Clarifying Service Area Analysis for Historical Walking Distances in ArcGIS Pro

[emointheworld]

Hello Esri Community,

I'm working on a project that involves generating 15-minute walking time isochrones using the service area layer in network analysis with ArcGIS Online data (using credits). My primary question centers around the historical traffic model: Is it designed to be applicable for pedestrian (walking) analysis as well, or is it solely for vehicular traffic?

To delve into this, I've conducted tests under various scenarios with Date and Time settings, including: 1) not using any time parameter, 2) setting a specific date (12/25/2010), 3) choosing a specific time (Wednesday at 12 PM), and 4) using the current time. In all these scenarios, I found no differences in the service areas generated, which makes me wonder if this is due to the nature of my data or an inherent characteristic of walking isochrones.

This leads to my main inquiry:

• Can historical traffic data in ArcGIS be effectively used to analyze historical walking service areas, or do you recommend alternative methods, particularly in the absence of historical road network data?

Also, is it a general assumption that walking isochrones don't change significantly over time?

Your insights or suggestions based on your experiences would be immensely valuable for my project. Thank you in advance for your guidance!

[MelindaMorang · Accepted solution]

The impedance attribute for walking time is just based on distance and a presumed constant walk speed (which you can change by creating a custom travel mode).  Historical traffic is only used for driving modes and has no impact on walking.  Setting a time of day also has no impact on walking, since the underlying calculations do not make use of any time-dependent attributes.

The ArcGIS Online data is updated multiple times a year based on the data we get from our data providers.  The walking time and distance will only change when the data is updated to add or remove roads (geometry changes) or to change whether certain roads as restricted or preferred for pedestrians (attribute changes) either because the status of the road changed in reality or because the data provider obtained some new information or corrected an error.

[emointheworld]

Thank you for your response! I'm analyzing walking accessibility to parks in 2000, 2010, and 2019. Considering that walking times are not affected by historical traffic data, I'm contemplating whether to use specific historical dates (like 1/1/2000, 1/1/2010, 1/1/2019) to potentially capture changes in road networks or pedestrian paths, or to simply use a 'no time' setting across all years.

Would setting specific historical dates provide any significant insights given the nature of updates in road geometry or pedestrian path attributes in ArcGIS Online data? Or is a consistent approach without time settings more advisable?

Any recommendations or insights would be greatly appreciated. Thank you!

[MelindaMorang]

I think I may not have fully explained the update process for the ArcGIS Online data.  We update the data used by the service several times a year, but the historical data is not maintained.  If you solve using ArcGIS Online, it will always use the current data, even if the date you enter for your analysis is from 2000 or 2019 or whatever.  It will not draw from any archived data that was applicable in those years.

Ultimately, if you are interested in modeling the changes in infrastructure between these years and how it impacts walking access, I don't think ArcGIS Online is the right tool for this job.  Instead, you would need to find or create network datasets using street features that represent the state of the infrastructure in those years and run the same analysis (same input points, etc.) using each of those datasets, and compare the results.