The Interpolate Points tool performs geostatistical interpolation to predict the values of a continuous variable everywhere between measured points.
Interpolation model creation
The Empirical Bayesian Kriging (EBK) tool is used to interpolate the Input points and Field to interpolate parameter values. Various parameters of EBK are determined by the Calculation precision parameter (see the Calculation precision section below for more information). The output of EBK is a geostatistical layer that stores the interpolation results and can be exported to various formats and representations.
Prediction layer creation
The GA Layer To Contour tool exports the geostatistical layer of predicted values to filled contour polygons. The values of the contours are determined by the Classification type, Number of classes, and Class break values parameters. If polygons are provided in the Clipping polygons parameter, the filled contour polygons are clipped using the Clip tool. The filled contour polygon layer is returned as the ResultLayer output layer of the tool.
Standard error layer creation
If the Output prediction errors parameter is checked, the GA Layer To Contour tool exports a second filled contour polygon layer for the standard errors of the predicted values. The standard error polygons are clipped to the clipping polygons, if provided. This polygon layer is returned as the PredictionError output layer of the tool.
Point prediction locations creation
If a point layer is provided in the Point prediction locations parameter, the GA Layer To Points tool uses the geostatistical layer to predict the values at the point locations. The predictions are stored in a point layer and returned as the PredictedPointLayer output layer of the tool.
The Calculation precision parameter sets various parameters of EBK to prioritize either the precision and accuracy of the interpolation results or the calculation speed. The parameter has three options: Speed, Balance, and Accuracy. The following table lists the parameter values used in EBK for each option:
Data transformation type
Semivariogram model type
Maximum number of points in each local model
Local model area overlap factor
Number of simulated semivariograms
Search neighborhood (Min neighbors)
Search neighborhood (Max neighbors)