3-DIMENSIONAL INTERPRETATION OF MULTIPLE-SOURCE BIPOLE-DIPOLE RESISTIVITY DATA USING THE APPARENT RESISTIVITY TENSOR

The bipole-dipole resistivity technique, which uses a single current source (bipole) to map variations in (apparent) resistivity has been much criticized in the past. A series of 3D models are used to show that the use of two distinct current bipoles in the same location but with different orientations, combined with analysis in the form of a previously defined tensor apparent resistivity, can greatly improve many aspects of bipole-dipole mapping. The model study shows that, for measurement stations more than a few bipole lengths from the current source, the apparent resistivity tensor behaves, to a close approximation, as though the current bipoles are idealized dipoles, and hence is independent of the orientation of the individual current sources used. Any pair of current bipoles (in the same location but with different orientations) can therefore be used to determine the tensor resistivity properties. The invariants of the apparent resistivity tensor have considerable advantages over the normal scalar apparent resistivities. Modelling shows that although the electric field vector corresponding to a single current bipole can be highly perturbed by a local inhomogeneity for some considerable distance beyond the inhomogeneity itself, the tensor invariants are virtually unperturbed beyond the extent of the inhomogeneity. Hence false anomalies, which are a characteristic of apparent resistivity measured using only single current bipole models, are almost completely eliminated by the use of tensor invariants. Of the possible tensor invariants, the invariant given by the square root of the determinant gives the best representation of a buried 3D body. Resistivity anomalies are localized, and occur only over the causative body. Even with complex models involving several buried bodies, the tensor invariants clearly delineate the extent of each body. Outside the bounds of perturbing bodies, the tensor data can be analysed by conventional techniques, for example, to determine layered structure.