Trend and Time Series Analysis of Reference Evapotranspiration (Case Study: Khorram Abad Plain)

This study examines the trend of ET0 changes using the Mann-Kendall nonparametric test and predicts it using time series analysis. To calculate FAO-Penman-Monteith ET0, Tmin and Tmax, RHmin and RHmax, sun-hours and wind-speedy of synoptic station Khorramabad during the 1991-2017 were used. The results showed that the trend of annual change in reference evapotranspiration with the Mann-Kendall statistics 0.2, was not statistically significant. The trend of monthly changes showed that the trend in September and July was decreasing by Mann-Kendall statistics -0.2 and -0.7, and in the rest of the months, it was increasing. In the stagnation time series check, the trend and ACF graph, showed an increasing trend in the series. for staging the series, the method of differentiation was used and the data were stagnant withfirstfirst-ordered differentiation. For modeling with SARIMA model, from period 1991-2015 for the modeling and from period ‎‎2016-2017 model verification was used, and condition of normality, accidentally and independent ‎of residual of the fitted model was examined. The results showed that the SARIMA‎(0,1,1)(1,0,1) for ET0 modeling, has the best accuracy. The RMSE and R2 ‎values for prediction with this model were 0.674 and 0.97 mm/month, which indicates that the ‎model is accurate.‎