Mesoscale wind field patterns play important roles in episodes of high particulate matter (PM2.5). Previous studies have reported the main synoptic meteorological features associated with high-PM2.5 days with two primary modes: long-range transport (LRT) and stagnation or local contribution (STL). However, due to the lack of more small-scale analyses spanning across multiple administrative regions, statistical quantification of representative mesoscale wind patterns with elevated PM2.5 episodes is needed to explore region-specific emissions reduction strategies in South Korea. This study applied k-means clustering to NCEP FNL reanalysis data from 2016 to 2020 to identify mesoscale wind field patterns associated with high-PM2.5 episodes across all 17 administrative regions of South Korea. Our statistical approach yielded region-specific wind fields for 17 individual administrative regions, and also empirically yielded a total of eight representative patterns, four patterns each for both LRT and STL modes, which were applicable across all regions. The eight patterns were consistent with most previous studies. In addition, we quantified the LTP and STL modes in all 17 regions, and showed that they were more prevalent in the western and southeastern administrative regions of South Korea, respectively. Our results serve as a reference for region-specific mitigation strategies in all administrative regions of South Korea, and will also be useful for analyzing relations between meteorology and high-PM2.5 episodes in the SIJAQ and ASIA-AQ studies.
Read full abstract