I am trying to find the total population in 2-mile buffers to find the population 2 miles around primary care clinics. So far I have done this:

  1. Added block group outlines and joined the areas with census data for the population for these areas. Calculated the area in square miles of each area
  2. Added 2-mile buffers.
  3. Intercepted buffers and block group areas.
  4. Calculated area of block groups within the buffers and used this to calculate the new population within each buffer
  5. Use a summary table to display the population for each clinic.

My issue is that the population values are way too high, and when I look at the data in Excel, many of the block group are listed multiple times per clinic. I wonder if this because the buffers are overlapping, as I do not see this for the buffers that have no other overlapping buffers.

What can I do to fix this?

  • Welcome to GIS SE! As a new user be sure to take the Tour to learn about our focussed Q&A format.
    – PolyGeo
    Jul 2, 2018 at 23:03

1 Answer 1


Overall your workflow is sound, but you're right, I can imagine the end result shows populations with inflated values for 1 of 2 reasons:

  1. The reason you mentioned regarding buffers. Without seeing your actual workflow, you may indeed be counting populations multiple times if you have overlapping buffers and then getting population count
  2. If your buffer only partially overlaps a blockgroup, you might be grabbing the entire population values. Instead you want a ratio. For example, 100 people might live in the block group. The buffer may only cover 50% of the block polygon, so the 'correct' answer would be 50 people. (This is best you can do unless you have more detailed information than people per block group)

So how to solve it?

  1. First lets make sure you have good buffered areas. Buffer has a dissolve option on it, but using this creates multipart features and that creates its own complexities. So, run buffer. Then run dissolve and uncheck the create multipart features options. You now have single features to put populations in. Note, this is best guess at what you want to do. There are other ways are handling overlapping buffers. This still has the potential to count people multiple times.
  2. Use Ratios. Before performing Intersect, open and run Make Feature Layer on your block polygons (with population). Inside the Field Info section, click the checkmark beside Ratio for your population attribute. Use the output of Make Feature Layer as input to Intersect (ratio'd population blocks + buffer areas).
  3. Investigate the Summarize Within tool as it handles a lot of this analysis.

If you're interested, I actually just did very similar analysis using arcgis.com and the online tools. Have a look at the end result here: https://www.arcgis.com/home/webmap/viewer.html?webmap=8ffbb4e72da948ec853c59269fbc1044

Methodology (these tools require a named user that has credits):

  • polygon DA blocks with population
  • splash pad locations
  • perform 15 minute walk times from each splash pad. The split option was selected: this means 1 'buffer' that does not overlap another will be created. This assumes an individual would walk to their closest splash pad
  • Use the summarize within tool to get populations per splash pad walk area (this tool automatically does ratio)
  • Derive new locations to get the areas not serviced by splash pads
  • Summarize Within run again to assign population values to the non-serviced locations.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.