Hydrogeological parameter estimation using geophysical data: A review of selected techniques

Subsurface environmental, engineering, and agricultural investigations often require characterization of hydraulic parameters. For example, groundwater flow modeling is often performed through an aquifer whose hydrological properties have been created using stochastic simulation techniques; these techniques use as input both hydraulic parameter point values and spatial correlation structure information. Conventional sampling or borehole techniques for measuring these parameters are costly, time- consuming, and invasive. Geophysical data can compliment direct characterization data by providing multi-dimensional and high resolution subsurface measurements in a minimally invasive manner. Several techniques have been developed in the preceding decade for using joint geophysical- hydrological data to characterize the subsurface; the purpose of this study is to review three methodologies that we have recently developed for use with geophysical-hydrological data to estimate hydrological parameters and their spatial correlation structures. The first two methodologies presented focus on producing high-resolution estimates of hydrological properties using densely sampled geophysical data and limited borehole data. Although we find that high-resolution geophysical data are useful for obtaining these estimates, in practice, geophysical profiles often sample only a small portion of the aquifer under investigation, and thus, the estimates obtained from geophysical data may not be sufficient to completely describe the hydraulic properties of the aquifer volume. The third and last section focuses on using high-resolution tomographic data together with limited borehole data to infer the spatial correlation structure of log-permeability, which can be used within stochastic simulation techniques to generate parameter estimates at unsampled locations. Our synthetic case studies suggest that collection of a few tomographic profiles and interpretation of these profiles together with limited wellbore data can yield hydrological point values and spatial correlation structure information that can be used to aid numerical aquifer model construction, calibration, and flow simulation. As this information is typically only obtainable from extensive hydrological sampling, use of geophysical methods may offer a more efficient and less invasive approach than traditional characterization campaigns. (C) 2000 Elsevier Science B.V.