SUMMARY
While traditional geoelectric array configurations, such as the Wenner–Schlumberger or the dipole–dipole, can provide very good images of 1-D or robust 2-D structures, they are not sufficiently sensitive to those inhomogeneities that have a small effect on the surface electrical potential distribution. The detection and description of such inhomogeneities become possible by applying quasi-null arrays, which provide very small (close to zero) signals above a homogeneous half-space. The imaging properties of the members of an array series containing such arrays, the so-called γ11n arrays (n = 1–7), are demonstrated and compared to those of the most popular traditional arrays. Although the field applicability of the quasi-null arrays has been heavily questioned, it was demonstrated by our quasi-field analogue modelling experiments. The quasi-field tests also validated all of the numerical modelling results as follows: (1) many or all of the γ11n arrays were able to detect prisms and vertical sheets located at depths larger than those detectable by traditional geoelectric arrays, including the optimized Stummer configuration; (2) the horizontal resolution of the γ11n arrays proved to be better than the horizontal resolution of traditional arrays; (3) with n increasing, the γ11n arrays proved to be less sensitive to 1-D, but more sensitive to 2-D bodies. In case of high n values, the γ11n arrays may even be entirely insensitive to any 1-D structure. On the basis of the quasi-field experiments, γ11n arrays are expected to be very efficient to indicate bodies, or variations in time that only have a small impact on the surface electrical potential distribution (e.g. caves, mines, tunnels, tubes, cables, fractures, dykes), or small changes in the subsurface conditions (monitoring of dams or waste deposits). Data acquisition by both a traditional and a γ11n array, individual inversion of their data, and a joint interpretation of the results are recommended to obtain both a robust image and fine details of the subsurface.
