Source tagging modeling study of heavy haze episodes under complex regional transport processes over Wuhan megacity, Central China

摘要

Wuhan as a megacity of Central China was suffering from severe, particulate matter pollution according to previous observation studies, however, the mechanism behind the pollution formation especially the impact of regional chemical transport is still unclear. This study, carried out on the Nested Air Quality Prediction Modeling System (NAQPMS) coupled with an on-line source-tagging module, explores different roles regional transport had in two strong haze episodes over Wuhan in October 2014 and quantitatively assesses the contributions from local and regional sources to PM2.5 concentration. Validation of predictions based on observations shows modeling system good skills in reproducing key meteorological and chemical features. The first short-time haze episode occurred on 12 October under strong northerly winds, with a hourly PM2.5 peak of 180 mu g m(-3), and was found to be caused primarily by the long-range transport from the northern regions, which contributed 60.6% of the episode's PM2.5 concentration (versus a total of 32.7% from sources in and near Wuhan). The second episode lasted from the 15-20 October under stable regional large-scale synoptic conditions and weak winds, and had an hourly PM2.5 peak of 231.0 mu g m(-3). In this episode, both the long-distance transport from far regions and short-range transport from the Wuhan-cluster were the primary causes of the haze episode and account for 24.8% and 29.2% of the PM2.5 concentration respectively. Therefore, regional transport acts as a crucial driver of haze pollution over Wuhan through not only long-range transfer of pollutants, but also short-range aerosol movement under specific meteorological conditions. The present findings highlight the important role of regional transport in urban haze formation and indicate that the joint control of multi city-clusters are needed to reduce the particulate pollution level in Wuhan. (C) 2017 Elsevier Ltd. All rights reserved.