louise_

Significance of Kriging Standard Error

Discussion created by louise_ on Jan 18, 2011
Hello!

I am a student working on a kriging interpolation of groundwater data.
I want to compare two models: One with the default settings of the GA Wizard and one i modified myself.
The created groundwater surfaces look almost identically. However the average standard errors are signifcantly different. As a result the root-mean-square standardized errors are 0,9 for "my own" model and 0,15 for the default options. From what I understand the default model has to be regarded as "not valid" for this reason.

So my questions are: Is the standard model really not valid? And what is the practical use of this standard error (and a corresponding map) if my predicition surfaces look almost the same?

I would appreciate your help on this very much!

Kind Regards,
Luise Keller

Outcomes