I am a newcomer to the subject of geostatistics and find the
going very tough.
I think I have a very simple question, but one which has been
bugging me for quite some time now.
The question has its origin from the field of traditional
Is there any way you can test for drift in a set of samples,
which have a distribution which are part of what appears to be a very lognormal
I get the impression from the literature that drift is
interpreted when the experimental variogram is not 'behaving' very well.
However, there appear to be numerous other factors which can explain this
Any comments, including tips on treating lognormally
distributed data, are welcome.