A fast hybrid 1D inversion approach for ground conductivity data

Abstract This paper presents a fast hybrid inversion procedure for ground conductivity data. The frequency domain data from these instruments are close to obeying the low‐frequency approximation, however, not infrequently need to be treated in the QuasiStatic approximation with full induction. In this paper, we suggest a hybrid inversion approach consisting of three steps: the first one is an inversion in the low‐frequency approximation to assess the overall conductivity regime; followed by an inversion step in the Born approximation with a non‐zero half‐space reference model; the final inversion step is an accurate inversion in the QuasiStatic approximation. The first step is fully linear, while the next approach is a linearization of the first step of a full iterative inversion. The final inversion step uses full‐accuracy forward responses and approximate derivatives. This hybrid inversion approach is very fast and covers all the needs of inversion of this type of data, from an almost instantaneous online/in‐field assessment of a set of measurements that offers a basis for making in‐field decisions, to an accurate inversion that produces final inversion models.