Is there a way to find the centroids of highest density areas in the density raster (red areas in the figure) created with the Optimized Hot Spot Analysis in ArcMap10.3. Theoretically I think I should do something like:

  1. Create poligons of different areas (green, yellow, orange, red in the figure)
  2. Select red areas
  3. X and Y of the red areas

I literally have no idea how to do it the stage 1 in ArcMap while I I know how to work out part 2 and 3. Or is there an easiest way to find the density of population (for example) and find out the coordinate of the highest density centers without using the Advanced Hot Spots tool?

enter image description here

  • 1
    Do you need geometric centroids or density-weighted centroids? (Both are easily obtainable in basically the same way: select the regions by comparing the values to a threshold; regiongroup the selection; then perform zonal means of the coordinates (or weighted zonal means of the weighted coordinates) using the regions as zones.) Another question is, what form should the results be in? Points, a binary grid, ordered pairs of coordinates, or something else?
    – whuber
    Mar 10, 2017 at 22:24

1 Answer 1


First you need to convert your Raster to Polygon. Then You can use Eliminate and\or minimum bounding geometry to clean up your polygons. Select your red polygons in any way you wish and run Feature to point tool to get the centroids. If such process is a daily routine you can create your own toolbox(model) in model builder.

  • 2
    First step, converting raster to integer is missing. Or perhaps reclassify
    – FelixIP
    Mar 10, 2017 at 22:50
  • Thank you for the tips, what I did was: 1- Reclassify ( so I could reshape the polygons as I prefer) 2- Raster to Integer 3- Raster to Polygon 4-Select by Attribute (selection of my red areas) 5-Features to point It worked out. Mar 13, 2017 at 10:55

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.