The Effects of Anisotropy on ERT Images for Vadose Zone Monitoring

The importance of anisotropic effects has been known almost since the inception of resistivity surveys and yet, these effects are almost always ignored when interpreting surface and cross-borehole (ERT) data. For the special (though common) case of horizontally layered and anisotropic media, resistivity measurements made using only electrodes in a vertical line (as in traditional resistivity well logging), are sensitive only to the horizontal and measurements made on a horizontal plane, (as in traditional surface resistivity measurements) measure the geometric mean of the horizontal and vertical resistivities. Thus, using either borehole or surface resistivity measurements alone produces consistent though somewhat erroneous results. But, ERT measurements cover the full range from vertical to horizontal arrays. Therefore, there is an incongruity between these types of measurements such that for a horizontally anisotropic earth there appears to be no coarsely layered isotropic earth that will closely match all of the measurements. This problem appears more prevalent in the vadose zone where the moisture content is largely controlled by grain size and thus the electrical resistivity often varies substantially over short intervals. It appears that the inverse solution must preserve the macroscopic anisotropy of the section as well as the average resistivity. The problem is illustrated using both synthetic model studies and field examples from a recent experiment at the the Sandia-Tech Vadose Zone (STVZ) facility. The STVZ site contains dense arrays of hydrologic probes and arrays of ERT electrodes allowing quantitative comparison of ERT and hydrological measurements. The goal is to construct 3-D images of the vadose zone electrical resistivity from which estimates of the subsurface geology, initial moisture content distributions, wetting fronts, and other hydrological parameters can be estimated. Inverting data using realistic values of anisotropy produces smoother, more realistic images that correlate well with hydrological measurements.