Spatio-Temporal Modeling of Intra-Urban Variations in Air Pollution Concentrations

• Activity Number: 59
• Type: Contributed
• Date/Time: Sunday, August 3, 2008 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Adam A. Szpiro*+ and Paul D. Sampson and Lianne Sheppard and Darren Wilton and Tim Larson and Thomas Lumley and Sara D. Adar
• Companies: University of Washington and University of Washington and University of Washington and University of Washington and University of Washington and University of Washington and University of Washington
• Keywords: Bayesian modeling ; Spatio-temporal modeling ; Air pollution

Abstract

To estimate the effect of air pollution on cardiovascular health in a cohort study, it is necessary to predict individual exposure based on sparse measurements. The U.S. EPA has few monitors in any given city as the network is primarily designed for regional air pollution levels. To address variation near roads, the EPA-funded MESA Air project is carrying out additional monitoring. Two features of the data present challenges. First, to represent small-scale variation in traffic-related pollution, we must pay close attention to covariates and how they relate to local meteorology. Second, to take advantage of irregularly sampled data we need a rich statistical model with space-time interactions in the correlation. Estimation of the parameters depends on specialized computational techniques. We describe an approach to addressing the challenges outlined above and give illustrative results.