Air pollution and health session at ISBA July 2014
Published: 31 March 2014
Results from the project to date were presented at an invited session of the annual 2014 International Society for Bayesian Analysis (ISBA) conference in Cancun on 14th - 18th July.
The ISBA session “Spatio-temporal methods for predicting air pollution” was chaired by Sudipto Banerjee and attended by around 35 people. The three speakers in this session were Dr Veronica Berrocal (University of Michigan), Dr Sabyasachi Mukhopadhyay (University of Southampton) and Dr Alastair Rushworth (University of Glasgow). The first speaker Dr Berrocal, described about a new approach for statistical downscaling, applied to ground-level ozone concentrations in the United states. Following this Dr Mukhopadhyay described innovative Bayesian space-time models for estimating and predicting concentrations of air pollutants across the UK. A number of new models were presented, and each were applied to a number of different types of pollutant, and resulting spatial maps of the fitted concentrations were shown. The final speaker, Dr Rushworth described a novel approach to modelling localised space-time autocorrelation for use in studying the relationship between air pollution and human health. A new conditional autoregressive model was presented and the method was illustrated on some simulated data, and on a large space-time data set of hospital respiratory admissions in English local health authorities. Copies of the talks presented are available on the "Deliverables and Outputs" page of this website.
First published: 31 March 2014