I am using ArcMap

Out of a raster, I reclassified it into only having the urban areas.

  • I ran this through zonal statistics and got the area of the urban area (in meters, which I presume are in square meters because it is area, correct?).
  • After getting the zonal stats, I joined into a census tract file which has immigrant population data (ex. dominicans, europeans,etc.).
  • I want to calculate the population density by square kilometer of these urban areas (1000 person/km2). What I did was taking the area field (results of the zonal stats) and dividing the presumed square meters into 1,000,000 to get the square kilometers.
  • After which, I divided the population group by the square km of that tract.

My question is, did I did this correctly, did I missed a step or are the results of the zonal stats not square meters?


1 Answer 1


Reading through your steps it sounds correct to me, although I would question

  1. Was your original dataset in meters so that your output from your first step, which you say is in meters, is actually in the same units. If the units are meters, then I think it's safe to assume the area will be in square meters.

  2. Do your census tracts match your areas so that the values from your census tracts directly relate to the same areas from your Urban Areas.

I think that if you can answer 'yes' to both of these then it appears that your steps are correct.

  • Thank you, this helps! I doubled checked everything and it seems that it was the correct method. This eases my worries!
    – XSC
    Apr 23, 2016 at 1:21
  • As a matter of habit, whenever I do any distance, area, or volume calculations with rasters, I always zoom to the level of single cell and then use the measure tool to double check the units & resolution of the raster.
    – Ezra Boyd
    May 3, 2016 at 2:45

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.