Spatial Interpolation and Prediction of Gaussian and Max-Stable Processes

This thesis deals with different aspects of spatial interpolation and prediction of random fields. In the case of Gaussian random fields, best linear predictors and conditional distributions are well-known, provided that the mean and covariance structure of the random field are given. For parametric estimation of the covariance function from data, we consider the flexible class of Whittle-Mat