ABSTRACT
This chapter examines the spatio-temporal model that fitted to describe the variability in data. It provides helpful suggestions on how to evaluate models for spatio-temporal data. The evaluation of a model through model checking, validation, and selection is a very important step in the model-building process. The chapter focuses less on how to implement the methods in R and more on the methods themselves. The reasons for this are twofold. First, several diagnostics are straight-forward to calculate once predictive distributions are available. Second, there are only a few packages that have a comprehensive suite of diagnostic tools. The idea of field comparison is to compare two or more spatial or spatio-temporal fields, in some sense, to decide if they are different. Field-matching approaches have attempted to deal with questions of scale decompositions and feature properties, and a summary of such methods from a geophysical perspective can be found in Brown et al. and Gilleland et al.
