Conceptual study on spatial distribution fluctuations of PM2.5 concentration on the Korean Peninsula in winter related to synoptic meteorological clusters

To classify prevailing synoptic patterns on the Korean Peninsula during the seasonal PM_2.5 management from 2015 to 2019, K-means clustering analysis was performed using 925 hPa geopotential height from the National Centers for Environmental Prediction (NCEP) Final Analysis (FNL) data. Additionally, we analyzed synoptic patterns based on the PM_2.5 pollution standards for each cluster to understand the differences in high and low PM_2.5 concentrations under similar synoptic meteorological conditions. The synoptic patterns were classified into five types. Clusters (C1, C3, C4) with a west-high, east-low pressure distribution experienced worsened PM_2.5 levels due to a reduction in the east-west pressure gradient. This weakening caused the northwesterly winds to slow down, leading to the accumulation of pollutants. Clusters (C2, C5) with a south-high, north-low pressure pattern were characterized by conditions of high PM_2.5, when a high-pressure system over the southeastern region of the Korean Peninsula approached or receded from the Korean Peninsula, accompanied by a reduction in the pressure gradient. This study confirms that the intensity and location of high pressure can lead to variations in air quality under similar synoptic meteorological conditions.