Impacts of cold waves on winter mortality in the Korean Peninsula

This study investigated the impacts of cold waves on winter mortality in the Korean Peninsula using long-term meteorological observations and mortality records from 2000 to 2023. Cold-wave days were first examined in terms of their temporal and regional characteristics, revealing substantial interannual variability and a tendency for increased cold-wave occurrence during winters characterized by the negative phase of the Arctic Oscillation. Spatially, cold-wave days were more frequent in inland and high-altitude regions such as Gyeonggi, Gangwon, and Chungbuk, whereas southern coastal regions and Jeju Island experienced few or no cold-wave events.
Mortality analyses showed distinct differences between direct cold-related deaths (T68–T69) and indirect causes, including cardiovascular and respiratory diseases (CVD+Resp). While T68–T69 deaths were concentrated under extremely low temperatures, CVD+Resp deaths occurred under relatively milder conditions, indicating the influence of multiple physiological and environmental factors. Time-lagged analyses demonstrated that excess mortality was most pronounced one day after cold-wave onset (lag 1), and that prolonged cold-wave duration (≥4 days) substantially amplified excess mortality, particularly for CVD+Resp deaths.
To further explore the relative importance and nonlinear effects of meteorological factors, a Gradient Boosting Machine (GBM) model was applied. Partial dependence analyses indicated that solar radiation and wind direction exerted the strongest influence on winter mortality, followed by minimum temperature, dew-point temperature, and diurnal temperature range. Two-way interaction analyses revealed that extreme cold combined with low solar radiation, specific wind regimes (northwesterly and easterly flows), dry atmospheric conditions, or large temperature variability significantly increased mortality levels. Overall, the results demonstrate that winter mortality associated with cold waves is governed by the combined effects of multiple meteorological factors rather than temperature alone. These findings highlight the importance of incorporating compound meteorological conditions into cold-wave risk assessment and public health adaptation strategies.