Walking distance is one of the most important factors that affect how walkable an area is and the creation of travel time and distance isochrones are an important tool for identifying these areas. In order to create these it’s first necessary to create your own network dataset in ArcGIS Pro or use a commercially available such as what is available through ArcGIS Online for a fee. While it’s not difficult to create these networks for driving as many municipalities publish datasets of their roads, these may not include the necessary features such as walking and bicycling paths that are required to create a network dataset that can represent modes of transportation outside of a car. However, by looking beyond the tools in ArcGIS Pro it’s possible to also use crowdsourced, free open data through OpenStreetMap for the same purpose and more as long you can credit their project.
Identifying area within walking distance to schools can provide important information to homebuyers looking to raise a family, this information is even more useful when combined with school boundaries. The intersection of the areas within walking distance to an elementary, middle and high school can also serve as a starting point to identify areas where infrastructure improvements to improve pedestrian safety can be focused. For students that walk to school this may also reduce stress, and improve health outcomes from increased physical activity (“Walking to School May Decrease Stress Reactivity, Improve Cardiovascular Health” 2010).
A mix of open data from both government sources and crowdsourced platforms were used: road network was created OpenStreetMap along with the location of schools while boundary files were acquired through government open data portals. Locations of Schools was collected from OpenStreetMap data by using Overpass Turbo, a tool for querying and extracting information from OpenStreetMap (n.d.), this was the crosschecked against information available from the Cupertino Union School District to schools that were closed or did not admit students purely on residency (2023).
Isochrones were created using OpenStreetMap data through Openrouteservice, which is both an online API and opensource routing engine capable of generating transportation isochrones similar to ESRI’s Service Area Analysis. Once created, these were clipped to each school’s attendance boundaries and the boundaries of the city (“Open Data Portal | City of Cupertino, CA” n.d.). To identify areas that were only within walking distance to an Elementary, Middle and High School the intersection of these were found. Finally, the using datasets from the U.S. Census Bureau the number of housing units within a walkable range was quantified (Bureau n.d.).
The schools with the highest number of housing units in the City of Cupertino that were within walking distance for each school level are Garden Gate Elementary (1,576 units), Lawson Middle (4,806 units) and Monta Vista High (8,106 units). Combined, a total of 4,397 housing units were within walking distance of an Elementary, Middle and High school.
| Elementary | Middle | High | |||
|---|---|---|---|---|---|
| Name | Housing Units | Name | Housing Units | Name | Housing Units |
| Garden Gate | 1,576 | Lawson | 4,806 | Monta Vista | 8,106 |
| Sedgwick | 1,389 | Kennedy | 3,301 | Cupertino | 6,790 |
| Eaton | 1,330 | Hyde | 2,164 | Homestead | 1,403 |
| Lincoln | 1,141 | ||||
| Collins | 1,106 | ||||
| Stevens Creek | 946 | ||||
| Total | 7,488 | 10,271 | 16,299 | ||
| K-12 Walkable | 4,397 |
The creation of travel distance isochrones proved to be the most difficult task primarily as result of the need to create isochrones to each entrance to the school rather than a single point as many schools have entrances that do not correlate one to one to their street address, rather providing access along adjacent streets to the side or behind school. For example, while across the street from an entrance to Monta Vista High School Google Maps will create a route that is 0.3 miles long to the front entrance which would lead to any isochrones not being accurate representations of the distance in the area where this was the closest entrance (n.d.). After creating entrances to each school through both personal knowledge and aerial imagery provided through the City of Cupertino 89 school entrances were identified.
The number of entrances then would prove a problem as three isochrones would be created for each point equating to 135.5 credits. For this reason, an alternative tool was sought to create the isochrones and Openrouteservice was decided on as it provided the option to self-host the service without worrying about API limits. While this wasn’t a consideration, using a self-hosted instance would allow for greater reproducibility as the road network would be frozen in time and would not be updated automatically like an online service would.
The use of crowdsourced information has the advantage of being low cost and can contain local knowledge absent outside of other sources. However, as it is not from an authoritative source it’s also necessary to crosscheck the data where possible. In the case of school locations, this is also published elsewhere such as the California Department of Education but for compiling a dataset for places such as stores it can be an alternative to accessing costly services that provide this information; at the very least serving as a starting point. Openrouteservice and another opensource routing engine, Valhalla may be useful to self-host in applications where the costs become excessive to use ESRI’s API service for generating isochrones or routes.