Spatio-temporal ARMA Modeling in a Geostatistic Framework

This paper investigates models of stochastic spatio-temporal dependence with geostatistical tools. Specifically, we use a time and space dependence ARMA model to propose new models of stochastic spatial and temporal tools such as variograms, correlograms and the madograms. we use a geostatistical tools as a spatio-temporal kriging to modling some health data. Again, we establish a link between the STARMA model and ordinary kriging, then apply it to covid-19 data.