Application of time series models for evapotranspiration forecasting in Bajgah station

The Penman FAO is one of the appropriate methods for the estimation of evapotranspiration at Bajgah experimental station, Shiraz University. Modeling and forecasting of weekly and monthly time series of evapotranspiration with Penman FAO method using seasonal auto-regressive integrated moving average (SARIMA) models as the linear stochastic models for this station were studied. The time series data of weekly and monthly average of evapotranspiration for the period 1366-1386 were separately considered as the input file of SARIMA models. On the basis of the corrected Akaike's information criterion (AIC) and significant coefficients, and models were selected for weekly and monthly time scales, respectively. The auto-correlation function plots of the residuals for these models have indicated that the residuals are uncorrelated. The weekly and monthly values of evapotranspiration from Khordad 1386 to Ordibehsht 1378 as the test data was forecasted using the fitted models. The results have indicated that the square of correlation coefficient between the observed and predicted values for weekly and monthly time scale is 0.92 and 0.99, respectively. The root mean square error for weekly and monthly time scale is respectively, 0.48 and 0.16 mm.