Predicting Traffic-Related Air Pollutants in Near-Road Urban Environments

• Activity Number: 91
• Type: Invited
• Date/Time: Sunday, August 9, 2015 : 9:30 PM to 10:15 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Owais Gilani* and Veronica J. Berrocal and Stuart Batterman
• Companies: University of Michigan School of Public Health and University of Michigan and University of Michigan School of Public Health
• Keywords: Air pollutant ; urban near-road ; Bayesian ; Dispersion ; Highway Traffic ; Spatiotemporal

Abstract

An improved understanding of concentrations of traffic-related air pollutants near roads enables us to predict exposure and assess their adverse impacts on health. In this study, we focus on concentrations of black carbon (BC) and nitrogen oxides (NO and NO2), three of a suite of traffic-related pollutants measured in nine transects across major roads in Detroit, MI over a period of one week in winter 2012. Repeated measurements were taken during rush-hour periods at sites in residential neighborhoods located from 50 to 500 m from both sides of the road. Bayesian spatial models were fitted to predict concentration of the various pollutants attributable to traffic sources. The models accounted for highway traffic volume and mix, wind speed, wind direction, time of day, and day of the week. Recognizing a different spatial dependence in the pollutant concentrations upwind and downwind of the highway, we employed non-stationary spatial covariance models in order to obtain improved exposure estimates of traffic-related air pollutants in near-road urban environments.