Interannual variation of air quality across an international airshed in Detroit (USA) and Windsor (Canada): A comparison of two sampling campaigns in both cities

Abstract

This study investigates air pollutant concentrations across the international airshed spanning Detroit, Michigan, USA and Windsor, Ontario, Canada. The Geospatial Determinants of Health Outcomes Consortium (GeoDHOC) measured air quality concurrently in Detroit and Windsor using consistent sampling methods and locations during separate, two-week periods in September 2008 and May/June 2009. This paper presents 2009 results and compares them to previously reported 2008 campaign results to assess spatial and temporal variability. A high-density array of 100 passive and 50 active samplers was used to measure nitrogen dioxide (NO2), 26 volatile organic compounds (VOCs), 23 polycyclic aromatic hydrocarbons (PAHs), and three size fractions of particulate matter (PM) in both campaigns. Geospatial and non-spatial tools were used to investigate changes in concentration distributions and site classification into low/medium/high tertiles between the campaigns. During the two-week 2009 campaign, general pollution patterns remained consistent with those observed during 2008. Higher concentrations in Detroit, strong correlations amongst the BTEX group, and consistency of sites falling into the low and high tertiles were observed. Similar spatial patterns for NO2, VOCs, BTEX, and PAHs were also found during both campaigns. Conversely, PM1-2.5 and PM2.5-10 distributions showed greater variability, as did toluene/benzene ratios. Observed differences are attributable to changes in point source emissions associated with changes in localized activities, possibly related to decreased economic and industrial activity in response to the downturn that began in late 2008. Repeated sampling at multiple locations is important to determine spatial and temporal variability in absolute concentrations. However, if the definition of relatively high and low concentrations regions within a well-established urban area is adequate to estimate future air pollutant exposures, a single multi-location campaign may suffice.