Model Evaluation
A jackknife procedure was used to evaluate the model.
The min and max temperatures at each of the 233 stations were predicted with the kriging model using all data
within the 2.2 degree cutoff, except for stations within 0.1 degrees.
This excludes the station itself and a few other very close stations.
The mean absolute error for this procedure 3.52°F.
Because of the well dispersed stations, this jackknife error is a near worse-case situation.
The best-case situation uses the station's own data when predicting that station.
For the best-case situation, the mean absolute error with the same kriging model is 0.023°F.
It is the small error variance that allows the model to reproduce the observed points so closely.
The typical mean prediction error of the kriging model is somewhere in between these values,
about 1.8°F.
The jackknife procedure was also used initially to evaluate possible model parameters,
but this procedure was too insensitive to the model parameters to get reliable best estimates.
This is common in kriging situations when the data dominate the results.
27 AUG 2001, updated on 30 OCT 2001
dlg@rapid.msu.montana.edu