Investigation the Pattern of Similar Gradient Regions of Average Annual Temperature Changes of Iran

Temperature is one of the important climatic elements which may be influenced by various geographical factors. Accordingly, investigating about this element can be helpful in order to discover the effects of those factors. The purpose of current study is to investigate the pattern of the regions with similar gradient of annual temperature changes in Iran. Thus, annual temperature of all Iran's territory was adopted for 50 years. Our findings showed that the gradient of annual temperature of Iran can be classified in to five major regions which contains high, moderate, low and increasing, stationary and decreasing trends. Region 1: with high increasing trends; region 2: with moderate increasing trends; region 3: with decreasing trends; region 4: with low increasing trends and finally region 5: with severe increasing trends. Our findings showed that all reigns, except for the third one which showed a decreasing trend, experienced an increasing trend by different values. Investigation of spatial distribution of these regions showed a tempo-spatial patterns over the regions which indicated that the regions are influenced by local factors, such as elevations. The representative pixels of each region was selected by using Principle Component Analysis (PCA) technique, and ARIMA modeling was conducted for representative of each region of similar gradient of changes. The findings showed that the dominant pattern on the cell temperature of each region was so simple. According to the models, temperature in all regions followed stochastic terms of one to two previous years. For instance, two models, including ARIMA (1, 0, 1), and ARIMA (0, 1, 1) were the best models which fitted on annual temperature of regions with decreasing and increasing trends respectively.