Ian P. O'Connor

The use of polar organic compounds to estimate the contribution of domestic solid fuel combustion and biogenic sources to ambient levels of organic carbon and PM2.5 in Cork Harbour, Ireland

PM(2.5) samples collected at Cork Harbour, Ireland during summer, autumn, late autumn and winter, 2008-2009 were analyzed for polar organic compounds that are useful markers for aerosol source characterization. The determined compounds include tracers for biomass burning primary particles, fungal spores, markers for secondary organic aerosol (SOA) from isoprene, α-/β-pinene, and d-limonene. Seasonal and temporal variations and other characteristic features of the detected tracers are discussed in terms of aerosol sources and processes. The biogenic species were detected only during the summer period where the contributions of isoprene SOA and fungal spores to the PM(2.5) organic carbon (OC) were estimated to be 1.6% and 1% respectively. The biomass burning markers, and in particular levoglucosan, were present in all samples and attributed to the combustion of cellulose-containing fuels including wood, peat, bituminous and smokeless coal. The contribution of domestic solid fuel (DSF) burning to the measured OC mass concentration was estimated at 10.8, 50, 66.4 and 74.9% for summer, autumn, late autumn and winter periods, respectively, based on factors derived from a series of burning experiments on locally available fuels. Application of an alternative approach, namely principal component analysis-multiple linear regression (PCA-MLR), to the measured concentrations of the polar organic marker compounds used in conjunction with real-time air quality data provided similar trends and estimates for DSF combustion during all seasons except summer. This study clearly demonstrates that, despite the ban on the sale of bituminous coal in Cork and other large urban areas in Ireland, DSF combustion is still the major source of OC during autumn and winter periods and also makes a significant contribution to PM(2.5) levels. The developed marker approach for estimating the contribution of DSF combustion to ambient OC concentrations can, in principle, also be applied to other locations.

Predicting hygroscopic growth using single particle chemical composition estimates

Single particle mass spectral data, collected in Paris, France, have been used to predict hygroscopic growth at the single particle level. The mass fractions of black carbon, organic aerosol, ammonium, nitrate, and sulphate present in each particle were estimated using a combination of single particle mass spectrometer and bulk aerosol chemical composition measurements. The Zdanovskii-Stokes-Robinson (ZSR) approach was then applied to predict hygroscopic growth factors based on these mass fraction estimates. Smaller particles with high black carbon mass fractions and low inorganic ion mass fractions exhibited the lowest predicted growth factors, while larger particles with high inorganic ion mass fractions exhibited the highest growth factors. Growth factors were calculated for subsaturated relative humidity (90%) to enable comparison with hygroscopic tandem differential mobility analyzer measurements. Mean predicted and measured hygroscopic growth factors for 110, 165, and 265 nm particles were found to agree within 6%. Single particle-based ZSR hygroscopicity estimates offer an advantage over bulk aerosol composition-based hygroscopicity estimates by providing additional chemical mixing state information. External mixing can be determined for particles of a given diameter through examination of the predicted hygroscopic growth factor distributions. Using this approach, 110 nm and 265 nm particles were found to be predominantly internally mixed; however, external mixing of 165 nm particles was observed periodically when thinly coated and thickly coated black carbon particles were simultaneously detected. Single particle-resolved chemical information will be useful for modeling efforts aimed at constraining cloud condensation nuclei activity and hygroscopic growth.

A dynamically segmented linear data model of the coast

This article discusses the rationale behind choosing a dynamically segmented linear data model as the basis for a coastal zone geographic information system (GIS). The salient features and conceptu...