Nonstationary Temperature Extremes in South Korea: Roles of Global Warming and Large-Scale Climate Variability

Regional heat extremes are intensifying under global warming, yet their evolution reflects the combined influence of anthropogenic forcing and internal climate variability. Here we investigate long-term changes in temperature extremes across South Korea using daily observations from 60 weather stations during 1974–2023. A non-stationary Generalized Extreme Value (GEV) framework incorporating global mean surface temperature (GMST) as a covariate reveals clear non-stationary behavior in both summer and winter extremes across event durations of 1–15 days. Winter cold extremes have weakened markedly, accompanied by a broadening of temperature distributions and a reduced frequency of severe cold events, indicating a systematic decline in cold-air outbreaks. In contrast, summer extremes exhibit greater societal relevance, with tropical night events showing the strongest sensitivity to global warming. These events have increased rapidly in both frequency and persistence, particularly along the west and south coasts, where enhanced moisture availability and elevated nighttime temperatures amplify heat stress. This intensification is likely linked to the strengthening of the North Pacific Subtropical High and warming of surrounding seas, which together enhance moisture transport and suppress nocturnal cooling over the Korean Peninsula.

Despite the persistent warming trend, heatwave activity exhibits pronounced decadal variability. A relative lull from the late 1990s to the early 2010s (period 1, P1) was followed by a rapid resurgence thereafter (period 2, P2). This shift is associated with changes in the North Atlantic Oscillation (NAO), whose planetary-scale teleconnections modulate atmospheric circulation over Northeast Asia. Positive NAO phases strengthen anticyclonic circulation over the Korean Peninsula, enhancing subsidence and surface warming. During P1, the NAO weakened with reduced interannual variability, likely linked to tropical Pacific forcing. In contrast, during P2, it strengthened, potentially driven by Atlantic forcing in conjunction with a phase shift in tropical Pacific variability. These results demonstrate that regional temperature extremes arise from the interplay between externally forced warming and internally generated climate variability, underscoring the importance of accounting for both processes in future climate risk assessment and adaptation.