Local and regional secondary organic aerosol: Insights from a year of semi-continuous carbon measurements at Pittsburgh

During the Pittsburgh Air Quality Study (PAQS) an automated semi-continuous thermal-optical transmittance (TOT) carbon analyzer was used to measure 2-4 h average particulate organic (OC) and elemental carbon (EC) concentrations from July 1, 2001 to August 13, 2002. To minimize the adsorption of vapor-phase organics, the sample air was drawn through a multi-channel parallel-plate diffusion denuder placed upstream of the carbon analyzer. Particulate OC and EC in the sample air were then collected on a quartz fiber filter (QFF) mounted inside the carbon analyzer, and analyzed immediately after collection. To account for any remaining organic vapors not retained by the denuder and collected on the sampling filter (positive artifact) a dynamic blank was run every two weeks. An upper-bound estimate of volatilization induced by the presence of the denuder upstream of the sampling filter (negative artifact) was also made. A detailed description of the operating protocol and quality assurance measurements is provided. The contributions of primary and secondary organic aerosol (SOA) to particulate OC were calculated using an "EC tracer method," which is codified herein. Annual average SOA accounted for 33% of particulate OC. SOA accounted for 30-40% of monthly average OC from June to November in Pittsburgh, similar to previous summertime estimates for Atlanta (Lim and Turpin 2002) and much larger than previous estimates of SOA in the Los Angeles Basin (Turpin and Huntzicke 1995). Examination of concentration dynamics suggests that multi-day formation and regional transport is an important contributor to the higher SOA contributions to OC in Pittsburgh and suggests that SOA is likely to be a particularly important contributor to particulate OC in locations that are recipients of long distance transport, such as the eastern United States.