برآورد تبخیر– تعرق پتانسیل بر اساس مدل‌های تصادفی سری‌زمانی (مطالعه موردی ایستگاه تبریز)

Evapotranspiration is important components of hydrological cycle, which is important in irrigation systems planning and evaluation of climate change impacts on water planning. In this study, evapotranspiration time series using Penman Monteith was studied in Tabriz synoptic station by the linear stochastic models such as ARIMA and SARIMA. The data had been used since 1986 to 2010. After calculating evapotranspiration, the first 20 years and last 5 years were used for model calibration and validation, respectively. This research was performed in two distinguish approaches. In the first approach all data were considered as unit series with an appropriate fitting model. In the second approach, the monthly series of using data were selected, separately. Then, for each month an appropriate model were fitted to data. The results showed that, the first approach was recognized as a better method regarded the second approach. The R2 and RMSE values were 0.964 and 14.85 first approach comparing to them of 0.963 and 15.52 in the second approach. In conclusion, the R2 and RMSE values of the approaches were relatively similar, with very small (0.67 mm/month) better error rate. In whole the two approaches don’t have significant difference but the first approach is recommendable because it includes fewer computations.