An Integrated Model of Surface Water and Groundwater Interactions at Yi-lan Area in Northeastern Taiwan

Interaction between surface water (SW) and groundwater (GW) plays an important role in local society and ecosystem, especially in areas with limited water resources. Historically, hydrologic simulations have not accounted for feedback looks between the GW system and other hydrologic processes. Integrated SW-GW modelling can provide a comprehensive and coherent understanding on basin-scale water cycle and better manage the water resources for sustainable usage. At Yi-lan area, hydrological modelling has been performed for both the entire SW and GW systems along, but fully integrated SW-GW modeling has not been attempted for this area. In order to enhance the efficiency of water useage, a coupled GW and SW flow model (GSFLOW), developed by U.S. Geological Survey, is selected as the numerical model to simulate the major processes of the hydrologic cycle. GSFLOW integrated PRMS with MODFLOW-2005 which perform surface hydrology simulation and 3-D groundwater simulation, respectively. The data of solar radiation, land use, precipitation, temperature, river stage, stream flow rate, groundwater level, and digital elevation model were collected from 2004-2012 to develop the simulation model. The coupled GSFLOW model is calibrated by automatic parameter estimation approach of using streamflows and groundwater levels. The singular value decomposition (SVD) method is performed to avoid the instability of solution during the model calibration. The calibrated results show that the state variables and fluxes in basin-scale water cycle can be simulated with high spatial and temporal resolutions, and all the important hydrologic processes can be characterized simultaneously in an integrated framework. The scenarios with different precipitation distributions and temperature patterns are conducted on the calibrated model to forecast the dynamic variations of hydrologic processes in the entire water basin. This study clearly demonstrated the benefits of using a physically based integrated SW-GW model, instead of a more conceptual one, in addressing complex real-world systems.