Hi,
I also think that transition probabilities might show a better
connection to the nature of the original variable, and mcuh less
cumbersome...
Regarding the sill problem, is it so surprising, that a de-trended
variable shows less variance than the original one? If the gaussian
variable has a
Dear List,
I am a PhD student working on geostatistical facies (i.e. categorical) models in the University of Barcelona (Spain).
I am using Truncated Gaussian Simulation to build facies models and I have some questions. My dataset is around 100 vertical wells penetrating a deltaic succession.
Hi Patricia
I think that an indicator (or a transition probability) approach
could be easier to interpret
(i.e. each indicator variogram
tell you something about the spatial pattern of each single facies).
Then, in regard to the sill of your variogram it depends if your
variogram is calculated