This is the fifth of six segments that show you how to improve the ozone prediction surface.

Cross-validation gives you an idea of how well the model predicts the unknown values.

For all points, cross-validation sequentially omits a point, predicts its value using the rest of the data, and compares the measured and predicted values. The calculated statistics serve as diagnostics that indicate whether the model is reasonable for map production.

In addition to visualizing the scatter of points around this 1:1 line, a number of statistical measures can be used to assess the model's performance. The objective of crossvalidation is to help you make an informed decision about which model provides the most accurate predictions. For a model that provides accurate predictions, the mean error should be close to 0, the root-mean-square error and average standard error should be as small as possible (this is useful when comparing models), and the root-mean-square stardardized error should be close to 1.

Here the term prediction error is used for the difference between the prediction and the actual measured value. For a model that provides accurate predictions, the mean prediction error should be close to 0 if the predictions are unbiased, the root-mean-square standardized prediction error should be close to 1 if the standard errors are accurate, and the root-mean-square prediction error should be small if the predictions are close to the measured values.

The Cross Validation dialog box also allows you to display scatterplots that show the error, standardized error, and QQ plot for each data point.

Click the QQPlot tab to display the QQ plot. From the QQPlot tab, you can see that some values fall slightly above the line and some slightly below the line, but most points fall very close to the straight dashed line, indicating that prediction errors are close to being normally distributed.

To select the location for a particular point, click the row that relates to the point of interest in the table. The selected point is shown in green on the scattergram.

You can save the cross validation to save a point feature class for further analysis of the results. Click Finish.

The Method Summary dialog box provides a summary of the model that will be used to create a surface.

Click OK. The predicted ozone map will appear as the top layer in ArcMap. By default, the layer assumes the name of the kriging method used to produce the surface (for instance, Ordinary Kriging).