In this research, the stochastic model (ARIMA) was applied to modeling the monthly temperature values for the area of Mosul, Northern Iraq, by utilizing the time-series data of monthly mean temperature for the period (1995-2010) using the Minitab Software.The performance of the proposed model has been approved through the prediction of monthly temperature values for the years 2009 and 2010 where the correlation coefficient value was (R2=0.99) with the values of the actual data for the same years. Then theproposed model was used topredict the mean monthlytemperature values for the years (2011-2012). Due to its importance in the preservation of water resources and rational use in line with the future state of water in the region, this data was used to estimate the future values of Reference Evapotranspiration (ETo) using different empirical methods basedessentially on temperature. The results ofBlaney-Criddle and Hamon methods showed high correlation with ETovalue calculated by Penman-Monteith model.
Key Words: Reference evapotranspiration, Temperature, ARIMA, Time series.