Source Apportionment Based On Long-term PM Compositional Monitoring At Multiple Stations Over The Pearl River Delta

In the past 30 years,the rapid economic development and urbanization in the Pearl River Delta?PRD?have resulted in significant air pollution problem.Although some achievement has been made in particulate matter?PM?pollution control in recent years,PM pollution is still widely spread with frequent non-attainment occurrences.Difficulties of PM pollution management include its complicated source categories,significant secondary formation,and large contribution from regional transport.In this regard,adopting comprehensive PM pollution source apportionment is of great importance.Based on the PM_2.5 compositional dataset with the largest spatial coverage and PM₁₀compositional dataset with the longest time span in the Pearl River Delta,this study developed an Uncertainty Assessment Framework?UAF?that assists in accurately identifying PM pollution sources in the PRD.The PM pollution sources were then comprehensively investigated in terms of their spatiotemporal distribution,long-term variation trends,local and non-local relative contribution,etc.Main work and conclusions are summarized below.First,the UAF for evaluating PM compositional monitoring data was developed for the first time.This UAF is important in improving the accuracy of PMF source apportionment result.By applying this UAF,this study depicted the optimal PM_2.5 factor profiles in the PRD and differentiated biomass burning from coal combustion sources which cannot be made by regular source apportionment methods.This UAF provided solid foundation for accurate source apportionment of PM in the PRD.Second,a quantitative estimation of local and non-local contribution approach was developed and applied in the PRD by combining the apportioned source contributions with backward trajectory analysis.Results showed that non-local contributions were consistently over half to PM_2.5 levels in the PRD in 2015.In comparison with Hong Kong and Macao stations,the inland stations of Guangdong had higher local contributions.Therefore,local contribution and regional co-control are both needed in air pollution management in Guangdong.The long-term analysis on PM₁₀ in Tsuen Wan of Hong Kong indicated that non-local contributions ranged from 43%-71%with over 60%in most years,indicating significant non-local transport.Regional and super-regional co-prevention and co-control is essential to the air quality management in Hong Kong.Third,the air-shed separation in the PRD in 2015 was conducted and local and non-local contributions in different air-sheds were identified by the similarities and differences in backward trajectory clusters.Results showed that PRD was in the same air-shed for more than half of the time.Inland and coastal areas tend to separate to form two air-sheds.PM_2.5concentration was higher in spring,summer and autumn in inland air-shed and in winter in coastal air-shed.In winter,coastal air-shed was readily receiving long-range transport bypassing Fujian while inland air-shed was more impacted by short-range transport in the direction of Jiangxi.Different air-sheds showed great difference in local and non-local contributions.Therefore,control measures should be formulated considering the local formation and non-local transport characteristics in different air-sheds.Finally,this study applied PSCF method to investigate the potential source areas for PM₁₀ in Tsuen Wan of Hong Kong.Results showed a clear transport corridor extending from central PRD,northeastern Guangdong,central Jiangxi,Fujian,Anhui and southern Jiangsu.Co-prevention and co-control with these regions are needed to reduce PM levels in Hong Kong.This study systematically investigated the main pollution sources of PM in the PRD,and provided new thinking and methodology and scientific support for the targeted internal and external air pollution management in the PRD.