Using probability kriging to create a probability map

Available with Geostatistical Analyst license.

Use probability kriging to produce a probability or standard error of indicators map. The data points need to be sampled from a phenomenon that is continuous in space, for example, temperature or air quality measurements.

Important parameters are the specified threshold (which determines which predictions will receive a 0 and which will receive a 1), covariance/semivariogram models, and search neighborhoods establishing the model. Estimation of semivariogram/covariance becomes difficult when the indicator variables are primarily all zeros or all ones. If possible, choose a threshold that gives a mix of zeros and ones.

1. Click the Geostatistical Analyst Wizard button on the Geostatistical Analyst toolbar.
2. Select Kriging/CoKriging, choose a dataset and attribute field, then click Next.
3. Choose Probability kriging.
4. Set Output type to Probability.
5. Type the Threshold value or click the Set button and set the threshold on the Classification dialog box.
6. Click either the Exceed or NOT Exceed option and click Next.
7. Specify the desired parameters on the Semivariogram/Covariance Modeling dialog box and click Next.
8. Specify the desired parameters on the Searching Neighborhood dialog box and click Next.
9. Examine the results on the Cross Validation dialog box and click Finish.
10. Click OK on the Method Report dialog box.

Feedback on this topic?