Spatial and Temporal Characteristics and Main Contributing Regions of High PM2.5 Pollution in Hong Kong

Fine particulate matter (PM2.5) pollution episodes occur frequently in most Chinese metropolises. Because Hong Kong is affected by the East Asian monsoon, its high PM2.5 pollution is potentially contributed to by a combination of local and regional emissions. This study identified the spatial and temporal characteristics and contributing regions of PM2.5 pollution episodes in Hong Kong. We analyzed the hourly monitoring data collected at various monitoring stations in Hong Kong for the year 2014 in addition to spatial distributions of PM2.5 in China in the days preceding high pollution episodes, and we identified the spatial and temporal characteristics of high PM2.5 pollution in Hong Kong. The relative contributions of local and regional sources to high PM2.5 episodes were then quantified. Finally, backward trajectories of air masses arriving in Hong Kong during high PM2.5 episodes were generated to determine the main contributing regions. The results indicated that PM2.5 levels were higher in the northwestern and southeastern parts of Hong Kong and the Yuen Long station observed the most serious pollution, with a highest daily average PM2.5 concentration of 118 mu g m(-3) ; other areas noted relatively lower concentrations, with maximum daily average concentrations less than 100 mu g m(-3). Regional sources accounted for almost 70% of the PM2.5 pollution and were mostly from the northern and northeastern areas of Hong Kong. Six main contributing regions were identified: the northern region of Guangdong province, Henan-Shandong, the Yangtze River Delta, Jiangxi, Fujian, and Anhui-Hubei. Among these, Guangdong province contributed the highest proportion of the pollution. Because PM2.5 is a major air pollutant in Hong Kong, the findings of this study will be useful for forecasting PM2.5 pollution episodes. Moreover, the developed methodology can be extended to other pollutants if their characteristics and main contributing regions can be identified.