0
$\begingroup$

I am working with a dataset of 1,284 settlements that are saved as point feature classes in ArcGIS. I have elevation values for every point based on a digital elevation model. I am attempting to see if there is some type of decision-making process that influences the elevation of the settlements or if it is the result of random chance. To do so, I created 99 datasets of 1,284 randomly placed points within my study area and got the elevation for those.

Is there some way I can compare the observed values against the simulated values to see if the differences are statistically significant?

Let me know if I should post any of the data...but here is a graph I made...

enter image description here

$\endgroup$
1
  • $\begingroup$ What exactly you want to compare? What is the question you want to answer? $\endgroup$
    – Tim
    Oct 4, 2020 at 8:03

1 Answer 1

1
$\begingroup$

If your data are a random sample from a population (as opposed to the population itself), then it seems a one-sample Wilcoxon test would show that the center of the population sampled differs from the center of your simulation (average value line).

In particular actual settlements seem to prefer 0-200 m, whereas your simulation shows many locations around 400 m.

Speculation: Maybe there is an advantage to being near sea level, and topography or soil around 400 m makes building difficult. Building at somewhat higher elevations may be considered worth the trouble because of spectacular views (or good soil for growing grapes or good landscape for sheep or dairy farms).

If data include essentially the entire population of settlements, then a test would not be appropriate, and your graph would suffice to tell the story.

$\endgroup$
2
  • $\begingroup$ I am a geographer who has worked with digital elevation models, but that background isn't needed to appreciate that settlements (should) avoid areas prone to flooding and can be much influenced by the need for relatively flat sites. Slope and also profile and plan curvature can be as important as altitude or elevation. $\endgroup$
    – Nick Cox
    Oct 4, 2020 at 9:56
  • $\begingroup$ Thanks, Nick, for the assistance. The settlements are archaeological sites, many known only due to identification during survey work and never excavated. The dataset contains nearly all known settlements, so I will just use the graph then to tell the story (but thanks for letting me know about the Wilcoxon test!). $\endgroup$
    – John
    Oct 6, 2020 at 15:12

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.