ABSTRACT
In this chapter, a general introduction for all cokriging methods is provided firstly, then the multivariate geostatistical methods that are available in R are introduced. These methods are 1) simple cokriging; 2) ordinary cokriging; and 3) kriging with an external drift (KED). Various examples are provided for how to apply these methods, including examination of assumptions and data requirements, parameter optimization, maximization of the predictive accuracy, and prediction and variances. The functions gstat.cv and fit.lmc in the gstat package and the function of krigecv in the spm2 package are introduced and used for variable selection, parameter optimization and predictive accuracy assessment. And the relevant functions are introduced for generating the spatial predictions of these methods. The petrel point data in the spm package and the pb. df grid data in Chapter 1 are used to demonstrate the applications of the multivariate geostatistical methods in R .
