Hierarchical Bayesian Spatio-Temporal Modelling of Regional Ozone Concentrations and Respective Network Design
Ground-level ozone concentration is a key indicator of air quality, which varies in space and time. A statistical modelling method of ground-level in a given region is undertaken in this talk. In addition, an environmental network design problem, is also explored. As an example, the ozone concentrations in the Pittsburgh region of Pennsylvania, United States is considered for demonstration purpose. In this region, there are 25 stations and all but one stations have missing observations. Covering this region by grid boxes, we obtain predictive distributions at the grid points by applying hierarchical Bayesian spatio-temporal modelling. In terms of an entropy criterion, the environmental network design problem is solved with use of the obtained predictive distributions. The model evaluation is also provided.