Basically, I want to generate artificial landscapes that vary in "tree cover" abundance and distribution. This should be relatively easy to do by varying Hurst exponents (level of aggregation) and adjusting the number of pixels representing tree cover in a matrix (in Matlab for example). The issue is that I also have to include a layer representing enhanced vegetation index. Since EVI is correlated with tree cover in real landscapes, my artificial landscapes should also represent this level of correlation. I think I can use the Band Collection Statistics tool to determine the correlation between tree cover and EVI of a "real" landscape", but is it then possible to "apply" this correlation value to generate a spatially correlated landscape with a layer representing tree cover and another representing EVI?
Any help is greatly appreciated!