Seasonal and regional temperature changes in China over the 50 year period 1951–2000

Statistical tests were used to detect the variations of winter (December to February), summer (June to August), and annual mean temperatures of 160 representative stations in China during the period from 1951 to 2000. Changes of temperature in the observed time series that exceed the confidence level of 0.01 were classified as significant temperature rises or drops. The stations with significant temperature rises were classified into several temperature rise types, by the use of a clustering method. A search was made for sudden, or abrupt, temperature changes at each of the 160 stations. Possible reasons for the observed temperature variations are discussed. The statistical analysis reveals that for the 160 stations, significant rises occur in 53.1% of their annual mean temperatures, 51.9% of their winter temperatures, and 13.8% of their summer temperatures. Temperature rises increase with latitude. In contrast, significant drops occur in annual mean temperatures at 1 station, winter temperatures at 3 stations, and summer temperatures at 5 stations. Approximately 50% of stations with significant temperature variations are in the form of abrupt changes. The abrupt temperature changes all occurred in the 1970s, most often in 1979 and 1980. The abrupt changes are shown to be related to the subtropical high in South China Sea. The temperature rises in winter are shown to be linked to the presence of an anomalously strong zonal circulation in Eurasia, and a weak polar vortex, since the 1980s. The summer temperature rise in South China also has a link with the increase in size and intensity of the subtropical high, over the same period. Some of the station temperature drops in summer were related to a weaker summer monsoon. The annual mean temperature rise at 71.8% of the stations was influenced mainly by their winter temperature rises. For 8.2% of the stations, the annual mean temperature rise was connected to their summer temperature rise, and for 12.9% of the stations it was related to both the summer and winter temperature rises.