SUMMARY: AI-GEOSTATS: 3D-kriging with soil data
- Dear List,
sorry that you had to wait for this summary.
I post the answers that I received in this Mail with the corresponding
names and email addresses. The answers were help for us to decide to do
kriging in 2-dimensional layers. However, so far do not have final results
but it seems to work fine this way.
Rudi Dutter <R.Dutter@...>:
I was involved in a geochemical analysis project at the peninsula Kola
had less data than you have, and also a much smaller amount of values in
z-direction. Nevertheless the results are documented in
C. Reiman et al. (1998): Environmental Geochemical Atlas of the Central
Region. Norges geologiske undersokelse (NGU), Geological Survey of Norway.
We experienced with kriging in 2-dimensional layers, and the results are
published. Only afterwards we started investigating the relations between
empirical structures between the layers. This seems to be very
however, no result is published.
I believe that 3d-kriging does not bring much, however, the comparisons
the structures of the layers are very challencing.
Denis Marcotte <Denis.Marcotte@...>:
It is probably better to treat this kind of data as 2D, soil type by soil
type. You can krige the thickness of each soil type and also the measured
variables; this will allow you to reconstruct vertical sections
everywhere. The data probably shows a complicated trend along the vertical
profile which complicates seriously the 3D study.
Med Bennett <mbennett@...>:
I would compute a vertical or "downhole" variogram to see if there if any
significant spatial correlation in that direction. If so, you should
attempt to use a 3D kriging method, as it would be wasteful not to use
information. You will have to use a very high value for anisotropy in the
Z direction (on the order of 1000 probably) or else leave the depth units
as cm and the horizontal units in m.
"Nel, Fanie" <fnel@...>:
I would suggest that you only krig within a horizon , not across horizons.
This assumes that the variables behave diffrently in each horizon. You
also have to estimate the width of each horizon.
If a variable shows a different variograms vertically and horizontally you
should do 3D kriging.
I hope there all who wrote to me listed above (there is some chaos in my
Mailbox these days).
Groner Landstr. 45
fon: +49 (0)551 / 5317388
fax: +49 (0)180 / 01212 5 109 12 817
mail: 1) mboehm@...-goettingen
On Thu, 7 Jun 2001, Markus Boehm wrote:
> Dear all,
> how should soil data (~3300 profiles, ~150m average horizontal distance,
> all about 1 meter deep with different thickness of horizons representing
> type of soil, humus contents, ...) be kriged?
> I need all information of the different horizons to calculate the somewhat
> like ground water infiltration rate.
> Does it make sense to use 3D-kriging methods? I think then there should a
> modification of the z-coordinate be made. I mean stretching of the
> z-coordinate. Does anyone know how to deal with the problem or some useful
> Or would you prefer to do 2D-kriging for a number of depths (10cm,
> 20cm,...), so there's no need to modify the z-coordinate?
> * To post a message to the list, send it to ai-geostats@...
> * As a general service to the users, please remember to post a summary of any useful responses to your questions.
> * To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
> * Support to the list is provided at http://www.ai-geostats.org
* To post a message to the list, send it to ai-geostats@...
* As a general service to the users, please remember to post a summary of any useful responses to your questions.
* To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
* Support to the list is provided at http://www.ai-geostats.org