Spatial data consist of two components, a spatial component and an attribute component (Haining, 2003). Spatial sampling can refer to the collection of information about either of these two components. An example of the sampling of the spatial component of a data set is the use of a global positioning system (GPS) to record locations in a landscape of point features such as trees. An example of sampling of the attribute component is the collection of soil properties such as clay content, cation exchange capacity, and so forth, according to some sampling pattern. One could also sample for both the spatial location and the attribute values. All four of the data sets used in this book conform to the second type of sampling, in which the location of the sample is speci–ed and some attribute value or values are recorded. It is implicitly assumed in this sort of sampling that the locations of the sites are measured without errors. This assumption is not always valid in real data sets, but it is a fundamental one on which much of the theory is based. Since the true location is usually close to the measured location, the usual way to take location uncertainty into account is to augment attribute uncertainty.