Inversão TDEM usando algorítmo de otimização local Levenberg–Marquardt

Abstract

The Time-Domain Electromagnetic (TDEM) method is a geophysical technique used to investigate subsurface physical properties, such as the electrical distribution and conductivity of materials. TDEM sounding data inversion is a technique aimed at estimating a subsurface model of these physical properties based on field data obtained using the method. In this study, the TDEM data inversion process is carried out by estimating the model parameters using the Levenberg–Marquardt algorithm, with the objective of minimizing the difference between the field data and the data calculated by the forward model. A 1D survey was used, in which physical properties vary only with depth, estimating the thickness and resistivity of one-dimensional layers. By implementing the algorithm on a set of synthetic data with noise, it was possible to obtain an approximate solution to the TDEM sounding data inversion problem