Time-frequency analysis of magnetotelluric data

Chant, Ian J.; Hastie, L. M.

doi:10.1111/j.1365-246x.1992.tb00586.x

Cited by 19 publications

(11 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…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.…”

Section: Methodsmentioning

confidence: 99%

Using empirical mode decomposition to process marine magnetotelluric data

Chen¹,

Heincke²,

Jegen³

et al. 2012

Geophysical Journal International

View full text Add to dashboard Cite

SUMMARY A major step in processing magnetotelluric (MT) data is the calculation of an impedance tensor as function of frequency from recorded time‐varying electromagnetic fields. Common signal processing techniques such as Fourier transform based procedures assume that the signals are stationary over the record length, which is not necessarily the case in MT, due to the possibility of sudden spatial and temporal variations in the naturally occurring source fields. In addition, noise in the recorded electric and magnetic field data may also be non‐stationary. Many modern MT processing techniques can handle such non‐stationarities through strategies such as windowing of the time‐series. However, it is not completely clear how extreme non‐stationarity may affect the resulting impedances. As a possible alternative, we examine a heuristic method called empirical mode decomposition (EMD) that is developed to handle arbitrary non‐stationary time‐series. EMD is a dynamic time series analysis method, in which complicated data sets can be decomposed into a finite number of simple intrinsic mode functions. In this paper, we use the EMD method on real and synthetic MT data. To determine impedance tensor estimates we first calculate instantaneous frequencies and spectra from the intrinsic mode functions and apply the impedance formula proposed by Berdichevsky to the instantaneous spectra. We first conduct synthetic tests where we compare the results from our EMD method to analytically determined apparent resistivities and phases. Next, we compare our strategy to a simple Fourier derived impedance formula and the frequently used robust processing technique bounded‐influence remote reference processing (BIRRP) for different levels of stochastic noise. All results show that apparent resistivities and phases which are calculated from EMD derived impedance tensors are generally more stable than those determined from simple Fourier analysis and only slightly worse than those from the robust processing. These results show that EMD has the potential to handle noisy data. Finally, as a test on real data, we apply our processing scheme to data measured from the Costa Rica subduction zone, and obtain similar impedance estimates as the BIRRP processing method.

show abstract

Section: Methodsmentioning

confidence: 99%

Using empirical mode decomposition to process marine magnetotelluric data

Chen¹,

Heincke²,

Jegen³

et al. 2012

Geophysical Journal International

View full text Add to dashboard Cite

show abstract

“…Conventional Fourier methods can be applied to short sections of data (Egbert & Booker 1986), and high‐resolution spectral methods (maximum entropy) can be used to define their frequency content (Tzanis & Beamish 1989). More recently, Chant & Hastie (1992) have computed evolutive time–frequency spectra and impedances, while Zhang & Paulson (1997) have analysed electromagnetic time series using wavelet transforms. The results given by Chant & Hastie (1992) suggest that methods based on Fourier transforms of short data sections are less effective than the evolutive time–frequency approach.…”

Section: Signal and Noisementioning

confidence: 99%

“…More recently, Chant & Hastie (1992) have computed evolutive time–frequency spectra and impedances, while Zhang & Paulson (1997) have analysed electromagnetic time series using wavelet transforms. The results given by Chant & Hastie (1992) suggest that methods based on Fourier transforms of short data sections are less effective than the evolutive time–frequency approach. In this paper, I have used complex demodulation (Banks 1975) to make time‐local estimates of impedances.…”

Section: Signal and Noisementioning

confidence: 99%

The effects of non-stationary noise on electromagnetic response estimates

Banks

1998

Geophys. J. Int.

View full text Add to dashboard Cite

The noise in natural electromagnetic time series is typically non‐stationary. Sections of data with high magnetic noise levels bias impedances and generate unreliable error estimates. Sections containing noise that is coherent between electric and magnetic channels also produce inappropriate impedances and errors. The answer is to compute response values for data sections which are as short as is feasible, i.e. which are compatible both with the chosen bandwidth and with the need to over‐determine the least‐squares estimation of the impedance and coherence. Only those values that are reliable are selected, and the best single measure of the reliability of Earth impedance estimates is their temporal invariance, which is tested by the coherence between the measured and predicted electric fields. Complex demodulation is the method used here to explore the temporal structure of electromagnetic fields in the period range 20–6000 s. For periods above 300 s, noisy sections are readily identified in time series of impedance values. The corresponding estimates deviate strongly from the normal value, are biased towards low impedance values, and are associated with low coherences. Plots of the impedance against coherence are particularly valuable diagnostic aids. For periods below 300 s, impedance bias increases systematically as the coherence falls, identifying input channel noise as the cause. By selecting sections with high coherence (equivalent to the impedance being invariant over the section) unbiased impedances and realistic errors can be determined. The scatter in impedance values among high‐coherence sections is due to noise that is coherent between input and output channels, implying the presence of two or more systems for which a consistent response can be defined. Where the Earth and noise responses are significantly different, it may be possible to improve estimates of the former by rejecting sections that do not generate satisfactory values for all the response elements.

show abstract

“…Such sources of contamination can be difficult to remove from data since they may be highly coherent. Such transient contaminating signals within the time-series data can not be dealt with adequately by stationary analysis methods such as the Fourier transform (Chant & Hastie 1992;Chant 1992).…”

Section: Introductionmentioning

confidence: 99%

“…Recently a method of robust non-stationary MT analysis has been suggested which uses the Cone kernel TFD to isolate transient source effects using time-frequency analysis techniques (Chant & Hastie 1992;Chant 1992;Zhao, Atlas & Marks 1990). In this paper we compare the apparent resistivity profiles produced by this TFD method of MT analysis with results using the stationary cross-frequency analysis method (Dekker 1983) on data from an MT deep sounding in south-east Queensland,…”

Section: Introductionmentioning

confidence: 99%

A comparison of the stability of stationary and non-stationary magnetotelluric analysis methods

Chant

Hastie

1993

Geophysical Journal International

Self Cite

View full text Add to dashboard Cite

S U M M A R Y A Magnetotelluric (MT) sounding was carried out at a site in south-east Queensland, in the Clarence-Moreton Basin. The synoptic recordings were taken over a period of four months at sampling frequencies from 500 Hz to 5 x Hz. The resulting data was analysed by the stationary cross-frequency and the Cone kernel time-frequency distribution (TFD) methods of MT analysis. The results were compared as apparent resistivities on a daily basis for frequencies above 1 Hz, as well as over all the available data. The TFD MT apparent-resistivity results were more stable and less noisy on an daily basis than the cross-frequency results. Similarly the TFD analysis gave less noisy results than the cross-frequency analysis when all available data was processed. Application of these new non-stationary analysis techniques to MT processing should decrease the bias error problem of the MT methods and so increase reliability and repeatability of MT soundings.

show abstract

Time-frequency analysis of magnetotelluric data

Cited by 19 publications

References 20 publications

Using empirical mode decomposition to process marine magnetotelluric data

Using empirical mode decomposition to process marine magnetotelluric data

The effects of non-stationary noise on electromagnetic response estimates

A comparison of the stability of stationary and non-stationary magnetotelluric analysis methods

Contact Info

Product

Resources

About