ABSTRACT
ABSTRACT: The modeling and understanding of temporal variability of groundwater discharge are important with respect to efficient water resources management, especially in low flow seasons. Baseflow, which is a major component of streamflow, can be considered as groundwater inflow or discharge. In this study, two stochastic modeling approaches, seasonal autoregressive integrated moving average (ARIMA) and Thomas-Fiering (TF) models, are used to model monthly baseflow time series in six different basins in the southwest of Iran. One-step-ahead forecasts for the test portion of the time series are generated using the selected set of candidate models. The major objective of this research is to understand the stochastic behavior of baseflow time series and develop reliable forecastingmodels. The results show that theARIMA(1,0,0)(0,1,1)model provides themost accurate forecasts for the particular basins and thismodel adequately describes the structure of the stochastic behavior of the monthly groundwater discharge time series.
