Large-scale perspective on the meteorological modulation of air quality over China in winter

Abstract

Rapid economic and industrial growth in China has led to serious air pollution with high concentrations of suspended fine particulate matter (PM2.5), in particular during winter. On a regional scale, meteorological conditions play a major role in modulating the accumulation, transport, removal and transformation of air pollutants. These meteorological conditions are affected by large-scale circulation patterns over China, dominated by the East Asian winter monsoon (EAWM).

However, previous studies of how the large-scale winter circulation modulates air quality in China primarily focused on the North China Plain. The study of regional differences in the dominant large-scale circulation patterns needed to project future climate-driven PM2.5 concentration changes is far from complete. On longer inter-annual timescales, the EAWM is in turn influenced by El Niño–Southern Oscillation (ENSO) and the ensuing Pacific-East Asia teleconnection pattern. Better understanding of the ENSO-EAWM relationship and changes in this relationship under global warming is needed. Firstly, the influence of large-scale circulation on daily PM2.5 variability through its direct effect on key regional meteorological variables over three major populated regions of China (Beijing–Tianjin–Hebei, BTH; the Yangtze River Delta, YRD; the Pearl River Delta, PRD) is examined, based on a new high-resolution air quality reanalysis dataset for China for five winters from December 2013 to February 2018. In BTH, a shallow East Asian trough curbs northerly cold and dry air from the Siberian High, enhancing PM2.5 pollution levels. Weak near-surface southerly winds in eastern and southern China, associated with a weakened Siberian High, suppress horizontal dispersion, contributing to air pollution accumulation over YRD. In PRD, weak southerly winds and precipitation deficits over southern China are conducive to high PM2.5 concentrations. To account for these dominant large-scale circulation–PM2.5 relationships, we propose three new circulation-based indices: a 500 hPa geopotential height-based index for BTH, a sea level pressure-based index for YRD and an 850 hPa meridional wind-based index for PRD. These three indices can effectively distinguish clean days from heavily polluted days in these regions, assuming PM2.5 variability is solely due to meteorology. Subsequently, the influence of the winter large-scale circulation on daily PM2.5 concentrations and on the sensitivity of PM2.5 to emissions over major populated regions of China with a focus on YRD is investigated, using the United Kingdom Earth System Model, UKESM1. Weak flow of near-surface cold, dry air from the north and weak inflow of maritime air are conducive to air pollution over YRD for 1999–2019. These provide favourable conditions for the accumulation of local pollution but limit the transport of air pollutants into YRD from the north for 2014– 2019. Based on the dominant large-scale circulation, we construct a new index using the north-south pressure gradient to project PM2.5 concentrations over the region. We show that this index can effectively distinguish different levels of pollution over YRD and explain changes in PM2.5 sensitivity to emissions from local and northern regions. We then project future changes in PM2.5 concentrations using this index under the weak climate and air pollutant mitigation scenario (SSP3- 7.0). We find an increase in PM2.5 concentrations over YRD due to climate-driven circulation changes that is expected to partially offset the effect of emission control measures in the near-term future. Finally, changes in the relationship between ENSO and the EAWM at various global warming levels during the 21st century are examined based on experiments from the Max Planck Institute Grand Ensemble (MPI-GE) that represent the upper boundary of the range of emissions scenario (RCP 8.5). The externally forced component of this relationship (i.e. forced by greenhouse gases and anthropogenic aerosol emissions) strengthens under moderate warming (+1.5 ◦ C), and then weakens for +3 ◦ C warming. These changes are characterised by variations in strength and location of the core of El Niño-related warming and associated deep convection anomalies over the equatorial Pacific leading to circulation anomalies across the Asian-Pacific region. Under global warming, the ENSO–EAWM relationship is strongly related to the background mean state of both the EAWM and ENSO, through changesin the EAWM strength and a shift of the ENSO pattern. Anthropogenic aerosols also play a key role in influencing the ENSO–EAWM relationship under moderate warming (up to 1.5 ◦C).

These results demonstrate the importance of understanding the occurrence of days with elevated PM2.5 concentrations and explaining changes in the sensitivity of PM2.5 to emissions from local and surrounding regions from a large-scale perspective. These findings could help project the occurrence of heavily polluted PM2.5 days during wintertime and assess future emission control strategies for PM2.5 air quality improvement under climate change. Furthermore, the ENSO–EAWM relationship is shown to have a substantial inter-decadal variation under global warming. This may further improve the accuracy of future predictions of air quality in China.