Traditionally spatial and temporal modelling assumes that the sampled locations (in time or space) are selected regardless of the values of the process under study. This phenomenon, called preferential sampling is present in many studies, whenever the process associated with the locations of the data and the process being modeled are stochastically dependent. One of the main objectives of this work is to extend the concept of preferential sampling to the temporal component. We aim at studying different models for the dependence structure in the spatial and temporal setting, different distributional assumptions for the observed stochastic process and to develop spatial and temporal predictors, under a model based approach, which give good estimates of the unobserved data. |