Comparison and overview of PM2.5 source apportionment methods

Serious air pollution in China now has great influence and threat on air quality, visibility, and human health. Accurate identification and quantification of sources play an important role in establishing relative policies, laws and control measures. This study briefly introduces development history and characteristics of three main methods to apportion sources of particles (emission inventory, source- oriented model and receptor model). Based on data platform of PM2.5 and chemical components conducted in Atlanta, USA, this study summarizes characteristics of major source apportionment methods of source-oriented model and receptor model and their performance towards different sources (i.e. vehicle emission, coal combustion, biomass burning, dust and secondary sources), points out the differences of various methods and potential reasons and analyzes their weakness and strength in application, suggesting that developing ensemble method should be an important direction for future source apportionment research.