I have two boundary sets. Intersecting reveals that there are genuine differences between them - as expected. However, there are also a host of slivers due to their being digitised by different agencies. These mismatches may have statistical implications.

I have tried several approaches to clearly differentiate one mismatch type from the other.

Current thinking:

  1. I get the MBR of each polygon in boundary set 1 and get the ratio of the area of the MBR to the area of the actual polygon. I do the same for boundary set 2.

  2. I get the MBR of each intersection (these have been disaggregated to singlepart) and get the ratio of the area of each intersection's MBR to the area of the actual intersection.

  3. The distribution of the ratio for slivers is likely to be materially different to the distribution of the ratio for genuine differences.


Do you agree that this seems like a sensible approach?

How easy is it to obtain the MBR (minXmaxXminYmaxY) for a polygon in ArcGIS 10.1 Advanced? Either point-and-click or Python suggestions welcome.

If my thinking is flawed, please point out how, and suggest an alternative.


In my daily work I have also these kind of challenges. Here is my workflow which working fine till now.

First step use Integrate by setting cluster tolerance and ranks. In this way you are making both layer semless based on the cluster tolerance and ranks. Make sure you always have the backup of both layers before using Integrate because this tool modifies the input data.

Second step now use the Intersect or Union to find genuine differences.

| improve this answer | |
  • Many thanks - apologies for the very delayed response. – Bruce Mitchell Jul 15 '14 at 10:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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