A robust two-step inversion of complex magnetotelluric apparent resistivity data

Though two-dimensional inversion is now a standard procedure, interpretation of magnetotelluric (MT) data under the assumption of isotropic and one-dimensional structures is a valuable procedure for a first step interpretation of exploratory and solid earth geophysics investigation data. Because its interpretation requires an efficient inverse modelling, we propose and evaluate an inversion procedure, which consists of two steps. Both steps employ jointly the modulus and the phase of the apparent resistivity function. The first one consists of the use of the asymptotic Bostick-Niblett approach. The second employs the result of the inversion obtained in the first step as a starting model to initialize the linearized inversion performed using a multiple re-weighted least-squares approach. We applied the analysis both to synthetic data and to field data from the Parana Basin in Brazil. The results show that the inversion procedure presents a faster convergence without loss of accuracy, increases the resolving power of the MT technique, and may improve its capability to delineate conductors up to a depth of one hundred kilometers. Therefore a reasonable interpretation of the data employing one-dimensional model can be achieved even in the presence of relatively noisy data, and under conditions that slightly violate the premise of lateral homogeneity.