Tem signal transformations to frequency domain for fast data inversion

Abstract

The paper presents a new data inversion technique for the transient electromagnetic method (TEM) by converting the measured signals into the frequency domain. The inversion involves a search for such earth model parameters that there is a consistency between the synthetic and field data. We use an optimization method, where through numerical simulation at each iteration the synthetic data are determined in accordance with the changing model parameters. Numerical simulation of TEM signals is a computationally expensive procedure, since the time-domain signal is usually calculated via the inverse Fourier transform of the frequency signal. Consequently, compared to the frequency signal, the time needed to calculate the time signal increases hundreds of times. It is proposed to transform the measured signals into the frequency domain and perform inversion therein, which significantly reduces the time expenditures. Transition into the frequency domain by the Fourier transform includes the extraction of the primary field from the signal, calculated by means of a special algorithm. This fact makes it possible to employ for the transformation a relatively small time interval actually used in TEM instead of an infinite one.