Hi,
This is a question I have been scratching my head lately.
In my data, there seems to be a global range, i.e. spatial autocorrelation
does imply some sort of homogeneity throughout the whole area. However, if you
look into more details, some of the sub-region contains outlier and hotspot
wh
Title: Message
About your question
on heterogeneity in spatial data. Sorry for my very lazy answer but one
elegant way to deal with such heterogeneity is Denis Allard's class kriging.
See http://www.ai-geostats.org/papers/_papers/000e.htm
Gregoire
PS: I just have
checked the ftp