Influence of Local Circulation on Short-term Variations in Ground-level PM_(2.5) Concentrations

10.4209/aaqr.240042

Su Jeong Lee；Sang-Hyun Lee；Hyung-Jin Choi；Joowan Kim；Maeng-Ki Kim

Local circulation ； Particulate matter ； Coastal regions ； K-means clustering

Aerosol and Air Quality Research

24卷10期（2024 / 10 / 01）

1 - 14

英文

Local air quality is greatly influenced by large- and small-scale weather systems through transport, deposition, and chemical transformation of emissions. Local circulation, in particular, can play a significant role under weak synoptic-scale forcing. To examine the influence of local circulation on daily ground-level PM_(2.5) concentrations, we utilized surface wind features observed at coastal stations in Midwest Korea, which hosts large industrial complexes and is located downwind of the Seoul Metropolitan area. Using K-means clustering, three circulation patterns were identified for the winter of 2021-2022, including one pattern under strong synoptic-scale forcing (Synoptic Cluster) and two local patterns (Sea Breeze Cluster and Stagnation Cluster). Each cluster is characterized by its unique wind patterns and different contributions to local air quality. The Stagnation Cluster, characterized by weak north-easterly winds with a comparatively short transport distance, was found to be most strongly linked to high PM_(2.5) levels, accounting for 57% of the high PM_(2.5) days (> 35 µg m^(-3)) during the 2021-2022 winter. Additionally, we discovered that the three most extreme PM_(2.5) events were all members of the Stagnation Cluster and that several consecutive stagnant days preceded each of these cases, facilitating local accumulations of nearby anthropogenic emissions. Overall, our findings emphasize that local air quality cannot be fully explained by synoptic-scale analysis, but can be better understood through the analysis of local circulation patterns. The study also highlights the importance of utilizing surface measurements and selecting features that can best describe the local circulation patterns in the region for the classification of local circulation, which contributes to better capturing both daily and hourly variability in PM_(2.5) concentrations under different weather regimes.