I want to detect change in two kernel density estimation rasters. These were created using the same bandwidth in ArcGIS 10.2. I want to see if the differences between these (which was done in Raster calculator as Float(map2 - map1) are significant.
Bowman and Azzalini (1997) describe a method for estimating whether points in both rasters differ with a confidence of 95%, however, I don't understand the formula well enough to manually put the numbers in myself. Several other authors have also used this method but not with data so I can't follow (i.e. Nelson, 2011, 'Quantifying wildlife home range changes'). The formula is:
change = (sqrt(pixel in map 1) - sqrt(pixel in map 2)) / sqrt((standard error of map 1)^2 + (standard error of map 2)^2)
The formula for determining standard error of the maps is very complicated and I don't understand it.
Does anyone know if there is a package/add on or a simple way of doing this in ArcGIS or in R?
Thanks for any help - happy to expand if needed.
أكثر...
Bowman and Azzalini (1997) describe a method for estimating whether points in both rasters differ with a confidence of 95%, however, I don't understand the formula well enough to manually put the numbers in myself. Several other authors have also used this method but not with data so I can't follow (i.e. Nelson, 2011, 'Quantifying wildlife home range changes'). The formula is:
change = (sqrt(pixel in map 1) - sqrt(pixel in map 2)) / sqrt((standard error of map 1)^2 + (standard error of map 2)^2)
The formula for determining standard error of the maps is very complicated and I don't understand it.
Does anyone know if there is a package/add on or a simple way of doing this in ArcGIS or in R?
Thanks for any help - happy to expand if needed.
أكثر...