3D hydraulic conductivity modeling of fractured granitic body using geostatistical techniques and its application to regional groundwater flow analysis

Kubo, Taiki*; Koike, Katsuaki*; Liu, C.*; Kurihara, Arata*; Matsuoka, Toshiyuki

Numerical simulations have been the most effective method for estimating flow pattern, flux, and flow velocity of the groundwater to precisely characterize large-scale groundwater systems. Spatial modeling of the 3D distribution of hydraulic conductivity over a study area is indispensable to obtain accurate simulation results. However, such spatial modeling is difficult in most cases due to the limitations of hydraulic conductivity data in terms of their volume and location. To overcome these problems and establish an advanced technique, we adopt geostatistics and combine a fracture distribution model with measured conductivity data, selecting the Tono area situated in Gifu Prefecture, central Japan for the field study. The size of the main target area covers 12 km (E-W) by 8 km (N-S), with a depth range of 1.5 km, and it is chiefly underlain by Cretaceous granite. Because the distribution of 395 hydraulic test data acquired along the 25 deep boreholes was biased, the data values were compared to the dimensions of simulated fractures using GEOFRAC. As a result, a positive correlation was identified. Using a regression equation for the correlation, hydraulic conductivity values were assigned to every simulated fracture. Then, a sequential Gaussian simulation (SGS) was applied to construct a 3D spatial model of hydraulic conductivity using those assumed values and actual test data. The plausibility of the resulting model was confirmed through the continuity of high and low permeable zones. The next step is a groundwater flow simulation using MODFLOW and the model. The simulation results were regarded to be appropriate because distribution of hydraulic head, locations of major discharge points, and anisotropy of hydraulic behavior of the Tsukiyoshi fault correspond to the results of observations.