Spatiotemporal trend and abrupt change analysis of temperature in Iran

The long-term trends in temperature over Iran were examined over 34 synoptic stations during a 50 year period (1961-2010) on seasonal and annual time scales. Two methods, a modified version of the Mann-Kendall test by eliminating the effect of all significant autocorrelation co-efficients and the regional Kendall test, were used in trend identification. The results revealed that the temperature had experienced significant positive trends in autumn, spring and especially summer over the study area. On an annual time scale and in the winter, the highest increasing trends were observed at stations located in the southern and southeastern parts of Iran. For regional analysis of trends, the stations were divided into five clusters based on the K-means clustering method and the silhouette index. Subsequently, the regional trend of temperature was analysed on seasonal and annual time scales using the regional Kendall test. The results of the regional Kendall test also indicated the rising trends in temperature during the last 50 years throughout the country on both seasonal and annual time scales. After verifying the presence of an increasing trend in the temperature time series, the non-parametric Pettitt test was used to detect the change points in the annual and seasonal time scales. The results showed that the change point of average temperature began from the summer of 1972 (Sabzevar station) and continued until the summer of 1998 (Zahedan station). The most frequent change point occurrence was between the years 1986 and 1994.