Dynamic changes of extreme temperature occurrence time in China

The dynamic changes in the occurrence time of temperature extremes is one of the important reasons for some natural disasters. In this study, a set of evaluation indices was defined to quantify the dynamic changes in occurrence time of temperature extremes, including 5 cold indices, 4 warm indices, 2 extreme indices, and 4 phenological indices. The dynamic changes of temperature extremes occurrence time in China were examined through the analysis of daily temperature data obtained from 596 meteorological stations spanning the period from 1960 to 2019. Additionally, an exploration of the correlation between the dynamic changes in temperature extremes occurrence time and atmospheric circulation was conducted. The findings indicate that in China: (1) All cold indices, such as the frost start date (SD0), cold day start date (SD3), very cold day start date (DTN90) and extreme low temperature start date (DTN95), showed a significantly delayed trend (P < 0.05). All warm indices, such as summer day start date (SD25), very hot day start date (DTX90) and extreme high temperature start date (DTX95), displayed a significantly advanced trend (P < 0.01). Phenological start indices, encompassing the growing season start date (SD10) and the warm season start date (SD5), showed a significantly advanced trend (P < 0.001), while the phenological end indices, including the growing season end date (ED10) and the warm season end date (ED5), displayed a delayed trend. However, the extreme indices, including the annual lowest temperature occurrence date (DTNn) and the annual highest temperature occurrence date (DTXx), showed an insignificantly advanced trend; (2) Dynamic changes in the occurrence time of temperature extremes correlated with topography and geographical location. The start date of all cold temperature events, extreme temperature events (DTNn and DTXx), and the end date of phenological events (ED10 and ED5) exhibited a more significantly delayed trend at higher elevations (P < 0.05). DTN90, DTN95, SD10, and SD5 exhibited a more significantly advanced trend at higher latitudes (P < 0.01). DTX90, DTX95, ED10, and ED5 displayed a more significantly delayed trend in higher latitudes (P < 0.001); (3) A correlation was found between dynamic changes in the occurrence time of temperature extremes and atmospheric circulation. The Asian Polar Vortex Intensity Index (APVII) exhibited a significant negative correlation (P < 0.01) with cold indices (SD0, SD3, DTN90 and DTN95). The Arctic Oscillation Index (AOI) demonstrated a significant positive correlation (P < 0.05) with cold indices (SD0, DTN90, DTN95 and DTN99). Additionally, there was a significant negative correlation (P < 0.01) between the Northern Hemisphere Subtropical High Area Index (NHSHAI) and warm indices (SD25 and DTX90). These findings contribute relevant reference information for climate change risk assessment and disaster warning.