High-Resolution Sampling and Analysis of Air Particulate Matter in the Pearl River Delta Region of Southern China: Dynamic Variations and Source Identification

Hourly-resolved PM_2.5 and PM_10-2.5 samples were collected in the industrial city Foshan in the Pearl River Delta region, China. The samples were subsequently analyzed for elemental components and black carbon (BC). A key purpose of the study was to understand the composition of particulate matter (PM) at high-time resolution in a polluted urban atmosphere to identify key components contributing to extreme PM concentration events and examine the diurnal chemical concentration patterns for air quality management purposes. It was found that BC and S concentrations dominated in the fine mode, while elements with mostly crustal and oceanic origins such as Si, Ca, Al and Cl were found in the coarse size fraction. Most of the elements showed strong diurnal variations. S did not show clear diurnal variations, suggesting regional rather than local origin. Based on empirical orthogonal functions (EOF) method, 3 forcing factors were identified contributing to the extreme events of PM_2.5 and selected elements, i.e., urban direct emissions, wet deposition and a combination of coarse mode sources. Conditional probability functions (CPF) were performed using wind profiles and elemental concentrations. The CPF results showed that BC and elemental Cl, K, Fe, Cu and Zn in the fine mode were mostly from the northwest, indicating that industrial emissions and combustion were the main sources. For elements in the coarse mode, Si, Al, K, Ca, Fe and Ti showed similar patterns, suggesting same sources such as local soil dust/construction activities. Coarse elemental Cl was mostly from the south and southeast, implying the influence of marine aerosol sources. For other trace elements, we found vanadium (V) in fine PM was mainly from the sources located to the southeast of the measuring site. Combined with CPF results of S and V in fine PM, we concluded shipping emissions were likely an important elemental emission source. An high-resolution meteorological dataset (1km x 1km) coupled with potential source contribution analysis (PSCF) was used to identify source emission locations. Health exposure risk of the elemental components were also estimated.