Calibration and uncertainty analysis of integrated SWAT-MODFLOW model based on iterative ensemble smoother method for watershed scale river-aquifer interactions assessment

Abstract River-aquifer interaction is a key component of the hydrological cycle that affects water resources and quality. Recently, the application of integrated models to assess the interaction has been increasing. However, calibration and uncertainty analysis of coupled models has been a challenge, especially for large-scale applications. In this study, we used PESTPP-IES, an implementation of the Gauss-Levenberg-Marquardt iterative ensemble smoother, to calibrate and quantify the uncertainty of an integrated SWAT-MODFLOW model for watershed-scale river aquifer interaction assessment. SWAT-MODFLOW combines the Soil and Water Assessment Tool (SWAT), a widely used watershed model, with a three-dimensional groundwater flow model (MODFLOW). The calibration performance of the model was evaluated, and the uncertainty in the parameters and observed ensemble, including the uncertainty in forecasting groundwater levels, was assessed. The results showed that the technique could enhance the model performance and reduce uncertainty. However, the results also revealed some limitations and biases, such as overestimating the groundwater levels in most monitoring wells. These biases were attributed to the limited availability of groundwater level in the first year of the calibration and the uncertainty in groundwater flow model parameters. The river-aquifer interactions analysis shows that water exchange occurs in almost all cells along the river, with most of the high-elevation areas receiving groundwater and flatter regions discharging water to the aquifer. The study showed that PESTPP-IES is a robust technique for watershed-scale river-aquifer modeling that can ensure model calibration and parameter uncertainty analysis. The findings of this study can be used to improve water resources management in watersheds and help decision-makers in making informed decisions.