ABSTRACT
Earth Sciences data are often collected in three-dimensional space as well as with reference to time at a given location. These data may be continuous along the parameter space but recorded/observed data are often discretized for convenience in computerized processing. A time series model is a suitable probability model in such a manner that each observation xt is assumed to have realized the value of a certain random variable Xt and the concept allows for the unpredictable nature of future observations. The class of linear time series models, which include ARMA models, form a basis for studying stationary processes. A stationary time series has properties that are invariant over the time. The chapter considers the time series and spatial series to be mathematically equivalent and as such, the parameter space includes time as well as spatial dimension(s).
