Estimation of aerosol sources and aerosol transport pathways using AERONET clustering and backward trajectories: a case study of Hong Kong

Abstract

Hong Kong, located adjacent to the rapidly growing urban-industrial region of south China, provides a case of mixed aerosol types (urban, industrial, marine, and long-distance, including dust) from diverse activities and has suffered many serious air pollution episodes over the last decade. However, the sources and transport pathways of aerosols measured and recorded in Hong Kong have not been well researched due to the lack of air quality monitoring stations in east Asia. Here, an integrated method combining Aerosol Robotic Network (AERONET) data, backward trajectories, and Potential Source Contribution Function (PSCF) modelling is used to identify probable transport pathways and magnitudes of source contributions for four characteristic aerosol types. These types, which are dominant in Hong Kong during defined climatic and environmental conditions, are urban fine aerosols, urban mixed aerosols, dust, and heavy pollution. They were defined by clustering a total of 730 AERONET data sets between 2005 and 2008. Results show that aerosol types 1 and 2 (urban fine and urban mixed) are associated with regional fine particulate urban emissions and predominantly local urban emissions, respectively, suggesting that mitigating measures taken within Hong Kong itself would be partially effective. Heavy pollution and dust (types 4 and 3) are more associated with short- and long-distance sources, notably heavy industries in nearby southern Guangdong and the Pearl River Delta region, and desert dust from arid regions in north China. The PSCF map representing dust aerosol type shows a wide range of eastward and southeastward trajectories from northwest China to Hong Kong. Although the contribution of dust sources is small compared to anthropogenic aerosols, a serious recent dust outbreak observed in Hong Kong was associated with an elevation of the air pollution index to 500, compared with 50–100 on normal days. The combined use of clustered AERONET, backward trajectories, and PSCF model can help to resolve long-standing issues about source regions and characteristics of pollutants carried to Hong Kong.