The Investigation of Variations and Time Series Modeling of Annual Temperature of Tehran's Station

Global warming has been associated with the changes of the pattern of the spatial-timing of temperature. Therefore, changes and fluctuations of this element have a scientific and practical importance. One of the approaches to examining the temperature variations is the use of statistical methods. Climatologists for the investigation of temperature variations use the time series methods. The main purpose of this study was to investigate the annual temperature changes of Tehran's station by using time series. For this purpose, the statistical data from the period of 1951-2008 were used. The results showed that in the spectral analysis with the confidence level of .95, cycles of 1, 21, and 22 were significantly observed. The annual temperature time series of the station also indicates the rising trend. Then, modeling in the family of polynomials also showed that the pattern of the temperature line of Tehran over the past 55 years has increased (.72±.08). Also, using a Markov chain estimated the possibility of change in situations from the cold year to hot year and vice versa 2.2 and 1.82, respectively. In other words, every 2.2 years a cold year and every 1.82 years, a warm year occurs. Finally, in the Arima modeling, after examining the Akaike value, the model of (1, 1, and 0) was selected as the most appropriate.