Using Geostatistical Analyst, you can compare the results of two mapped surfaces. This allows you to make an informed decision as to which provides more accurate predictions of ozone concentration based on cross-validation statistics.

To do so, right-click the Trend Removed layer and point to Compare. You will be comparing the Trend Removed layer with the Default layer you created in the second segment.

Because the root-mean-square prediction error is smaller for the Trend Removed layer, the root-mean-square standardized prediction error is closer to one for the Trend Removed layer, and the mean prediction error is also closer to zero for the Trend Removed layer, you can state with some evidence that the Trend Removed model is better and more valid. Thus, you can remove the Default layer since you no longer need it.

Close the Cross Validation Comparison dialog box. Right-click the Default layer and click Remove.

Click the Trend Removed layer and move it to the bottom of the table of contents so that you can see the sample points and outline of California.

Click Save on the Standard toolbar.