Long-term source apportionment of PM2.5 across the contiguous United States (2000-2019) using a multilinear engine model.

Identifying PM2.5 sources is crucial for effective air quality management and public health. This research used the Multilinear Engine (ME-2) model to analyze PM2.5 from 515 EPA Chemical Speciation Network (CSN) and Interagency Monitoring of Protected Visual Environments (IMPROVE) sites across the U.S. from 2000 to 2019. The U.S. was divided into nine regions for detailed analysis. A total of seven source types (tracers) were resolved across the country: (1) Soil/Dust (Si, Al, Ca and Fe); (2) Vehicle emissions (EC, OC, Cu and Zn); (3) Biomass/wood burning (K); (4) Heavy oil/coal combustion (Ni, V, Cl and As); (5) Secondary sulfate (SO42-); (6) Secondary nitrate (NO3-) and (7) Sea salt (Mg, Na, Cl and SO42-). Furthermore, we extracted and calculated secondary organic aerosols (SOA) based on the secondary sulfate and nitrate factors. Notably, significant reductions in secondary sulfate, nitrate, and heavy oil/coal combustion emissions reflect recent cuts in fossil-fueled power sector emissions. A decline in SOA suggests effective mitigation of their formation conditions or precursors. Despite these improvements, vehicle emissions and biomass burning show no significant decrease, highlighting the need for focused control on these persistent pollution sources for future air quality management.

The impacts of regional transport on anthropogenic source contributions of PM2.5 in a basin city, China.

PM2.5 pollution events are often happened in urban agglomeration locates in mountain-basin regions due to the complex terra and intensive emissions. Source apportionment is essential for identifying the pollution sources and important for developing local mitigation strategies, however, it is influenced by regional transport. To understand how the regional transport influences the atmospheric environment of a basin, we connected the PM2.5 source contributions estimated by observation-based receptor source apportionment and the regional contributions estimated by a tagging technology in the comprehensive air quality model with extensions (CAMx) via an artificial neural network (ANNs). The result shows that the PM2.5 in Xi'an was from biomass burning, coal combustion, traffic related emissions, mineral dust, industrial emissions, secondary nitrate and sulfate. 48.8 % of the PM2.5 in study period was from Xi'an, then followed by the outside area of Guanzhong basin (28.2 %), Xianyang (14.6 %) and Weinan (5.8 %). Baoji and Tongchuan contributed trivial amount. The sensitivity analysis showed that the transported PM2.5 would lead to divergent results of source contributions at Xi'an. The transported PM2.5 from the outside has great a potential to alter the source contributions implying a large uncertainty of the source apportionment introduced when long-range transported pollutants arrived. It suggests that a full comprehension on the impacts of regional transport can lower the uncertainty of the local PM2.5 source apportionment and reginal collaborative actions can be of great use for pollution mitigation.