Prediction of the mass flow rate of the Hirmand River: The application of the SARIMA time-series technique

A river’s flow prediction and simulation for the sake of predicting the river inflow in future periods are among the main and practical issues of water resource management. The present research assessed and determined a proper model to predict the mass flow rate of the Sistan and Paryan rivers in the Hirmand watershed using the SARIMA time-series model over the period 2000-2018. The statics of the model was checked by the Box-Cox method using the autocorrelation and partial autocorrelation tests. Based on the results and the validation, Akaike and Schwarz-Bayesian criteria, SARIMA (1,1,1)(2,1,0)12 showed the lowest Akaike criterion (1325.91) and Schwarz-Bayesian criterion (1895.40) for the Sistan river and SAMIRA (1,1,1)(1,1,1)12 exhibited the lowest values of Akaike and Schwarz-Bayesian criteria (-328.429 and 328.151, respectively) for the Paryan river. Therefore, they were selected as the best model to simulate the mass flow rate in the Hirmand watershed. The model is applied to simulate the mass flow rate in future periods.