Abstract
Resolution of a thin sandwiched layer employing DC resistivity or electromagnetic (EM) sounding methods is a tricky problem and often eludes adequate detectability. The inversion of resistivity data, from various electrode arrays, in general yields the product resistivity x thickness or conductivity x thickness depending on whether the target layer is resistive or conductive. On the other hand EM data, obtained using TE excitation, can resolve a conductive layer to some extent but fails miserably when the target layer is resistive. A joint-inversion of the two data sets (DC resistivity and EM), however, has been found to enhance the resolution of such thin layers to a great extent. In the present study we have tried to assess the performance of resistivity sounding data from various electrode arrays in resolving a thin sandwiched (conductive or resistive) layer, by using joint-inversion with multi-frequency EM data obtained from a large rectangular loop system. The study reveals that a very thin layer or layers with very high contrast and buried at great depths can not be resolved very accurately even by employing the joint-inversion. It is also found that, in general, the joint-inversion of EM data with that from dipole arrays yields better resolution with faster rate of convergence. The study is extended to simulate the field situation by introducing random noise to theoretical response curves. Deterioration of inverted model parameters is observed as the noise level is increased. Dipole arrays are found to provide better estimation of layer parameters in comparison to other arrays.