The interesting story given by Tim
Glover is a good example of "spatial outliers". The dumptruck loads of
polluted soils are too different from the neighbourhood, and thus they should be
regarded as spatial outliers. When one cannot get a decent variogram after
trying all possible data transformations and robust calculations, another way
worth trying is to detect such spatial outliers (e.g., using local Moran's
About two years ago, I dealt with a dataset of soil
organic carbon in Ireland, and found that exclusion of spatial
outliers significantly improves the structure of a variogram. When doing
kriging, the excluded values may be put back to preserve information of raw
This may be an explanation to the conflict
between failed variogram and spatial structure. Spatial outliers may destroy
spatial structures and thus result in failed variograms.
----- Original Message -----
Sent: Thursday, September 02, 2004 3:30
Subject: [ai-geostats] spatial
> Dear oh Dear, I am failing to communicate (again).
> As far as I know, I didn't say you could not use
> geostatistics when a
trend is present! I regularly use
> Universal Kriging for data with a
trend and kriging
> with an external drift when the trend is governed
> an outside factor (see free tutorial at website).The question originally posed what how does one decide
geostatistics is not appriate. The answer
> Gregoire and myself gave was
"when you cannot get a
> semi-variogam graph" after trying all
> variations of transforms, interpretation andde-trending.
> I recently worked with an orange grove in
> (bugs on oranges) which showed no decenteven though rough inverse distance maps
> looked reasonable. It turned out
they had two
> different kinds of tree in the orchard. Separating
> 'rootstocks' yielded a vastly improved semi-variogramdecent geostatistical analysis.
> failure to obtain a
> My additional point was that
> semi-variogram model simply means that there is
> 'distance related' structure. It does NOT mean therespatial structure.
> is NO
> ___________________________________________________________ALL-NEW Yahoo!
Messenger - all new features - even more fun! http://uk.messenger.yahoo.com>
> * By using the ai-geostats mailing list you
agree to follow its rules
> ( see href="http://www.ai-geostats.org/help_ai-geostats.htm">http://www.ai-geostats.org/help_ai-geostats.htm )
> following in the subject or in the body (plain text format) of an email message
> * To unsubscribe to ai-geostats, send the