A.R Ghodrati - Faculty Member of Agricultural and Natural Research Center of Guilan
Mhoammd taher - Department of Soil science, Karaj Branch, Islamic Azad University, Karaj, Iran

The purpose of analysis of time series is finding changes model and forecasting. In this paper, temperature of theCaspian southern coasts was modeling by sARIMA or seasonal Autoregressive integrated moving average.Therefore monthly mean temperature related to Anzali, Ramsar and babolsar synoptic stations with long termdataset has been studied through (1955 – 2006).At first in order to study the climate change process in theregion, three phenomena; homogeneity, trend and discontinuity of the series and temperature extreme valueswere analyzed. On the other hand, each series were analyzed by multiplicative decomposition method and themain components of the time series, namely Trend (trt), Cyclical changes (clt), Seasonal changes (snt) andIrregular changes (irt) were determined. Then sARIMA model was performed and temperature was predicted.Inpreparation of the time series to use ARIMA model, the time series transformed to normal and stationary seriesusing Box-Cox and differencing method. After selection of some suitable models and estimation of parametersby maximum likelihood method, independence and normality of model residuals( t a ) should be considered.Then Akaike information criteria (AIC) and (SBC) determine the best model.sARIMA (1,0,0)(0,1,1)12 wasselected for Anzali and Babolsar and sARIMA (0,0,2)(0,1,1)12 was selected for Ramsar mean monthlytemperature. Temperature at all stations was predicted with the high accuracy, in comparison with the actual datain two years 2005 and 2006 as the gauge by sARIMA model and multiplicative decomposition method.Correlation coefficient between the actual and fitted data was nearly 0.97 and the absolute and relative errorswere very small.

time series, Caspian, Anzali, Ramsar, Babolsar