Buffer tool vs. Select by Location > Within a Distance

287
9
3 weeks ago
Craig_Eissler_Iceman
Occasional Contributor III

Hello,

When I count the number of points within a specified buffer using the Buffer tool vs. a spatial query (Select by Location > Within a Distance), I get a different answer. In this case, it's 24 compared to 22 points, respectively.

I was expecting the results to be the same. Why are they not?
Or, does it have something to do with geodesic vs. planar measurement modes, "if" those defaults are different?

Thanks, CCE

0 Kudos
9 Replies
Dale_Honeycutt
Occasional Contributor II

The difference between Planar and Geodesic could definitely account for the difference! Run both tools with the same option.  For Select By Location, choose "Within a distance (geodesic) for the Relationship parameter and for Buffer choose "Geodesic (shape preserving" for the Method parameter.  You should get the same number of features within the distance you chose. 

Here's a post on basically the same subject:

https://community.esri.com/t5/arcgis-pro-questions/geodesic-option-on-points-yields-incorrect-buffer...

 

Craig_Eissler_Iceman
Occasional Contributor III

Thanks, Dale. I tried this, and yes, I got the same result using the 'Geodesic" method when using the Buffer tool/Select by Intersect tools and then just a Select Within a Distance tool.

But when BOTH tools are set to Planar (Euclidean), I get different results. ...Why?

...By the way, the Input points are in projected coordinates (UTM), and the Distance used is 500 feet.

0 Kudos
Dale_Honeycutt
Occasional Contributor II

I don't have an answer for the differences when using Planar.  Can you upload your data so I can take a look?

0 Kudos
Craig_Eissler_Iceman
Occasional Contributor III

...sorry, I was out of town. Attached is the data I'm using.

The workflow was to create a 500 foot buffer around the LPST points;  then see how many of those point buffers intersect Streams (using Select by Location/Intersect). I get 24.

However, if I just use Select by Location/Within a Distance of 500 feet from those LPST points, I only get 22.

Thanks, CCE

0 Kudos
Dale_Honeycutt
Occasional Contributor II

Craig:

I downloaded your data and tried to reproduce this issue.  I can't.  Here are some things I did to try to reproduce.

  • I started out by setting the Map's spatial ref to that of your points/streams. I then created the buffers and got 24 when I used Select By Location/Intersect
  • I then used Select By Location using the points and Within a distance 500 US Survey Feet. I got 24 again. Used International Feet (whatever that is) and got 24.
  • I changed the spatial reference of the map to several different projections, including 1984 Web Mercator, to see if that would somehow affect the selection.  It didn't.
  • For grins, I used the geoprocessing Select Layer By Location tool instead of Select By Location button found on the app's ribbon. No difference.

I'm at a loss to explain why you're getting 22.  The only way I could get a number ^= 24 was by selecting some of the streams and using that selected set in Select By Location.

So I'm at a loss.  Sorry.

0 Kudos
Dale_Honeycutt
Occasional Contributor II

I was embarrassed that I didn't know what International Feet was.  Here's a topic about it

International Feet: "0.3048 meters. Required for new coordinate systems used by U.S. government agencies beginning in 2023."

US Survey Feet: "1,200/3937 meters or 0.3048006096012192. Used in existing U.S. coordinate systems based on NAD27 and NAD83. Usage will be phased out starting in 2023."

0 Kudos
Craig_Eissler_Iceman
Occasional Contributor III

...No worries, this is a bit confusing. So, yeah, I'm curious to see how the "phasing out" is applied. ...As of now -- and it's one year later -- I'm still seeing both options in ArcGIS ...but maybe I'm missing something

0 Kudos
Dale_Honeycutt
Occasional Contributor II

I've been wondering how it affects my car's speedometer readings  🤔

0 Kudos
Craig_Eissler_Iceman
Occasional Contributor III

Hmmm, thanks for trying that -- it lead me to think of something... I actually have 2 Streams layers. One is in UTM coordinates which produces 24 intersections with the buffers; and the other is in State Plane coordinates which produces 22 intersections with the buffers ...so, therein lies the difference! ...And, yes, of course, it's best to have all operational layers in the same projected coordinate system for spatial analysis ...and here's a good example 😉 Thanks again for you help!