Magnetotelluric Signals Research Articles

A Novel Frequency-Division Deep Learning Approach for Magnetotelluric Data Quality Enhancement

High signal-to-noise ratio magnetotelluric (MT) data are crucial for accurately interpreting subsurface structures. Recently, deep learning has become popular for MT denoising due to its ability to avoid parameter tuning and enable real-time processing. These methods typically fit or predict signals in noisy segments after identifying and segmenting signal and noise in the time domain. However, these methods struggle to preserve both low- and high-frequency signals effectively due to high noise levels in these segments. To address this issue, we propose a novel deep learning denoising method that separately recovers low- and high-frequency signals using distinct strategies. Low-frequency signals are fitted using an inverse autoencoder with a channel attention mechanism, effectively removing high-frequency components. High-frequency signals are then predicted using a bidirectional long short-term memory network (BiLSTM) combined with a squeeze-and-excitation (SE) mechanism, enhancing prediction by considering both global and local signal characteristics. Additionally, we introduce the multivariate state estimation technique (MSET) for real-time signal-noise identification. MSET analyzes residuals after separating low-frequency signals to identify noise. Denoising is performed only on segments with significant noise, preserving more effective signals. Finally, the fitted low-frequency dominant component and predicted high-frequency component are combined to form the denoised MT signals. This combined approach significantly improves the restoration quality of effective signals compared to existing methods. Experimental results demonstrate that our method exhibits superior denoising capabilities in both quantitative and qualitative evaluations, including apparent resistivity-phase curves and polarization direction analysis, offering enhanced performance over current deep learning methods.

Vehicle noise characteristics in magnetotelluric data and vehicle noise removal using waveform fitting

Magnetic field fluctuations due to vehicle noise were observed in magnetotelluric (MT) time-series data measured near roads. The observed vehicle noise had magnitudes ranging from tens to thousands of μA/m, whereas the observed weak natural MT signal magnitudes were approximately tens of μA/m. A small signal-to-noise ratio made it difficult to apply robust processing for removing vehicle noise. In addition, vehicle noise severely distorts the MT response in the MT deadband from 0.01 Hz to 0.3 Hz, where the MT signal is very weak, and methods to remove it are required for deep structure imaging. In this study, magnetic field fluctuations due to moving vehicles were simulated with a magnetic dipole and attempted to be removed using a waveform fitting method. A total of 378 vehicle noises were extracted from the near-road MT data and synthesized with the remote MT data without vehicle noises to investigate the effect of vehicle noise on the MT response. Removal of vehicle noise from synthesized remote MT data resulted in substantial restoration of the apparent resistivity and phase curves around the MT deadband and below 0.001 Hz. In the MT field data, the vehicle noise was simulated and removed with two moving dipoles; the magnitude of the remaining vehicle noise was reduced by approximately half compared to a single dipole, and very stable apparent resistivity and phase curves were obtained. Although electromagnetic noise distortion remains after vehicle noise removal, the waveform fitting method significantly improves the apparent resistivity and phase curve response in the 0.01–0.3 Hz frequency band.

Data augmentation and denoising of magnetotelluric signals based on CS-ResNet

Data augmentation and denoising of magnetotelluric signals based on CS-ResNet

Magnetotelluric sampling theorem

We consider a general case of a magnetotelluric (MT) study to reveal three-dimensional (3D) distribution of the electrical conductivity within the Earth based on measurements of electromagnetic (EM) fields by a two-dimensional (2D) array. Such an MT array observation can be regarded as a spatially discrete sampling of the MT responses (impedances), and each observation site can be regarded as a sampling point. This means that MT array measurements must follow the Nyquist–Shannon sampling theorem. This paper discusses how the sampling theorem is applied to MT array studies and what kind of consideration is required in the application on the basis of synthetic model calculations, with special attention to spatial resolutions. With an aid of the EM scattering theory and the sampling theorem, we can show that an observation array resolves some features of the MT impedance but does not others. We call the resolvable and unresolvable features the MT signal and noise, respectively. This study introduces the spatial Fourier transform of array MT data (impedances) which helps us investigating sampling effects of lateral heterogeneity from a different angle (in the wavenumber domain). Shallow heterogeneities cause a sharp spatial change of impedance elements near structural boundaries. High wavenumber Fourier components are required to describe such a feature, which means the site spacing must be sufficiently short to be able to resolve such features. Otherwise, a set of array MT data will suffer from aliasing, which is one of the typical causes of MT distortion (MT geologic noise). Conversely, a signal due to a deep-seated conductivity anomaly will have more reduced amplitude at higher wavenumbers, which means focused imaging of such an anomaly is generally difficult. Finally, it is suggested to properly consider the sampling theorem in an observation array design, so as to have best performance in resolving MT signals.Graphical abstract

A de-noising method of magnetotelluric signals based on the generalized S-transform

The generalized S-transform has excellent time-frequency resolution capability. It can effectively demonstrate the non-stationary characteristics of signals and the dynamic changes in energy. This provides the conditions for denoising in the time-frequency domain. The magnetotelluric (MT) survey method is often affected by various artificial noises, and these noises are very prominent in certain frequency bands. To address this problem, a method based on the generalized S transform is proposed to de-noise MT data. The generalized S transform is used to improve the time-frequency localization of the different signal components, meanwhile the electromagnetic noise can be conveniently suppressed in the time-frequency domain. Simulated signals and measured MT signals contaminated by common sinusoidal and triangular wave interference are processed. The results were compared with those of the Wavelet threshold method. Results show that the presented method is more conducive to the identification and suppression of electromagnetic noise and shows better de-noising results than the Wavelet threshold method. After de-noising, the useful information of the MT signal affected by strong interference is highlighted, which provides favorable conditions for power spectrum and impedance estimation in the future. Further, the response parameter curve calculated from the de-noised data becomes stable. The proposed method provides a new alternative for de-noising of MT signals, especially those with low signal-to-noise ratios.

Marine magnetotelluric and controlled source electromagnetics signal simulative generator

Ocean bottom electromagnetic receivers (OBEMs) are invaluable tools for observing seafloor electromagnetic signals, primarily used in magnetotelluric (MT) sounding and marine controlled source electromagnetics (MCSEM). However, the high operational costs of deploying OBEMs require thorough pre-deployment testing to enhance the data reliability and quality. Commercial signal generators used for MT and controlled source electromagnetic (CSEM) testing often fall short of the restricted terms of the frequency range, signal-to-noise ratio, storage depth, time synchronization, and amplitude dynamics. To address these limitations, we developed a specialized marine MT and CSEM signal simulative generator. The generator employs pseudorandom number code-based white noise to simulate MT signals and uses multi-frequency synthesis and amplitude modulation (AM) to generate CSEM signals. The system is developed using a low-power microcontroller unit and a field programmable gate array platform, integrating various circuits such as power, clock, encoding, control, data storage, chopper, and amplification circuits. The generator simultaneously outputs four-channel white noise and AM signals. The white noise feature provides a wide and adjustable frequency band and amplitude, while the AM feature provides a large dynamic amplitude range and multi-frequency synthesis at various carrier frequency points. Comprehensive testing verified the performance of the proposed simulative generator. The test results demonstrate that the MT simulated signals cover a frequency range from 1 E4s to 100 Hz, providing adjustable amplitudes between 10 μVpp and 10 mVpp. The CSEM simulated signals include standard square waves, third harmonic waves, and seventh harmonic waves, with amplitudes adjustable from 10 mVpp to 10 Vpp. The proposed simulative generator enhances the efficiency of OBEM functional testing and provides essential equipment support for maintaining the quality of marine data.

Natural source electromagnetic survey for geothermal application in industrial area with strong electromagnetic noise

The magnetotelluric (MT) method is a widely used geophysical technique for deep structure exploration. However, rapid industrial development has resulted in significant anthropogenic interference for practical applications of this method in recent years. To carry out MT survey in regions with strong electromagnetic (EM) noise, we develop a strategy, in which part of MT sites are set for night time period and a newly proposed automatic clustering method based on MT signal patterns is applied to process measured data from Yingde area in Guangdong. The MT sites are employed in night time to avoid human activity and measure the quiet time MT signal to make sure that we can capture portion of MT signal. Based on cosine similarity, we can use clustering algorithms to separate the signals from the noise. The application of this strategy for geothermal exploration in Yingde area in Guangdong, indicates the validity of our proposed strategy for real geothermal detection.

Research and Application of Deep Learning Selection Method for Magnetotelluric Power Spectrum Based on Rhoplus Correction

Magnetotelluric (MT) is an electromagnetic exploration method using natural field sources, which are unstable, irregular, and vulnerable to interference from electromagnetic signals. The total power spectrum of the magnetotelluric signal is superimposed by a sub-power spectrum of segmented time series. To reduce the influence of electromagnetic interference, the pre-processing of magnetotelluric data will delete the severely interfered sub-power spectrum. The selection of power spectrum often requires human-computer interaction, which is time-consuming and laborious, heavily dependent on personal experience, and has great randomness. This article innovation is put forward based on the classification of Rhoplus correction and deep learning automatic power spectrum method, first of all, the MT power spectrum impedance estimation results are Rhoplus one dimensional inversion and obtain fitting, using neural network learning fitting differential characteristics and classification, and then for the power spectrum of MT again after low-quality quantum power spectrum to eliminate overlay, Impedance estimation is performed, and the above process is repeated to find the optimal power spectrum combination. Through the comparison and verification of measured MT data, the automatic power spectrum selection method based on Rhoplus correction and deep learning classification achieve similar results to manual editing.

Garnet pyroxenites explain high electrical conductivity in the East African deep lithosphere

In Tanzania, the deep lithospheric mantle (> 70 km depth) is characterized by significantly higher electrical conductivity within the cratonic root than in the Mozambique belt. Such contrasts are typically attributed to changes in volatiles and/or melt content, with changes in mineralogy deemed insufficient to impact conductivity. To test this assumption, electrical conductivity measurements were conducted at pressure-temperature conditions relevant to the Tanzanian lithosphere (1.5 and 3 GPa; from 400 to >1500 °C) on dunite (depleted) and pyroxenite (fertile) xenoliths from Engorora, Northern Tanzania. Once garnet becomes stable in the fertile mantle rock (> 60 km, 1.7 GPa), it nucleates at grain boundaries, forming the backbone of a conductive network. At 3 GPa, such garnet-rich networks increase conductivity by a factor of 100 regardless of temperature. Numerical models demonstrate that the observed low (< 10−2 S⋅m−1) and high (> 10−1 S⋅m−1) conductivity values can be explained by low and high degrees of garnet connectivity, respectively. Such high electrical conductivities in cratonic roots can be explained by the presence of connected garnet clusters or garnet pyroxenites, suggesting mantle fertilization. This new source for electromagnetic signal generation at appropriate pressures and temperatures must be factored in where interpreting magnetotelluric signals at relevant depths in the lithosphere. Plain language summaryThe deep root of the Tanzanian craton is characterized by high electrical conductivity. In this study, we reproduce this anomaly in the laboratory. We use powders of Tanzanian xenoliths, which are natural rocks similar to what is expected in the deep Tanzanian lithosphere. Two extremes of mantle composition are tested: dunite and pyroxenite, respectively typical of a depleted mantle (already produced melts) and fertile mantle (high chemical diversity). Experiments at high pressures and temperatures representative of the deep Tanzanian lithosphere reveal that the high electrical conductivity is generated by the establishment of a connected network of garnet pyroxenites. Minerals such as garnet and pyroxenes are known to conduct electricity much better than olivine – the major component of most mantle rocks – especially when they contain limited amounts of hydrogen. This “water” is incorporated in crystals and at their boundaries and promotes high-conductivity pathways. A significant increase in conductivity occurs once the pressure is high enough for garnet to become stable. These results suggest that the composition of the deep cratonic root in Tanzania is closer to a pyroxenite than a peridotite. This difference in composition compared to the average mantle is expected as an impact of the East African superplume.

Data processing method for magnetotelluric sounding based on cepstral analysis

Magnetotelluric (MT) signals exhibit the characteristics of being weak and having a wide frequency band. The acquired field data are susceptible to various types of noise, which poses challenges in accurate identification and processing. Currently, there exist numerous MT data processing methods; however, they lack efficiency and physical meaning. To address this issue and improve the signal-to-noise ratio of the acquired data, this study proposes a MT data processing method based on cepstral analysis. By employing cepstral analysis on the MT data, the cepstrum is obtained, and an appropriate truncation position is selected for processing. The experimental results demonstrate that this method obtains smoother and more continuous apparent resistivity curves with fewer errors. Compared with other methods, the cepstral analysis method can effectively suppress different types of MT noise, and the method is simple and efficient with clear physical significance.

Magnetotelluric Power Line Noise Removal Using Temporally Varying Sinusoidal Subtraction of the Grid Utility Frequency

The magnetotelluric method relies on variations of natural electromagnetic fields, which in the vicinity of human settlements are persistently distorted by anthropogenic electromagnetic noise. A large source of noise to the magnetotelluric response is caused by the harmonic oscillations of the power network utility frequency centered on 50/60 Hz along with the associated higher harmonics. Removing this type of noise is essential for high frequency magnetotelluric measurements used for shallow surveys. There are a large number of approaches for how to treat power line noise in magnetotelluric signals, however, commonly used methods do not take into account time variations/instabilities of the utility frequency. That is not serious problem in vicinity of well balanced grid networks, but can cause issues in regions with larger utility frequency variations. Under such conditions, commonly used methods loose more of the natural signal, which is undesirable especially in case of very noisy datasets. Hence, we adopted approach for removing of power line noise with respect to time variations of the utility frequency and applied it to magnetotelluric signals to preserve more of natural signal. The method is based on modelling of the grid network harmonic oscillations by the optimum utility frequency and its integer multiples. The resulting sum of sinusoidal signals is subsequently subtracted from recorded data and only particular noise frequencies are removed from the original signal with high precision, while frequency ranges around power line harmonics are cleaned.

Magnetotelluric noise suppression via convolutional neural network

It is well known that a magnetotelluric (MT) signal with high signal-to-noise ratio is an important prerequisite for correct interpretation of subsurface structures. However, MT signals collected in the environment of strong cultural noise often are of low data quality due to noise pollution, which seriously affects the accuracy of interpretation. As can be seen from the MT time-domain waveform, the noise is highly energetic, diverse, and random. This means MT denoising methods should have strong applicability to guarantee accurate and effective separation of MT signal from noise data. Therefore, we propose a deep-learning-based data nonlinear mapping method for MT signal-to-noise separation. First, this method focuses on learning the nonlinear mapping relationship between a large amount of noise data and the corresponding noise contour by using the convolutional neural network (CNN) in advance. Then, the mapping transformation of noise data to noise contour in the measured data is realized by CNN model. Specifically, based on the features of MT noise data, we construct a large amount of training data very close to it by mathematical functions. At the same time, we also select some of the measured data to be added to the training set. This not only expands the amount and diversity of the training set but also improves the adaptability of CNN when dealing with complex data. Finally, we evaluate the denoising performance of the proposed method in terms of time-domain waveforms, apparent resistivity-phase curves, and polarization directions before and after denoising. The processing results of the simulated data and the measured data collected in Luzong area have verified the feasibility and effectiveness of the proposed method in MT data denoising.

Magnetotelluric data denoising method combining two deep-learning-based models

The magnetotelluric (MT) data collected in an ore-concentration area are extremely vulnerable to all kinds of noise pollution. However, separating real MT signals from strong noise is still a difficult problem, and the noise in MT data is quite distinct from clean data in morphological features. By performing the signal-noise identification and data prediction, we develop a deep learning method to denoise MT data containing strong noise. First, we use the convolutional neural network (CNN) to learn the feature differences between the samples of massive noise and clean data and use the learned features to realize signal-noise identification of the measured data. Second, we use the measured clean data obtained by CNN identification to train the long short-term memory (LSTM) neural network and perform the prediction denoising of the noisy data. The simulation results clearly demonstrate the following two facts: (1) the predicted data output from LSTM basically matches the time-frequency domain features of the real data and (2) our CNN method performs significantly better than the features parameter classification method in dealing with signal-noise identification. In addition, the validity of our method is verified by the processing results of the measured data.

Removal of strong noise in magnetotelluric data using grey wolf optimized wavelet threshold

AbstractTraditional magnetotelluric signal processing usually uses time–frequency transformation method. Wavelet is also a time–frequency transformation method that used to suppress the magnetotelluric noise. However, the selection of the threshold is very significant, and the unsuitable threshold will lead to excessive distortion of the reconstructed signal. Thus, we propose a method for magnetotelluric noise suppression using grey wolf optimized wavelet threshold. First, the magnetotelluric signal is decomposed by wavelet with appropriate wavelet basis and decomposition layers. The generalized cross‐validation criterion is used as the fitness function of grey wolf optimizer algorithm, which optimizes the threshold of each decomposition layer. Then, the detail coefficients of each layer and the maximum layer of the approximation coefficients are used in the optimized threshold. Next, the inverse wavelet transform is performed. Finally, the noise contour is obtained through iteratively searching for the optimal threshold, and the useful magnetotelluric signal is reconstructed. Simulation experiments and measured magnetotelluric data processing show that the large‐scale interference can effectively be suppressed, and the reconstructed magnetotelluric signal retains the more abundant low‐frequency of useful information. Compared with the remote reference method, fixed threshold method and Birge–Massart layered threshold method, the proposed method realizes the wavelet denoising with adaptive threshold selection in the magnetotelluric noise suppression. The results obtained show smoother and more continuous apparent resistivity–phase curves, which verifies the effectiveness of the optimization.

Magnetotelluric Noise Attenuation Using a Deep Residual Shrinkage Network

Magnetotelluric (MT) surveying is an essential geophysical method for mapping subsurface electrical conductivity structures. The MT signal is susceptible to cultural noise, and the intensity of noise is growing with urbanization. Cultural noise is increasingly difficult to be removed by conventional data processing methods. We propose a novel time-series editing method based on the deep residual shrinkage network (DRSN) to address this issue. Firstly, the MT data are divided into small segments to form a dataset system. Secondly, we use the dataset system to train the denoising model. Finally, the trained model is used for MT data denoising. The experiments using synthetic data and actual field data collected in Qinghai and Luzong, China, show that the DRSN can effectively remove the cultural noise and has better adaptability and efficiency than traditional MT signal processing methods.

Identification and Suppression of Magnetotelluric Noise via a Deep Residual Network

The magnetotelluric (MT) method is widely applied in petroleum, mining, and deep Earth structure exploration but suffers from cultural noise. This noise will distort apparent resistivity and phase, leading to false geological interpretation. Therefore, denoising is indispensable for MT signal processing. The sparse representation method acts as a critical role in MT denoising. However, this method depends on the sparse assumption leading to inadequate denoising results in some cases. We propose an alternative MT denoising approach, which can achieve accurate denoising without assumptions on datasets. We first design a residual network (ResNet), which has an excellent fitting ability owing to its deep architecture. In addition, the ResNet network contains skip-connection blocks to guarantee the robustness of network degradation. As for the number of training, validation, and test datasets, we use 10,000,000; 10,000; and 100 field data, respectively, and apply the gradual shrinkage learning rate to ensure the ResNet’s generalization. In the noise identification stage, we use a small-time window to scan the MT time series, after which the gramian angular field (GAF) is applied to help identify noise and divide the MT time series into noise-free and noise data. We keep the noise-free data section in the denoising stage, and the noise data section is fed into our network. In our experiments, we test the performances of different time window sizes for noise identification and suppression and record corresponding time consumption. Then, we compare our approach with sparse representation methods. Testing results show that our approach can obtain the desired denoising results. The accuracy and loss curves show that our approach can well suppress the MT noise, and our network has a good generalization. To further validate our approach’s effectiveness, we show the apparent resistivity, phase, and polarization direction of test datasets. Our approach can adjust the distortion of apparent resistivity and phase and randomize the polarization direction distribution. Although our approach requires the high quality of the training dataset, it achieves accurate MT denoising after training and can be meaningful in cases of a severe MT noisy environment.

ADMM-Based Method for Estimating Magnetotelluric Impedance in the Time Domain

Traditional magnetotelluric (MT) impedance estimations are based on Fourier theory and carried out in the frequency domain, which has a strict stationarity requirement for the analyzed signal. However, the stationarity assumption cannot be satisfied when the data possess a low signal-to-noise ratio and/or a short observation period. These shortcomings can cause significant errors in the MT impedance estimations, especially in the low-to-medium frequency bands. The alternating direction method of multipliers (ADMM) and polarization analysis of the electromagnetic signal can be attached to the time-domain MT impedance estimation to address these shortcomings. This time-domain technique calculates the impedance in the time domain without Fourier transformation, and the ADMM and polarization analysis are applied to further improve the stability and convergence of the impedance estimation. Here, we present an ADMM-based method for MT impedance estimations (ADMM-MT). Various noise are added to a noise-free MT signal to test the performance of ADMM-MT. The results show that ADMM-MT yields impedance estimates with relative recovery errors below 0.2, even when the signal-to-noise ratio of the data is 0 dB and the observation period is 60 min. We then apply ADMM-MT to field data in Inner Mongolia, China. The results indicate that the apparent resistivity curves obtained by ADMM-MT using a short time series are smooth over the 0.001–1-Hz band, which is consistent with the remote reference processing results obtained using a long time series. In contrast, the curves obtained by traditional robust estimation method are strongly biased.

De-Noising of Magnetotelluric Signals by Discrete Wavelet Transform and SVD Decomposition

Magnetotelluric (MT) sounding data can easily be damaged by various types of noise, especially in industrial areas, where the quality of measured data is poor. Most traditional de-noising methods are ineffective to the low signal-to-noise ratio of data. To solve the above problem, we propose the use of a de-noising method for the detection of noise in MT data based on discrete wavelet transform and singular value decomposition (SVD), with multiscale dispersion entropy and phase space reconstruction carried out for pretreatment. No “over processing” takes place in the proposed method. Compared with wavelet transform and SVD decomposition in synthetic tests, the proposed method removes the profile of noise more completely, including large-scale noise and impulse noise. For high levels or low levels of noise, the proposed method can increase the signal-to-noise ratio of data more obviously. Moreover, application to the field MT data can prove the performance of the proposed method. The proposed method is a feasible method for the elimination of various noise types and can improve MT data with high noise levels, obtaining a recovery in the response. It can improve abrupt points and distortion in MT response curves more effectively than the robust method can.

Bayesian magnetotelluric inversion using methylene blue structural priors for imaging shallow conductors in geothermal fields

In geothermal exploration, magnetotelluric (MT) data and inversion models are commonly used to image shallow conductors typically associated with the presence of an electrically conductive clay cap that overlies the main reservoir. However, these inversion models suffer from nonuniqueness and uncertainty, and the inclusion of useful geologic information is still limited. We have developed a Bayesian inversion method that integrates the electrical resistivity distribution from MT surveys with borehole methylene blue (MeB) data, an indicator of conductive clay content. The MeB data were used to inform structural priors for the MT Bayesian inversion that focus on inferring with uncertainty the shallow conductor boundary in geothermal fields. By incorporating borehole information, our inversion reduced nonuniqueness and then explicitly represented the irreducible uncertainty as estimated depth intervals for the conductor boundary. We used the Markov chain Monte Carlo and a 1D three-layer resistivity model to accelerate the Bayesian inversion of the MT signal beneath each station. Then, inferred conductor boundary distributions were interpolated to construct pseudo-2D/3D models of the uncertain conductor geometry. We compare our approach against deterministic MT inversion software on synthetic and field examples, and our approach has good performance in estimating the depth to the bottom of the conductor, a valuable target in geothermal reservoir exploration.

Separation of magnetotelluric signals based on refined composite multiscale dispersion entropy and orthogonal matching pursuit

Magnetotelluric (MT) data processing can increase the reliability of measured data. Traditional MT data denoising methods are usually applied to entire MT time-series, which results in the loss of useful MT signals and a decrease of imaging accuracy of electromagnetic inversion. However, targeted MT noise separation can retain part of the signal unaffected by strong noise and enhance the quality of MT responses. Thus, we propose a novel method for MT noise separation that uses the refined composite multiscale dispersion entropy (RCMDE) and the orthogonal matching pursuit (OMP) algorithm. First, the RCMDE is extracted from each segment of the MT data. Then, the RCMDEs for each segment are input to the fuzzy c-mean (FCM) clustering algorithm for automatic identification of the MT signal and noise. Next, the OMP method is utilized to remove the identified noise segments independently. Finally, the reconstructed signal consists of the denoised signal segments and the identified useful signal segments. We conducted simulation experiments and algorithm evaluations on electromagnetic transfer function (EMTF) data, simulated data and measured sites. The results indicate that the RCMDE can improve the stability of multiscale dispersion entropy (MDE) and multiscale entropy (ME) by analyzing the characteristics of the signal samples library, effectively distinguishing MT signals and noise. Compared with the existing technique of denoising entire time series, the proposed method uses the RCMDE as characteristic parameter and uses the OMP algorithm for noise separation, simplifies the multi-feature fusion, and improves the accuracy of signal-noise identification. Moreover, the denoising efficiency is accelerated, and the MT response in the low-frequency band is greatly improved.