28PROBSTAT Analyzing extreme diurnal temperature range in spring on the Korean peninsula using a sliding window approach

We analyzed extreme values of the diurnal temperature range (DTR), daily maximum and minimum temperature, and daily average temperature during the spring season. The data were collected from 16 automated synoptic observation system (ASOS) stations in South Korea over a 59-year period spanning from 1964 to 2022. The sliding window approach was employed to capture the nonstationarity of the extreme values. By applying the peak over threshold (POT) method to sample the extremes, we model the POT extremes using the generalized extreme value (GEV) distribution. Trends in POT extremes are detected using the Mann-Kendall trend test, and the DTR has decreased due to a significant increase in the minimum temperature. Strong increasing trends in maximum, minimum, and average temperatures have appeared at most of the stations. The extreme DTR, as indicated by the return periods, shows clear nonstationarity, suggesting potential changes in the frequency of DTR increase or intensity decrease. Despite the reduction in extreme DTR due to climate change, major stations still exhibited a high DTR of approximately 17 K or more, which could pose physical risks.