A dynamical space-time calibration model for the airborne particular matter data

The PM10 concentrations are often measured by different instruments situated in different sites. As some of these instruments are more precise than others we utilize the data from the former ones to correct the less precise data, obtaining an homogeneus dataset. To do this we propose a Geostatistical Dynamical Calibration model (GDC) based on the state - space approach of the spatio temporal modelling. This approach is a geostatistical extension of the DDC model (?). We assume that the observed data are random fields composed by a linear function of the “true” levels and error components, where the “true” concentrations are unobservable processes and represent the state equations of the model.Considering the TEOM and LVG data of the Piemonte region, we show some preliminary analysis.