Evident differences of haze days between December and January in north China and possible relationships with preceding climate factors.

Air pollution of fine particulate matter (PM2.5) is severe in winter and causes vast premature deaths in north China. Many previous studies have concerned winter‐mean haze conditions and identified some useful climate factors. However, obvious differences in the number of haze days also existed between December and January (HDJ‐D) that lacked of understandings. The HDJ‐D maximum was 6.6 days, which occurred between December 2005 and January 2006, accounting for more than 45% of the monthly mean number of haze days in these 2 months. Differences of large‐scale atmospheric circulations were distributed like a Rossby wave pattern in mid‐troposphere, with centres in the west of Europe (+), the north to the Caspian Sea (−) and north China (+). Associated meteorological conditions in north China resulted in more adverse ventilation conditions and stronger chemical productions, and thus larger number of haze days in January than in December. In addition, significant relationships between HDJ‐D and three preceding climate factors during 1979–2018 were also revealed, including August–September–October sea ice over the Laptev Sea and the north of the Kara Sea (correlation coefficient = −0.53), November snow depth in eastern Europe and western Siberia (correlation coefficient = −0.44) and sea surface temperature over the Mid‐Atlantic Ridge (correlation coefficient = −0.48). Abundant analyses, basing on the observation data, large‐ensemble climate simulations and atmospheric chemical simulations, all reproduced the influences of each climate factor on large‐scale atmospheric circulations and local meteorological conditions associated with HDJ‐D changes. [ABSTRACT FROM AUTHOR]