Spatial heterogeneity of Covid-19 cases in Italy

We study weekly Covid-19 incidence rates in Italy from the 24th February 2020 to late July 2021 using space-time disease mapping models. These models often include an interaction term to account for the complexity in the data. In a Bayesian hierarchical framework, prior elicition of precision parameters, re- sponsible for the smoothness of the process, remains a difficult task. We use an intuitive reparametrization of space-time interaction models by means of mixing parameters that control the proportion of variability explained by each term.