Bayesian Gaussian models for point referenced spatial and spatio-temporal data

When data is correlated both spatially and temporally, spatial and spatio-temporal modelling is useful for meaningful interpretation of the parameters of the covariates and for reliable predictions. In this paper we discuss some modelling strategies for point referenced spatial and spatio-temporal data. We describe Gaussian models in this context and use Bayesian hierarchical approaches for model based inference and predictions through the Markov chain Monte Carlo (MCMC) algorithm. Yearly average precipitation data from Western Australia is used to illustrate the models.