Prediction of unit hydrograph fluctuations for Sahneh aquifer using time series analysis

Nowadays, Box-Jenkins models are considered one of the applicable tools to forecast different time series, especially water table fluctuations. This study addresses the prediction of unit hydrograph fluctuations of Sahneh plain aquifer using Box-Jenkins models. The study area is Sahneh Plain located in the province of Kermanshah. Using monthly data related to 21 piezometers available in the plain, unit hydrograph time-series of Sahneh plain aquifer were computed and plotted for 19 water years (1991-92 to 2009-10). With the use of autocorrelation and partial autocorrelation functions, static time series of multiple seasonal Box-Jenkins models were fitted to the data. After reviewing the results, six top models were introduced. After verifying, ultimately, SARIMA (1,1,4)(3,1,0)12 model was chosen as a best one to forecast unit hydrograph fluctuations in the Sahneh aquifer. The results show that the selected model is capable of predicting seasonal fluctuations, but does not correctly represent the period or cycle.