Spatial Access Measures

Over the last year, Statistics Canada and Infrastructure Canada collaborated on the implementation of a set of spatial measures of access to services and amenities using active and public modes of transportation. The purpose of this collaboration was to generate data and analytical work in support of more sustainable and resilient active and public transportation systems across Canada.

There are seven types of amenity categories within the Spatial Access Measures: primary and secondary educational facilities (EFs), postsecondary educational facilities (PSEFs), health care facilities (HFs), places of employment (EMPs), grocery stores (GSs), cultural and arts facilities (CAFs), and sports and recreational facilities (SRFs). For each amenity, there are four variants based on the transportation mode: access via public transit during peak hours, access via public transit during off-peak hours, access via cycling and access via walking.

Downloading the Spatial Access Measures

Spatial Access Measures data

The Spatial Access Measures contain 28 measures of access for 7 types of amenities and 4 modes of transportation combined.

Concise description of the data and methods

The Spatial Access Measures are a set of indicators that quantify the ease of reaching destinations of varying levels of attractiveness from an origin dissemination block (DB). DBs are the smallest census geography, about the size of a block in urban areas and considerably larger in less urban areas. There are seven destination amenities, which include EFs, PSEFs, HFs, EMPs, GSs, CAFs and SRFs. For each amenity, there are four variants based on the transportation mode: access via public transit during peak hours, access via public transit during off-peak hours, access via cycling and access via walking.

The travel times via public transportation were calculated using r5r^Note , an R package that accounts for the transit stops and schedules provided in the collected General Transit Feed Specification data. Meanwhile, travel times via cycling and walking were calculated using Valhalla. To calculate the distance travelled between two DBs to generate the travel time matrices, a representative point was calculated based on the building footprint in each DB located closest to all other buildings in that DB. The building footprints used for this calculation are a compilation of footprints from the Linkable Open Data Environment (LODE), OpenStreetMap (OSM) and Microsoft.

The gravity model adopted for this analysis builds on the methods developed by Alasia et al. (2021), who produced the first Proximity Measures Database for Canada.^Note The gravity model is based on the principle that the likelihood of interaction between two locations is proportional to the attractiveness (mass) of the destination and inversely proportional to the distance or duration of travel between them. The distance or duration can be further transformed using an impedance or distance decay function to account for the variation in willingness to travel to different types of amenities and use different modes of transportation. Finally, the level values are rescaled into an index by min-max normalization; that is, the minimum value for all Canada is set to 0 and the maximum value for all of Canada is set to 1. The output of the dual model, which is used exclusively for GSs, is the time it takes to travel in minutes to reach the nth closest GS.^Note

For CAF, EF and PSEF destinations, the total mass of a DB is the sum of the masses of the facilities within it (i.e., count of observations). The masses were derived from the LODE and assumed to be uniform, with each destination given a value of 1.

For SRF destinations, a subset of OSM data was used. The masses of SRFs were derived from the processed OSM data and assumed to be semi-uniform; that is, while each SRF was given a count value of 1, areas with multiple digitized SRF features were scaled to not completely outweigh single features (i.e., fewer than four features were weighted as 1, four or five features weighted as 2 and over five features weighted as 3 ). The total mass of a DB is the sum of the masses of the facilities within it (i.e., count of weighted observations).

The masses for GSs were derived from a filtered subset of GSs (North American Industry Classification System [NAICS] 44511, supermarkets and other grocery [except convenience] stores) maintained by the Business Register (BR). BR grocery stores used for this project were assumed to be uniform in size; that is, each grocery store was given a value of 1. The total mass of a DB is the sum of the masses of the facilities within it (i.e., count of observations).  

The masses of EMPs were derived from the employment counts of all businesses maintained by the BR. Employment counts were partitioned into eight bins for which the numeric categorical code, ranging from one to eight, was used as an EMP mass. The total mass of a DB is the sum of the masses of the businesses within it.

The masses of HFs were derived from a subset of HFs (NAICS 6211, offices of physicians; 6212, offices of dentists; 6213, offices of other health practitioners; 621494, community health centres; and 622, hospitals) maintained by the BR and supplemented with data from the Open Database of Healthcare Facilities. Employment counts were partitioned into eight bins for which the numeric categorical code, ranging from one to eight, was used as a mass. The total mass of a DB is the sum of the masses of the HFs within it.