackerson 2005).
Road intersections
are identified from the
streets dataset using the “Network Analyst”-tool, and connectivity is
based upon the number of connections at each intersection with
neighbourhood streets and major streets. Intersections with highways or
highway ramps are excluded as the are considered not walkable. The
intersections are spatially joined to the CT in which they
lie.
In the
intersection density calculation, only intersections with three or more
unique intersecting streets are included. Different intersections
were identified depending on whether an intersection is a three-way
intersection (moderate walkability) or a four-way intersection (good
walkability) and depending on whether an intersection
only connects neighbourhood roads (good walkability) or also major roads
with a high traffic volume (only moderate walkability). The density is based on
the number of intersection per CT area. The results of the
density calculations are classified into deciles and normalized on a
scale from 1 to 10. The final intersection density of a CT is created
by summing the different densities with a higher weight given to
four-way intersections and intersections between neighbourhood roads. Map: Street Connectivity, GVRD
Measurement of
Land-use Mix
The land-use mix
measure, or entropy score,
was complex to create as the accessible land-use data is grouped into
very broad classes (residential, commercial, governmental and
institutional, resource and industrial, parks and recreation, and open
area). As the GVRD does not only contain urban, but also less populated
areas, non developed areas (parks and recreation, open area) encompass
large areas, especially at the CTs located at the edges of the GVRD.
These “non developed” areas are not suitable for measuring a
land-use mix that encourages walking. However, entropy score
calculations just
reflect the mixture of land-uses, not the land-use categories which are
contained in that mixture.
To overcome
the problem of large open areas and the potential to skew the land-use
mix calculation, the land-use mix measure was based only on the
developed area. Developed area was defined as the sum of the areas
belonging to the following land-use categories: residential,
commercial, governmental and institutional, and industrial. In GIS, the
land-use parcels are geometrically intersected with the CTs in order to
assign them to the CT in which they are located. The area covered by
each land-use category within a CT is calculated using a Pivot table in
a spreadsheet application. Based on this, entropy land-use scores are
created for each CT, calculated via the following formula:
k is the category of land-use, p is
the proportion of the developed land area devoted to a specific
land-use, N is the number of land-use categories in a CT.
The results
of the entropy calculation range from 0 to 1, with 0 representing a
completely homogeneous area (the CT is covered by one single
land-use type) and 1 representing heterogeneity (all land-use categories
covering the area of a CTs are equally distributed). Map: Land use, GVRD
Map:
Land use mix, GVRD
Measurement of Topographic
Variation
The DEMs covering the GVRD area were converted to polygons and the polygons
were spatially joined to the CTs. The Standard Deviation of
elevation values within a CT is measured in order to characterize
variations in terrain. As the GVRD area also covers parts of the
North Shore Mountains, the obtained values were not classified into
deciles, as this would have biased the classifications scheme. Rather, all CTs with an extremely high standard
deviation were group into one class. Map: Topographic Variation Measurement of Access
to Public Transport
A core
component to the success of a city development oriented away from
car-dependency rests in the capacity of residents to access transit
stops. Schlossberg and Brown (2004) found that people are most likely
to walk to transit stops if the stops are within a five minute walk. Generally, this equals a distance of 400 m.
Buffers of
400m are created around each busstop. In order to minimize the
influence of the differences in CT size, the buffers are clipped to
the developed area within each CT. Then, the proportion of the
developed area per CT which falls in this buffer is calculated, ranging
from 0 (no bus stop buffer in the developed area of this CT) to 1 (developed
area of this CT is fully covered by buffer around bus stops.
Map:
Bus stop buffer, GVRD Map: Access to public transport, GVRD
|
