A Comparative Analysis of Water Table Oscillations using GMS Software and Time-series Model in Ajabshir Plain

Prediction of water table oscillation, especially in arid and semiarid areas is essential for better planning. In this study, monthly data of water table elevation, during the period of 1380 to 1390 for Ajabshir plain with an area of 130 km2 were used by application of two methods to estimate and predict the water table elevation. The date of 1380 to 1390 were used for estimating and the water table elevations were predicted for the three years of 1391 to 1393. The first method is solving partial differential equation for consecutive time steps and the second one is time series model. The GMS software was used for numerical solution of the differential equation by finite difference method. The correlation coefficient value of 0.9 and RMSE value of 0.41 between the estimated and observed amonts of groundwater levels were obtained using this method. The best fitted model for the water table elevation data using time series was ARMA (3, 2), correlation coefficient and RMSE values of this method were 0.85 and 0.49 respectively. According to the evaluation criteria, the partial differential equation method was more accurate than the time series method.