ABSTRACT
Meteorological quantities (temperature, precipitation, etc.) exhibit variations both in time and space. On the other hand, a substantial degree of similarity usually is present between observations of a meteorological quantity taken either: (a) at the same location separated by short periods of time; or (b) at the same time at different locations separated by small distances. To represent such meteorological quantities in a realistic manner, it is necessary to consider probabilistic models that allow for both variability and dependence. Such models are useful for characterizing meteorological processes in terms of a few meaningful parameters, and they are also necessary to make valid statistical inferences about meteorological data.
