Evaluation of Sources of Organic Aerosol Components Measured by an Aerosol Mass Spectrometer (AMS) during NACHTT Field Campaign using Positive Matrix Factorization (PMF)

An Aerodyne compact Time-of-Flight Aerosol Mass Spectrometer (C-ToF-AMS) was deployed to determine size resolved chemical composition of aerosols in Nitrogen, Aerosol Composition and Halogens on a Tall Tower (NACHTT) field campaign at Erie (40.0500°N, 105.0039°W), CO between February 17 and March 14, 2011. Vertical profiles of organic aerosols along with other aerosol components (SO42-, NO3-, NH4+ and Cl-) were gathered along a 300 m tall tower during the study. The average total mass of the aerosols analyzed by the AMS varied from 2-7 μg m-3 and average contribution of organics to total mass changed from 17 to 32 %. Among the inorganic parameters measured, NO3- depicted the highest contribution to total mass with a value of 24 %. Positive Matrix Factorization (PMF), an advanced factor analysis tool used for source apportionment of atmospheric pollutants, was applied to the generated data set to find the sources affecting the chemical composition of the organic aerosols measured during the field campaign. Time periods with relatively higher organic concentrations measured were chosen as an input to PMF. PMF resolved three factors: Hydrocarbon-like Organic Aerosol (HOA) and two oxygenated organic aerosols (OOA-I and OOA-II). In the first factor, the mass spectrum was dominated by m/z 41, 43 and 55, which are known as characteristics of hydrocarbons. In addition, a small signal of m/z 60 and 73 was identified, implying that this factor may have contributions from biomass burning. The correlation of this factor with other biomass burning tracers will be further investigated to confirm this finding. Based on the intensity of the m/z values, it can be concluded that OOA-I corresponds to more aged secondary organic aerosol (SOA) and OOA-II represent fresher SOA with higher volatility. Correlation with measured inorganic parameters in addition to gas phase VOC, NOx and O3 data will further provide useful information about the sources of OA determined in this study.