Next you will create (interpolate) a surface of ozone concentration using the default settings of Geostatistical Analyst. You will use the ozone point dataset (ca_ozone_pts) as the input dataset and interpolate the ozone values at the locations where values are not known using ordinary kriging. You will click Next on many of the dialog boxes, thus accepting the defaults. Don't worry about the details of the dialog boxes in this exercise. Each dialog box will be revisited in later exercises. The intent of this exercise is to create a surface using the default options.

The Semivariogram/Covariance Modeling dialog box allows you to examine spatial relationships between measured points. You assume things that are close are more alike. The semivariogram allows you to explore this assumption. The process of fitting a semivariogram model while capturing the spatial relationships is known as variography.

The crosshairs show a location that has no measured value. To predict a value at the crosshairs, you can use the values at the measured locations. You know that the values of the closest measured locations are most like the value of the unmeasured location that you are trying to predict. The red points in the above image are going to be weighted (or influence the unknown value) more than the green points since they are closer to the location you are predicting. Using the surrounding points, with the model fitted on the Semivariogram/Covariance Modeling dialog box, you can predict a more accurate value for the unmeasured location.

The Cross Validation dialog box gives you an idea of how well the model predicts the values at the unknown locations. You will learn how to use the graph and understand the statistics in Exercise 4.

Notice that the interpolation continues into the ocean. You will learn in Exercise 6 how to restrict the prediction surface to stay within California.