Abstract:Based on the daily temperature dataset of 156 weather station records, a set of statistical methods, including trend analysis, Wavelet analysis, Mann-Kendall test, accumulative anomaly analysis, Pettitt test, and principal component analysis was employed to investigate the spatial and temporal variations of the extreme temperature events from 1961 to 2014 in the coastal area of China. Results of the trend analysis demonstrated an upward trend in warm extremes and a downward trend in cold extremes as well as diurnal temperature range (DTR). The decadal trend rates of the night extremes were obviously higher than those of the day extremes in the coastal area of China. Generally, a decrease was observed in the multi-year averages of the frost days (FD0), ice days (ID0) and diurnal temperature range (DTR), and an increase was observed in the mutil-year averages of the summer days (SU25), tropical nights (TR20), minimum value of daily maximum temperature (TNx), minimum value of daily minimum temperature (TNn) and growing season length (GSL) from north to south. However, a little variation was observed in the multi-year averages of the cool days (TX10p), cool nights (TN10p), warm days (TX90p), warm nights (TN90p), maximum value of daily maximum temperature (TXx), maximum value of daily minimum temperature (TNx), cold spell duration index (CSDI) and warm spell duration index (WSDI) between the sub-regions and the entire coastal area of China. The primary period of extreme temperature indices varied from 2 to 8 years in the sub-regions of the coastal area of China, and no significant decadal period was detected. The mutation time of extreme temperature indices occurred mainly in the 1980s and the 1990s in all sub-regions. Additionally, the cold extremes and minimum values of daily maximum (minimum) temperature mutated earlier than those of the warm extremes and maximum values of daily maximum (minimum) temperature. After mutation, the extreme warm events and extreme value events tended to occur frequently, whereas the occurrence of extreme cold events decreased gradually. The extreme temperature indices holding high load in the first principal component showed strong positive correlations with each other and exhibited high contributions to the daily average and daily maximum (minimum) temperature. On the other hand, the extreme temperature indices holding low load in the first principal component presented weak correlations with other extreme indices and showed low contributions to the daily average and daily maximum (minimum) temperature.