Analysis of the impacts of synoptic air pressure distribution on the spatial distribution of PM10 in the Korean Peninsula

Abstract:

In this study, we classified wind patterns and analyzed meteorological fields in the Southeastern part of Korean Peninsula during the last 5 years (2014-2018) to understand its impacts on spatial distribution of PM₁₀ concentration. We carried out K-means cluster analysis using u, v components at 900 hPa of NCEP FNL data on whole days from 2014 to 2018 (1795 days) and high PM₁₀ concentration episodes (514 days, over 50 ㎍/㎥). As a results, the wind patterns and meteorological fields were clearly distinguished according to its pressure patterns characteristics. The wind patterns on whole days were classified four patterns(C1-C4) and those associated with high PM₁₀ concentration episodes were five patterns (C_H1-C_H5). We found that the wind patterns of four clusters for every data was divided into five wind patterns on high PM₁₀ episodes and the high PM₁₀ concentrations in the southeastern part of Korean Peninsula from 2014 to 2018 was induced by long-range transport and local emissions. In addition, one cluster containing many high-concentration episodes is corresponded with two clusters when analyzing clusters using only high PM₁₀ concentration data. The two separated clusters have some differences in the distribution of ground level air pressure. These differences often lead to differences of the PM₁₀ spatial distribution on the southeastern part of Korean Peninsula. Therefore, the understanding of pressure pattern over the Korean Peninsula can work the one of major factors in the accuracy of local PM₁₀ prediction.