Research

Between-airport heterogeneity in air toxics emissions associated with individual cancer risk thresholds and population risks

Ying Zhou¹ and Jonathan I Levy²

¹  Department of Environmental and Occupational Health, Rollins School of Public Health, Emory University, 1518 Clifton Road NE, Atlanta, Georgia 30322, USA

²  Department of Environmental Health, Harvard School of Public Health, Landmark Center 4th Floor West, PO Box 15677, Boston, Massachusetts, 02215, USA

author email corresponding author email

Environmental Health 2009, 8:22doi:10.1186/1476-069X-8-22

Published:	8 May 2009

Abstract

Background

Airports represent a complex source type of increasing importance contributing to air toxics risks. Comprehensive atmospheric dispersion models are beyond the scope of many applications, so it would be valuable to rapidly but accurately characterize the risk-relevant exposure implications of emissions at an airport.

Methods

In this study, we apply a high resolution atmospheric dispersion model (AERMOD) to 32 airports across the United States, focusing on benzene, 1,3-butadiene, and benzo [a]pyrene. We estimate the emission rates required at these airports to exceed a 10^-6 lifetime cancer risk for the maximally exposed individual (emission thresholds) and estimate the total population risk at these emission rates.

Results

The emission thresholds vary by two orders of magnitude across airports, with variability predicted by proximity of populations to the airport and mixing height (R² = 0.74–0.75 across pollutants). At these emission thresholds, the population risk within 50 km of the airport varies by two orders of magnitude across airports, driven by substantial heterogeneity in total population exposure per unit emissions that is related to population density and uncorrelated with emission thresholds.

Conclusion

Our findings indicate that site characteristics can be used to accurately predict maximum individual risk and total population risk at a given level of emissions, but that optimizing on one endpoint will be non-optimal for the other.