Abstract
The negative impact of extreme high-temperature days (EHDs) on people's livelihood has increased over the past decades. Therefore, an improved understanding of the fundamental mechanisms of EHDs is imperative to mitigate this impact. Herein, we classify the large-scale atmospheric circulation patterns associated with EHDs that occurred in South Korea from 1982 to 2018 using a self-organizing map (SOM) and investigate the dynamic mechanism for each cluster pattern through composite analysis. A common feature of all SOM clusters is the positive geopotential height (GPH) anomaly over the Korean Peninsula, which provides favorable conditions for EHDs through adiabatic warming caused by anomalous downward motion. Results show that Cluster 1 (C1) is related to the eastward-propagating wave train in the mid-latitude Northern Hemisphere, while Cluster 2 (C2) and 3 (C3) are influenced by a northward-propagating wave train forced by enhanced convection in the subtropical western North Pacific (WNP). Compared to C2, C3 exhibits strong and eastward-extended enhanced convection over the subtropical WNP, which generates an anomalous high-pressure system over the southern part of the Kamchatka Peninsula, reinforcing EHDs via atmospheric blocking. Our results can contribute to the understanding of East Asia climate variability because wave trains influence the climate dynamics of this region.