Handling measurement error when modeling semivariogram and covariance functions
Etapes :
-
Start the Geostatistical Analyst Wizard
, choose a kriging method, then begin to construct your model.
For an example on how to start constructing a kriging model, refer to Using ordinary kriging to create a prediction map.
-
When you reach the Semivariogram/Covariance Modeling page, enter a value for the Measurement Error parameter.
The value can be entered as a percentage of the Nugget or as a numeric value (which has to be less than or equal to the Nugget value).

If you have multiple measurements per location (coincident data), let Geostatistical Analyst compute the measurement error for you by selecting Estimate value from the drop-down list.

If the measurement error is set to zero, you'll get "exact" kriging.
4/26/2014