Service area analysis:
Assessing equity in spatial
accessibility of veterinary facilities
The second exercise in the course on GIS for Transportation related transportation to accessibility to public and commercial services. Access to services is a vital prerequisite for many daily human activities and needs, regardless of place of residence or socioeconomic status. The aim of transportation is to provide access to key services to as many people as possible, but if the social implications of inaccessibility are not mitigated by more efficient transportation, it is likely to become a factor of social exclusion and segregation.
Specifically, in this exercise, the task was to ascertain whether there are population inequalities in accessibility to public and commercial services in a study area of your choice in Västra Götaland County. Using ArcGIS Pro, the basic workflow was to create a series of service areas based on my previously created multimodal network dataset (see earlier coursework: Building a Network) and the public/commercial services of my choice and relate this information to the socioeconomic and demographic composition of the population in each service area. In doing so, I was then able to compare the results among the different bands of the service areas, as well as uncovered zones, in my study area, and determine whether the characteristics of people differ according to accessibility to different services.
In my case, I chose to look at veterinarian facilities for this exercise. Specifically, I wanted to see how many dogs are registered per municipality in Västra Götaland and whether there exists an inequality between accessibility-by-proximity and accessibility-by-mobility of veterinary services in the county. However, my analysis was only done using one travel mode: the car. Hence, veterinary services close by or far away were equally considered by driving a car on the road network, regardless of whether or not other travel modes (i.e., walking, cycling) were better for accessing the service.
The map to the right shows the service area of and around each veterinary facility in the county with respect to the layout of the road network. Each color represents an interval of travel time by car towards the nearest facility, where dark green = 0-15 minutes, light green = 15-30 minutes, yellow = 30-45 minutes, orange = 45-60 minutes, and red = 60-82 minutes driving time (82 minutes being the longest travel time needed to reach the nearest facility).
With that, I joined a table containing the number of dogs registered in each municipality to the service areas and calculated the statistics of the spatial relationship between the two.
Specifically, I wanted to calculate the sum of dogs per service area, % of total dogs per service area, and the mean number of dogs per service area (in hindsight, calculating the mean proved to be a meaningless statistic). However, before I could do that, I had to join the dog registry table to the municipal boundaries layer. Exporting the joined layers simply created a new point feature class with all of the attributes centered in the geographical middle point of each municipality (see map below to the left).
Surely, the geographical middle point of each municipality is not an accurate representation of the true distribution of where the dogs are registered. For that, I would have needed data showing the address of each and every dog, or at least the total number of dogs per urban area. However, this sufficed to depict in a simple manner the general accessibility of veterinary facilities for each municipality. From there, I could use the ‘Spatial Join’ tool in ArcGIS Pro to join the dog registry point feature data to the service area interval polygons and calculate the statistics from there. The map below shows the final result.
The final result looks almost identical to the map above but includes contextual information such as the municipal boundaries, and the statistics presented in a three-column legend list.
The conclusion is that there does not seem to be any noticeable inequality in the accessibility of veterinary facilities by car in the county. Most dogs reside within a driving distance of between 0 - 15 minutes (38,169 dogs or 25%) and 15 - 30 minutes (87,319 or 57%) away from the nearest facility, amounting to a total of 125,488 dogs or 82%. The service areas furthest away, i.e., 45 - 60 minutes and 60 - 82 minutes away, are indeed at a disadvantage, but only if we assume that they do not travel to the nearest facility in a neighboring county. Therefore, since the veterinary service area coverage is quite good throughout Västra Götaland, it is reasonably safe to assume that the same can be said about neighboring counties and that the remaining 6% of people could simply travel to another county instead.
