Title: RE: AI-GEOSTATS: Non-Monotone Variogram model
Dear Yadollah,
have a look to see if you have any strong outliers using a variogram cloud. Your potential drift may simply be a single erroneous or extreme value, which once removed will cause your variogram to stabilise,
Benjamin Warr
Hello list members
My Empirical Variogram for a spatial data set has an cubic like form(see below). Then
the ordinary variogram models such as Exponential, Gaussian, Linear, Spherical models
is not best fit to my data( Because these models are monotone (increasing) functions).
Then I need to
Dear Waghei,
At first I would not trust only in the traditional semivariogram. You
should better use more robust measures to evaluate the semivariogram
structures. There are several possibilities like correlogram or other
standardized variogram estimators (Srivastava, Parker 1989 - see below
A graph of semivariogram
gamma(h)
|.
| .
| . .
|..
| .. .
|. . .
| . ..
| .
|
