Exploring the regional pollution characteristics and meteorological formation mechanism of PM2.5 in North China during 2013–2017

[Display omitted] •The domain-averaged PM2.5 in North China decreased significantly from 2013 to 2017.•Firstly established two automatic algorithms to identify regional pollution types and events.•The relationships of synoptic circulations and regional PM2.5 levels were quantified.•Starting areas, synoptic and local meteorological mechanisms were analyzed based on persistent regional pollution events.•Quantifying the effects of weather changes on the interannual PM2.5 variations. In the last decade, North China (NC) has been one of the most populated and polluted regions in the world. The regional air pollution has had a serious impact on people’s health; thus, all levels of government have implemented various pollution prevention measures since 2013. Based on multi-city in situ environmental and meteorological data, as well as the meteorological reanalysis dataset from 2013 to 2017, regional pollution characteristics and meteorological formation mechanisms were analyzed to provide a more comprehensive understanding of the evolution of PM2.5 in NC. The domain-averaged PM2.5 was 79 ± 17 µg m−3 from 2013 to 2017, with a decreasing rate of 10 μg m−3 yr−1. Two automatic computer algorithms were established to identify 6 daily regional pollution types (DRPTs) and 48 persistent regional pollution events (PRPEs) over NC during 2014–2017. The average PM2.5 concentration for the Large-Region-Pollution type (including the Large-Moderate-Region-Pollution and Large-Severe-Region-Pollution types) was 113 ± 40 µg m−3, and more than half of Large-Region-Pollution days and PRPEs occurred in winter. The PRPEs in NC mainly developed from the area south of Hebei. The number of Large-Region-Pollution days decreased notably from 2014 to 2017, the annual number of days varying between 194 and 97 days, whereas a slight decline was observed in winter. In addition, the averaged PM2.5 concentrations and the numbers and durations of the PRPEs decreased. Lamb-Jenkinson weather typing was used to reveal the impact of synoptic circulations on PM2.5 across NC. Generally, the contributions of the variations in circulation to the reduction in PM2.5 levels over NC between 2013 and 2017 were 64% and 45% in summer and winter, respectively. The three most highly polluted weather types were types C, S and E, with an average PM2.5 concentration of 137 ± 40 µg m−3 in winter. Furthermore, three typical circulation dynamics were categorized in the peak stage of the PRPEs, namely, the southerly air