[Characteristic Analysis and Source Apportionment of VOCs in Urban Areas of Beijing in Summer].

Refined characterization of volatile organic compound (VOCs) components and source apportionment can provide scientific and effective support for ozone (O) pollution prevention and control. Using hourly-resolution VOCs online data monitored at urban sites in Beijing from July to August in 2020, the chemical characteristics of VOCs and ozone formation potential (OFP) in environmental receptors during high and low ozone concentration periods were analyzed, and refined source apportionment was conducted with a positive matrix factorization (PMF) model. The results showed that the average [total volatile organic compounds (TVOCs)] at the monitoring sites during the observation period was 12.65×10, and the (TVOCs) during the high and low ozone concentration periods were 13.44×10 and 12.33×10, respectively, with an OFP of 107.6 μg·mand 99.2 μg·m, respectively. Ozone production was controlled by VOCs, with the highest reactivity of aromatic hydrocarbons and the top three species contributing to OFP being isoprene, toluene, and -xylene. The main sources of VOCs in environmental receptors during low O periods included vehicular emissions (26.4%), background emissions (15.7%), solvent using (13.0%), auto repair (12.8%), secondary generation sources (9.7%), biomass combustion (6.1%), printing industry (5.7%), LNG-fueled vehicles (5.5%), and vegetation emissions (5.0%), of which background emissions, secondary generation, and printing industry sources have been little discussed in recent studies of VOCs source apportionment in Beijing. The contribution of auto repair sources and secondary generation sources increased by 3.4% and 2.6%, respectively, during the high O periods compared to those during the low O periods, and vehicular emissions remained the most significant source of VOCs contribution in the urban area of Beijing. Vegetation emissions rose from 07:00 pm and reach a maximum in the late afternoon. The contribution of background emission sources was less variable; vehicular emissions and LNG-fueled vehicle sources showed a morning and evening peak, with a relatively low contribution in the afternoon.