The relationship between aviation activities and ultrafine particulate matter concentrations near a mid-sized airport Hsiao-Hsien Hsu a, * , Gary Adamkiewicz a , E. Andres Houseman a, b , Jose Vallarino a , Steven J. Melly a , Roger L. Wayson c , John D. Spengler a , Jonathan I. Levy a, d a Harvard School of Public Health, Department of Environmental Health, Landmark Center 4th Floor West, 401 Park Drive, Boston, MA 02215, USA b The Warren Alpert Medical School of Brown University,121 South Main St., Providence, RI 02912, USA c Volpe National Transportation Systems Center, 55 Broadway St., Cambridge, MA 02142, USA d Boston University School of Public Health, Department of Environmental Health, Talbot T4W, 715 Albany St., Boston, MA 02118, USA article info Article history: Received 18 August 2011 Received in revised form 30 November 2011 Accepted 1 December 2011 Keywords: Air quality Aircraft Ground measurements Regression Source attribution Ultrafine particulate matter abstract Aircraft contribute to emissions of ultrafine particulate matter (UFP) and other air pollutants, with corre- sponding impacts on community-level exposures near active airports. However, it is challenging to isolate the contribution of aircraft from local road traffic and other nearby combustion sources. In this study, we used high-resolution monitoring and flight activity data to quantify contributions from landing and take-off operations (LTO) to UFP concentrations. UFP concentrations were monitored with 1-min resolution at four monitoring sites surrounding T.F. Green Airport in Warwick, RI, in three one-week campaigns across different seasons in 2007 and 2008. Along with pollutant monitoring, wind data were collected and runway-specific LTO data were obtained from airport officials. We developed regression models in which wind speed and direction were included as a nonparametric smooth spatial term using thin-plate splines applied to wind velocity vectors and fitted using linear mixed models. To better pinpoint the timing in the LTO cycle most contributing to elevated concentrations, we used regression models with lag terms for flight activity (ranging from 5 min before to 5 min after the departure or arrival). Results suggest positive associations between UFP concentrations and LTO activities, especially for departures when an aircraft moves near or passes a moni- toring site. Departures of jet engine aircrafts on a runway proximate to one of the monitors have a maximal impact 1 min prior to take-off, with median absolute contributions during those minutes of 7400 particles cm 3 (range: 1100e70,000 particles cm 3 ). Across all observations, our models indicate median (95th, 99th percentile) percent contribution for all LTO activities of 9.8% (54%, 72%) and 6.6% (39%, 55%) for the two sites proximate to the airport’s principal runway, and 4.7% (24%, 36%) and 1.8% (22%, 31%) for the remaining two sites. Our analysis illustrates the complexity of aviation impacts on local air quality and allows for quanti- fication of the marginal contribution of LTO activity relative to other nearby sources. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction Aviation-related air pollution, including both emissions from aircraft and other airport-related activities, can potentially contribute to elevated exposures in communities near airports for multiple air pollutants. Detailed plume characterization studies in experimental settings have shown that aircraft engines have power-dependent emissions of multiple air pollutants, with the highest emission rates typically under thrust and increased fuel flow rate conditions commonly associated with take-off (Agrawal et al., 2008; Kinsey et al., 2010; Wey et al., 2006, 2007). Field studies have also demonstrated potential influences of aviation activity on a variety of VOCs, PAHs, and criteria pollutants (Carslaw et al., 2006; Herndon et al., 2008, 2005; Yu et al., 2004). However, it can be difficult to quantitatively separate the contri- butions of aviation from other local sources and background concentrations. This is in part because distinguishing aircraft emis- sions from other local combustion sources such as roadway traffic during monitoring studies is difficult and requires correlation over time and space with aviation activity. Studies that have attempted to evaluate the marginal contribution of airports to ambient pollutant levels have generally either not quantitatively estimated marginal contributions (Hu et al., 2009; Westerdahl et al., 2008; Yu et al., 2004) or have done so with 1-h or longer averaging times (Adamkiewicz et al., 2010; Dodson et al., 2009), which are more challenging to interpret given the correlations among sources and complex and rapidly changing source characteristics. Dispersion models are often * Corresponding author. Tel.: þ1 617 384 8806; fax: þ1 617 384 8859. E-mail address: leonhsu@hsph.harvard.edu (H.-H. Hsu). Contents lists available at SciVerse ScienceDirect Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv 1352-2310/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2011.12.002 Atmospheric Environment 50 (2012) 328e337