Improvement of MT data processing using stationary and coherence tests

Abstract

The great majority of data processing methods for magnetotelluric measurements are based on an average Fourier spectral analysis to estimate either the transfer function or the coherence function. This assumes that the time‐series data are stationary. The aim of this paper is to present a preprocessing method able to extract the stationary and coherent part of the original signals. The practitioner can then apply the magnetotelluric method of his choice to these new data. This preliminary data sorting is done in four steps: (i) slow drifts are eliminated using a high‐pass filter whose cut‐off frequency is determined by an iterative procedure. Based on run tests, this procedure is also able to remove segments with non‐independent samples in the time series; (ii) non‐stationary segments are eliminated after band‐pass filtering; (iii) non‐coherent segments are eliminated before spectral analysis; (iv) the impedance tensor value is then retained, at a given frequency, only if the signals are coherent. This preprocessing method was tested on the simplest, but still used, magnetotelluric method which uses only two field components, and it was found that the average resistivity standard deviation decreased significantly from 14.6 Ωm without sorting to 8.6 Ωm after sorting.

Using empirical mode decomposition to process marine magnetotelluric data

“…In practice, noisy measurements make it necessary to utilize a number of sophisticated statistical methods to derive robust and reliable impedance estimates (e.g. Chave & Thomson 1989; Sutarno & Vozoff 1991; Chant & Hastie 1992; Spagnolini 1994; Egbert 1997; Lamarque 1999). We will discuss these issues in relation to our proposed method below.…”

Synthesis of natural electric and magnetic Time‐series using Inter‐station transfer functions and time‐series from a Neighboring site (STIN): Applications for processing MT data

“…Although there are many sources of noise with differing impacts on MT data, it is easy to visually recognize cultural noise in the time domain by picking out features such as spikes, steps, and harmonic noise. Several approaches have been developed to differentiate between noise and MT signals, measuring different properties such as coherence [ Jones , ; Lamarque , ], wavelet coefficient amplitude [ Trad and Travassos , ], variance [ Kappler , ], and polarization [ Escalas et al ., ]. Weckmann et al .…”

“…Although there are many sources of noise with differing impacts on MT data, it is easy to visually recognize cultural noise in the time domain by picking out features such as spikes, steps, and harmonic noise. Several approaches have been developed to differentiate between noise and MT signals, measuring different properties such as coherence [Jones, 1981;Lamarque, 1999], wavelet coefficient amplitude [Trad and Travassos, 2000], variance [Kappler, 2012], and polarization [Escalas et al, 2013]. Weckmann et al [2005] suggested an approach for noise removal based on a combination of frequency domain editing with subsequent single-site robust processing, analyzing several parameters such as spectral power density, coherence, and distribution of response functions.…”

Generation of synthetic wide-band electromagnetic time series

“…Computer simulation of electromagnetic field components has frequently been used for testing estimation techniques of the magnetotelluric impedance tensor (Goubau et al, 1978;Mc-Mechan and Barrodale, 1985;Yee et al, 1988;Larsen et al, 1996) and for assessing preprocessing methods of extraction of the stationary and coherent part of signals corrupted by noise (San Filipo and Hohmann, 1983;Lamarque, 1999). The testing and assessing tasks are generally accomplished in the frequency domain through the power spectra of the measured signals.…”

“…Larsen et al (1996) used measured magnetic time series and an estimated 1D transfer function, modified by a distortion function, to generate the electric time series in a test area. Simple pseudo-random numbers were used to represent the noise-free random components of the time series (Lamarque, 1999) or the real and imaginary parts of the incident magnetic field components (Goubau et al, 1978). The latter authors calculated the electric field spectra using a bi-dimensional impedance tensor made by simple complex relationships.…”