Statistical modeling of long-term behavior at Yazd synoptic station

Temperature, as one of the most important climate factors, is known as an appropriate tool with which to find out about climatic changes. This is due to the fact that even a small change in temperature has a considerable effect on echosystems.
In this study, some statistical methods, trending tests through parametric and none-parametric methods, and relative assessment trend methods based on observations are used to find out the temperature behavior. Also, spectrum analyses are done in order to evaluate the trend of Yazd annual temperature. In this article, the annual synoptic data of Yazd station collected between 1962 and 2010 are analyzed, and the ARIMA statistical model is used to achieve research goals. Several pattern tests are examined, and the results indicate that ARIMA (0, 1, and 2) is the most appropriate statistical model. Based on this model, temperature forecast is performed under a 95% confidence interval for the next 10 years. The test results, with a 5% error, show that the variation of the annual temperature is significant in Yazd and has an increasing trend. Also, under a confidence level of 95%, in addition to the first harmonic, the 22th is also another significant harmonic on temperature time series. In other words, this attribution is another factor that moderates the trend in observations.