Spatio-temporal Clustering of PM2.5 in northern Italy using a Bayesian model

Florian Wolf, Alessandra Guglielmi, Alessandro Carminati

Abstract: Northern Italy is a well-known hotspot of air pollution in Europe, due to its high population density, characteristic geography and specific climate conditions. Consequently, there is a considerable interest in investigating the temporal patterns of air quality across numerous stations scattered throughout the region. In this work, we apply a Bayesian spatio-temporal product partition model for clustering particulate matter (PM2.5) concentration in northern Italy and compare its performance with a baseline spatial-only model.