Universal kriging of multivariate spatial data under multivariate isotropic power type variogram model

The purpose of the present paper is to establish universal kriging formula for multivariate spatial data observed over a second order stationary random field. In this study we consider multivariate isotropic parametric variogram model of power type. The parameters of the model including the cross-covariance parameter are estimated by applying least squares method. Numerical solution of the least squares equations are approximated by applying graphical approach. The main assumption adopted in this work is that the spatial process is regarded as isotropic in that the covariance function depends on the Euclidean distance among the two considered points. We demonstrate the application of the method to a bivariate spatial process of corn plants for construction the simultaneous kriging maps of the process which are generated by scatting the contour plot of the optimum prediction of the corn plants over the experimental region under bivariate parametric power type variogram model.