Distribution Characteristics and Source Identifications of Black Carbon and Brown Carbon Aerosols in Typical Cities in Northern and Southern China-Xi'an and Hong Kong

Abstract

Light-absorbing carbonaceous (LACs) aerosols play important roles in climate change through direct and semi-direct radiative forcing, including the most potent absorber of black carbon (BC) at the visible wavelengths (400-700 nm) and strong light-absorbing organic compounds (brown carbon (BrC)) at the near-ultraviolet and blue wavelengths. This group of studies investigated the mass concentrations, optical properties and source identifications of BC and BrC and examined their interaction with gas pollutants (NO2, NOx, SO2, O3, and CO) and PM2.5 chemical compositions at various meteorological conditions. The study was conducted in two geographically distinct cities-Xi’an and Hong Kong (hereafter referred to as XA and HK) - in China, during polluted periods in 2014-2016. To understand the BC sources and their transport paths, the relationship between BC and other pollutants, the backward trajectory of BC, the potential source contribution function (PSCF) model and an advanced source analysis model of multilinear engine (ME-2) were established and analyzed during the sampling periods. Advanced simulations and near-source experiments were also conducted to predict the UV-visible absorption spectrum of different chemical structures of BrC in this study. The major findings of this dissertation were described below.

1) Our results have shown distinctive seasonality for LACs’ (including BC and optical-BrC) optical properties and their formations during different periods in Xi’an. The chemical-derived light extinction coefficient (Chemical-bext) was the highest during winter haze (WH) days (1019.2 Mm-1), and decreased to ~1/4 during summer haze (SH) days (237.6 Mm-1). The light-absorbing coefficient of BC (babs-880nm, BC) contributed a high fraction to both WH bext (10.9%) and SH bext (8.7%), while Optical-BrC’s light absorption coefficient (babs-370nm, BrC) contributed ~4.0% to WH bext, which was reduced by 10 orders of magnitude for SH bext (0.4%), indicating that the UV light absorptions changes significantly due to the seasonal variations from different sources. Further, the higher haze/non-haze ratios were found in babs-880nm, BC (1.78 for winter and 1.40 for summer) than those in babs-370nm, BrC (0.99 for winter and 1.01 for summer), implying that the particle’s light absorptions in the visible spectrum were enhanced from non-haze to haze periods. There was a remarkable difference of diurnal pattern between babs-370nm, BrC and babs-880nm, BC during haze periods, showing that BC leads to a severe visibility reduction during traffic rush hours. The abundant BrC can be attributed to varied secondary formation processes, i.e., SH BrC at noon was generally induced by photochemically derived secondary organic aerosol (SOA) and WH BrC peaked at 19:00-20:00 due to aqueous-SOA.

2) This study investigated the “roadside-to-ambient” evolution of BC and optical-BrC in the typical urban atmosphere of Hong Kong. The high AAC_ Full spectrum value (>1) were found in both roadside and ambient particles, demonstrating the presence of BrC. Particle-bound polycyclic aromatic hydrocarbons (p-PAHs), an indicator of BrC, showed a significant UV light-absorbing ability. The average BC and p-PAHs concentrations were 3.8 μg·m−3 and 87.6 ng·m-3 respectively at the roadside site, but were only 1.5 μg·m−3 and 18.1 ng·m-3 at the ambient site, suggesting BC and p-PAHs concentrations were strongly driven by the local traffic emissions. The significantly positive correlation (R=0.93) and high slope (S=41.3) between BC and p-PAHs loadings were observed at the roadside, demonstrating the likelihood of common emission sources for BC and p-PAHs. By contrast, a distinct reduction of the correlation coefficient (R=0.58) and lower regression slope (S=19.6) of BC versus p-PAHs were found at the ambient site, indicating that more complex processes affect the mixture of BC and p-PAHs in ambient aerosols as they transport and mix with fresh nearby pollutant emissions. The converse trend of non-sea-salt sulfates (nss-SO42-)/BC was observed in comparison with p-PAHs/BC ratios, supporting the concept that enhanced secondary inorganic coatings covered the inner layer of p-PAHs was nss-SO42-. Moreover, we found that the (NH4)2SO4 was the principal SIA which formed during atmospheric aging processes. The elevated babs-BrC,370nm and AAC values at UV wavelengths (AAC_UV band) in ambient particles reflected the impacts of aging process from roadside to ambient contributing to significant changes in PM2.5 UV light absorptions. The large difference between babs-370nm, BrC and babs-880nm, BC in fresh roadside particles appears to be due to the abundance of photochemical secondary BrC at noon time. Whereas in aged ambient particles, both aqueous secondary organic aerosol (SOA) and photochemical SOA contributed to the peak values of babs-370nm, BrC at midnight and noon, respectively, highlighting that secondary BrC has different sources, and shows the impact of atmospheric aging on BrC and BC properties as well as their roles in aerosol light absorption over Hong Kong regions.

3) For parallel observations in winter, both high BC (7.9 μg·m−3) and PM2.5 (182 μg·m−3) concentrations showed high levels in Xi’an (XA) and were much higher than those in Hong Kong (HK) (BC, 3.2 μg·m−3; PM2.5, 34.5 μg·m−3, respectively). In contrast, the contribution of BC to PM2.5 in Hong Kong reached to 10.7%, which was ~1.5 times of that in Xi’an (7.6%). These results emphasized that BC played an important role in HK PM2.5. The statistical distribution of HK babs-880nm, BC was tightly correlated with vehicle emissions during the daytime, which peaked in the heavy traffic-time, while the distribution XA babs-880nm, BC was flat over the whole day as results of increasing stationary BC sources and stable meteorological conditions. 72-h back trajectories and PSCF analysis showed that the anthropogenic pollution of XA BC mainly originated from local emissions while nearly half HK BC contributed from more distant sources, such as the ocean ship’s emissions. These anthropogenic BC sources were found to be regional in nature based on the ME-2 analysis: XA BC source consisted of three parts: 24.13% from coal burning, 18.72% from biomass burning and 37.55% from vehicle emissions; while virtually the majority of HK BC contributions refer to vehicle and ship emission (82.13%) and only ~2.31% were coal burning which originated from the emissions by residential combustion, industry and power plants in inland China.

4) To better quantify the optical-chemical properties and the sources of PM2.5 BrC, methanol-soluble aqueous-BrC was extracted in this study and relationships with various PM2.5 sources were established during a winter period. The babs-365nm, BrC and mass absorption coefficient (MAC-BrC) were 55.6 Mm-1 and 2.4 m2·g-1 in Xi’an, which were ~9 times and 2 times of those in Hong Kong, respectively, indicating that heavy loadings of BrC were presented in Xi’an. Good correlations (R>0.7, p<0.001) among babs-365nm, BrC, water-soluble K+ and primary organic carbon (POC) were found. Unlike the high correlation coefficient of babs-365nm, BrC and secondary organic carbon (SOC) (R=0.73, p<0.001) in Hong Kong, the XA babs-365nm, BrC showed much poorer correlation with SOC (R=0.4, p<0.001). These phenomena emphasized that XA BrC mainly originates from primary combustion while the HK BrC were both influenced by primary and secondary sources in winter months. Source profiles of primary-BrC optical parameters in Xi’an were detected, indicating that open biomass burning emissions contribute more BrC absorbance than the fossil fuel burning and compressed biomass burning. This model predicted that the POA of biomass or wood burning showed slightly light absorptions once emitted. As for SOA, the absorption of toluene SOA was much enhanced with higher levels of NOx and longer aging time.