Abstract: Using the daily mean temperature, maximum and minimum temperature data at 27 observational stations in Guangdong province from 1961 to 2010, the climate change characteristics of high temperature weather were analyzed by methods of a linear trend, a Morlet wavelet analysis, a Mann-Kendall test, an EOF (empirical orthogonal function) and a REOF (rotated empirical orthogonal function). The results show that the mean high temperature days are in a significant decreasing trend from south to north in Guangdong province. However, it is in an increasing trend in the recent 50 years in the most areas of Guangdong province, especially in the middle area. The long term evolution of high temperature days is in a significantly increasing trend, and the turning occurs in 1990s with small gradient before 1990s and large gradient afterwards. The number of high temperature days increases significantly after 2005 and an abrupt change appears in 1998. It has a cycle of 20-23 years for high temperature days, while the variation of cycle is not significant. The standard anomaly field of high temperature days in Guangdong province can be divided into three regions, i.e. central, north and south regions.

Key words: High temperature')">

High temperature, REOF (rotated empirical orthogonal function) method, Abrupt change