Understanding the inter-city causality and regional transport of atmospheric PM2.5 pollution in winter in the Harbin-Changchun megalopolis in China: A perspective from local and regional

Harbin-Changchun megalopolis (HCM) is the typical cold urban agglomeration in China, where PM2.5 pollution is still serious in winter against the backdrop of continuous improvement in annual air quality in China. To further understand interactions of atmospheric pollution among HCM cities, inter-city causality and regional transport of PM2.5 in the winter in the HCM were comprehensively investigated by using convergent cross mapping (CCM) and CMAQ-BFM methods. CCM analysis results suggest strong bidirectional causal relationships between cities in the HCM, and the causality during polluted episodes were significantly larger than that during clean period. In addition, the influence on local PM2.5 from the HCM western cities were larger than that from cities in the southeast. Inter-city and regional transport contributions results demonstrated that although local emission were the largest contributors among 14 sub-regions for most HCM cities, interactions among cities were strong. Regional transport (42.8%–77.4%) largely contributes to HCM cities’ PM2.5 concentrations. Among three regions outside the HCM, NMG (including part of inner Mongolia and Baicheng city in Jilin, 9.1%) was the largest contributor to the PM2.5 concentration in the whole HCM, followed by JLS (including Liaoning Province, Tonghua and Baishan cities in Jilin province, 5.1%) and HLJ (including cities of Heihe, Yichun, Jiamusi, Hegang, Shuangyashan, Jixi, Qitaihe in the Heilongjiang province, 3.8%). Regional transport contribution to the most HCM cities increased significantly from excellent to heavily polluted days. Furthermore, close relationships between transport paths/intensity and wind direction/speed in studied region suggests that we can quantitatively guide the regional joint emergency prevention and control before and during heavily polluted events based on regional weather forecasts in the future.