Spatially interpolating point data

Spatial interpolation is the procedure by which the behavior of a certain phenomenon of interest is predicted in locations where it has not been measured. For this purpose, we need a spatial prediction model—a set of procedures to obtain the predicted values given the calibration data. The two types of calibration data usually encountered are:

• Field measurements: Available for a limited set of locations (usually points), for example, meteorological data from stations in Spain

• Covariates: Available for each location within the area of interest, for example, elevation data from Spain's DEM

The spatial prediction model of our choice is calibrated using the calibration data. This model can then be used to calculate the predicted level of the phenomenon of interest in any location (usually points). The two main types of spatial interpolation methods recognized are:

• Deterministic model: In this model, model parameter values are arbitrarily determined

• Statistical...