Winter haze over North China Plain from 2009 to 2016: Influence of emission and meteorology.

Analysis of observed PM2.5 in Beijing since 2009 reveals that winter haze over North China Plain (NCP) peaked in 2012 and 2013 and there was an improvement in air quality until 2016. The variation of wintertime PM2.5 from 2009 to 2016 is influenced by both emission changes and meteorology conditions, and we quantified the relative contributions from these two aspects. Sensitivity simulation by GEOS-Chem suggested that emission reductions over NCP in 2013-2017 caused 10% decrease of regional mean PM2.5 concentration in 2016 winter compared to 2012 winter level. We removed emission influence on PM2.5 to get PM2.5 that influenced by meteorology (met-influenced PM2.5). For met-influenced PM2.5, compared to original-observed PM2.5(the US Embassy data), percentage of clean days (daily PM2.5 ≤ 75 μg m-3) decreases while that of polluted (75 μg m-3 < daily PM2.5 ≤150 μg m-3) and heavily polluted (150 μg m-3 < daily PM2.5 ≤ 250 μg m-3) days increases. However, proportion of extremely polluted (daily PM2.5 > 250 μg m-3) days stays unchanged, even if emission reduction is doubled, indicating that the extremely polluted situation over NCP is dominated by meteorological conditions, and emission control from 2013 to 2017 has little effects on the extremely polluted days. We developed an effective haze day index (HDI) to represent the weather conditions conducive to haze days. HDI is constructed based on the normalized near surface meridional wind (V850), temperature difference (δT) between near surface (850 hPa) and upper atmosphere (250 hPa), and the relative humidity at 1000 hPa (RH1000). HDI is skillful to detect 72% of the severe haze days (daily PM2.5 > 150 μg m-3). On average, the anomalously high V850 is the main cause of severe haze, while in 2012 winter, RH1000 favorable for secondary aerosols' formation is the largest contributor to haze.