I am trying to do spatial autocorrelation of sorts based on a layer of polygons of mean rainfall data. I want to see how rainfall deviates from a certain point on the landscape. Essentially, I want to see how far someone would need to go in order to find another polygon of rainfall that is most different to the polygon they live in. I have found the interpretation I need from the semivariogram function in Geostatistical Analyst, but I want to be able to click on a polygon and see where the closest, yet most different (in terms of rainfall amounts) polygons are.
I want to try doing one of the methods of spatial autocorrelation, but it appears all of the methods I have tried calculates z-scores based on the mean of the ENTIRE area of polygons, not the single polygon that I want to calculate the correlation from.
Do you have any ideas on how to best do this? Is spatial autocorrelation not the best route and are there other spatial statistics that I may want to pursue? Or is there a way to fiddle with the settings on spatial autocorrelation to ensure it is calculate the Zscore from a single point?