Sources of PM<sub>2.5</sub> carbonaceous aerosol in Riyadh, Saudi Arabia

Abstract

Abstract. Knowledge of the sources of carbonaceous aerosol affecting air quality in Riyadh, Saudi Arabia, is limited but needed for the development of pollution control strategies. We conducted sampling of PM2.5 from April to September 2012 at various sites in the city and used a thermo-optical semi-continuous method to quantify the organic carbon (OC) and elemental carbon (EC) concentrations. The average OC and EC concentrations were 4.7 ± 4.4 and 2.1 ± 2.5 µg m−3, respectively, during this period. Both OC and EC concentrations had strong diurnal variations, with peaks at 06:00–08:00 LT and 20:00–22:00 LT, attributed to the combined effect of increased vehicle emissions during rush hour and the shallow boundary layer in the early morning and at night. This finding suggested a significant influence of local vehicular emissions on OC and EC. The OC ∕ EC ratio in primary emissions was estimated to be 1.01, close to documented values for diesel emissions. Estimated primary organic carbon (POC) and secondary organic carbon (SOC) concentrations were comparable, with average concentrations of 2.0 ± 2.4 and 2.8 ± 3.4 µg m−3, respectively. We also collected 24 h samples of PM10 onto quartz microfiber filters and analyzed these for an array of metals by inductively coupled plasma atomic emission spectroscopy (ICP-AES). Total OC was correlated with Ca (R2 of 0.63), suggesting that OC precursors and Ca may have similar sources, and the possibility that they underwent similar atmospheric processing. In addition to a ubiquitous dust source, Ca is emitted during desalting processes in the numerous refineries in the region and from cement kilns, suggesting these sources may also contribute to observed OC concentrations in Riyadh. Concentration weighted trajectory (CWT) analysis showed that high OC and EC concentrations were associated with air masses arriving from the Persian Gulf and the region around Baghdad, locations with high densities of oil fields and refineries as well as a large Saudi Arabian cement plant. We further applied positive matrix factorization to the aligned dataset of EC, OC, and metal concentrations (Al, Ca, Cu, Fe, K, Mg, Mn, Na, Ni, Pb, and V). Three factors were derived and were proposed to be associated with oil combustion, industrial emissions (Pb based), and a combined source from oil fields, cement production, and local vehicular emissions. The dominant OC and EC source was the combined source, contributing 3.9 µg m−3 (80 %) to observed OC and 1.9 µg m−3 (92 %) to observed EC.