Inverse modeling of saturated-unsaturated flow in site-scale fractured rocks using the continuum approach: A case study at Baihetan dam site, Southwest China

Modeling saturated-unsaturated flow in fractured rock formations remains a challenging issue due to the difficulties in properly calibrating the unsaturated flow properties for fractured rocks. On the basis of the continuum approach, this study uses inverse modeling to determine the unsaturated hydraulic parameters of fractured rocks, loose sediments, and geologic structures in a large-scale slope under difficult hydrogeologic conditions at the Baihetan dam site, Southwest China. The saturated hydraulic conductivities were determined a priori by packer tests. Field data, including time-series of groundwater level observations and available discharge measurements, is used. A combined procedure of orthogonal design, finite element forward analysis, artificial neural network, and genetic algorithm is adopted for the inverse analysis with high computational efficiency. The numerical results well reproduce the typical features of multiple water tables and wedge-shaped unsaturated zones controlled by gently-sloping tuff zones of low vertical permeability, showing 46-65% of groundwater flowing subhorizontally along each tuff zone and the remaining 35-54% infiltrating vertically and recharging deeper layers. The drastic drawdown of groundwater level after 10-year site characterization is also reproduced. The calculated discharge and water tables agree reasonably well with field observations, indicating that the estimated parameters are representative of the unsaturated hydraulic properties at the site. This work provides an efficient and effective methodology for seeking an optimal solution in large-scale inverse modeling of saturated-unsaturated flow, and further manifests the feasibility of the continuum approach in predicting unsaturated flow in site-scale fractured rock formations.