A method for depicting deep and large faults based on low-frequency electromagnetic information recovery enhancement

Abstract

This invention provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. The method includes: Step S1: acquiring electromagnetic data and preprocessing it to obtain preprocessed electromagnetic data; Step S2: obtaining the apparent resistivity and phase in the frequency domain based on the preprocessed electromagnetic data; Step S3: establishing a shallow stratum electrical model; Step S4: obtaining a depth domain apparent resistivity profile based on the shallow stratum electrical model; Step S5: characterizing the distribution of deep and large faults based on the depth domain apparent resistivity profile. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults.

Description

Technical Field

[0001] This invention relates to the field of geophysical exploration, and in particular to a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. Background Technology

[0002] Hot dry rock resources are an emerging geothermal energy source, referring to rock masses buried at depths of 3–10 km, containing little or no fluid, with temperatures exceeding 180℃, whose thermal energy can be utilized under current technological and economic conditions. With its enormous energy reserves, wide distribution, clean and low-carbon characteristics, and stable and continuous thermal energy, it is considered a strategic alternative energy source with great development potential. The selection of target areas for hot dry rock geothermal resources involves multiple factors, including "source, pathway, reservoir, and caprock." Deep, large faults, as a crucial link in "pathway," involve depths of 10–30 km, making their characterization extremely challenging.

[0003] Electromagnetic exploration (EMB) technology is an effective means of identifying deep and large faults (10km-30km), and deep, low-resistivity characteristics are often associated with thermal channels. This paper systematically reviews the various technical aspects of EMB, clarifying that the signal-to-noise ratio (SNR) of the acquired signal and the inversion algorithm play a decisive role in the EMB inversion effect. Due to various random noises generated by natural and man-made interference sources, magnetotelluric time series signals are scattered with various types of noise, leading to reduced detection accuracy and potentially introducing false anomalies. Therefore, denoising of the time series signal is essential before processing and inversion of the acquired data. Improving the SNR and extracting useful information from the electromagnetic signal using appropriate denoising methods is of significant practical importance. In practical applications, various noise removal methods have emerged, such as filtering, least squares, remote reference, Fourier analysis, and wavelet transform. However, due to the complex square wave noise, step noise, and impulse noise contained in magnetotelluric time series signals, these traditional methods have certain limitations in their noise removal capabilities.

[0004] To address the characteristics of natural electromagnetic fields—wide bandwidth, small amplitude, high randomness, and uneven energy distribution with frequency—a denoising method based on a combination of nonlinear filters and singular spectrum decomposition (SSD) was designed. Nonlinear filters demonstrate excellent processing capability and adaptability for complex and variable nonlinear and nonstationary signals, effectively filtering out noise and interference. SSD analysis, a principal component analysis method for studying nonlinear time series data, utilizes the basic idea of ​​phase space reconstruction to extract the main features of the filtered electromagnetic signal. Combining nonlinear filters with SSD analysis further enhances the denoising effect of electromagnetic signal processing, offering significant technical advantages over traditional methods. Therefore, introducing the combination of nonlinear filters and SSD analysis into the preprocessing stage of magnetotelluric time series data reveals a novel method for noise removal in magnetotelluric time series, improving the signal-to-noise ratio of electromagnetic signals and enhancing low-frequency electromagnetic signals, thus providing accurate foundational data for further apparent resistivity conversion and inversion.

[0005] Furthermore, in the electromagnetic inversion process, conventional inversion methods often fail to effectively recover low-frequency information. Even with constrained inversion, the model primarily considers the electrical constraints at the depth of the high-resistivity basement top surface, typically at a depth of 10 km. For thermal exploration of hot dry rock, which requires predicting the electrical structure within 50 km, a model built only at a 10 km depth is difficult to fit in the low-frequency band, resulting in low resolution in deep inversion and affecting the accuracy of deep structure characterization. Therefore, based on the constraints of shallow strata electrical models, low-frequency electromagnetic signals are extracted, and low-frequency electromagnetic inversion is focused on deep layers. Multiple rounds of iterative inversion highlight the detection advantages of low-frequency signals and improve deep resolution. Summary of the Invention

[0006] In view of the above problems, the present invention is proposed to provide a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement to overcome or at least partially solve the above problems.

[0007] According to one aspect of the present invention, a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement is provided, the characterization method comprising:

[0008] Step S1: Collect electromagnetic data and perform preprocessing to obtain preprocessed electromagnetic data;

[0009] Step S2: Based on the preprocessed electromagnetic data, obtain the apparent resistivity and phase in the frequency domain;

[0010] Step S3: Establish a shallow formation electrical model;

[0011] Step S4: Obtain the depth domain apparent resistivity profile based on the shallow formation electrical model;

[0012] Step S5: Based on the depth domain apparent resistivity profile, characterize the distribution of deep fractures.

[0013] Optionally, step S1: acquiring electromagnetic data and preprocessing it to obtain preprocessed electromagnetic data specifically includes:

[0014] Step S101: Suppress electromagnetic time series noise based on a combination of nonlinear filters and singular spectrum decomposition to obtain electromagnetic data with high signal-to-noise ratio.

[0015] Optionally, step S101: Suppressing electromagnetic time series noise and obtaining high signal-to-noise ratio electromagnetic data based on a combination of nonlinear filters and singular spectral decomposition specifically includes:

[0016] Step S1011: Use a nonlinear filter to denoise the acquired one-dimensional electromagnetic time series signal;

[0017] Step S1012: Construct a trajectory matrix from the filtered electromagnetic time series signal and decompose it to construct a covariance matrix. Calculate the eigenvalues ​​and eigenvectors of the covariance matrix to obtain the singular values ​​of the matrix.

[0018] Step S1013: Perform singular vector grouping based on the singular values ​​of the matrix to separate the electromagnetic time series signal and the noise signal;

[0019] Step S1014: Singular vector diagonal averaging, converting the grouped matrix into an electromagnetic time series signal of a certain time length, and reconstructing the denoised electromagnetic time series signal.

[0020] Optionally, step S1012: constructing a trajectory matrix and decomposing the filtered electromagnetic time series signal into a covariance matrix, calculating the eigenvalues ​​and eigenvectors of the covariance matrix, and obtaining the singular values ​​of the matrix specifically includes:

[0021] When constructing a trajectory matrix from the denoised one-dimensional magnetotelluric time series signal, a window length L is defined, and the trajectory matrix X is rearranged. The trajectory matrix X is of order L×K, where K=N-L+1, (1<L≤N / 2). Then, the trajectory matrix X of the signal vector X is defined as...

[0022] Where X i =[x i x i+1 ,.....x i+m-1 The element x at position (i,j) of the trajectory matrix X. ij =x i+j-1 All elements on the anti-diagonal are equal;

[0023] The method for trajectory matrix decomposition is to construct the covariance matrix C = XX.T Calculate the eigenvalues ​​and eigenvectors U of the covariance matrix, and follow the formula (λ1≥λ2≥…≥λ). L ), (U1,U2,…,U L )arrangement;

[0024] In the singular vector grouping process, L components are divided into M disjoint sets I1, I2, ..., I M , where ∑|I m |=L,M∈[1,M], let I=[i1,i2,···,i p [A group is defined as matrix X] I Defined as X I =X i1 +X i2 +···+X ip Then the trajectory matrix can be represented as X = X I1 +X I2 +···+X IM ;

[0025] Each matrix X I The contribution of the trajectory matrix is ​​related to the eigenvalues, and the contribution degree ηI is defined as...

[0026] Optionally, step S1013: grouping singular vectors according to the singular values ​​of the matrix to separate the electromagnetic time series signal and the noise signal specifically includes:

[0027] Diagonal averaging of singular vectors, for each matrix Y∈R LxK The purpose of calculating the average of its diagonals is to calculate its average contribution rate.

[0028] Define the projection of Y as

[0029]

[0030] Optionally, step S2: obtaining the frequency domain apparent resistivity and phase based on the preprocessed electromagnetic data specifically includes:

[0031] Step S102: Perform time-frequency conversion and power spectrum estimation on the preprocessed electromagnetic data to obtain the apparent resistivity and phase in the frequency domain.

[0032] Optionally, step S3: establishing a shallow formation electrical model specifically includes: step S103: establishing a shallow formation electrical model and providing an inversion constraint model.

[0033] Optionally, step S4: obtaining the depth-domain apparent resistivity profile based on the shallow formation electrical model specifically includes:

[0034] Step S104: Obtain the depth domain apparent resistivity profile by low-frequency electromagnetic inversion based on the constraints of the shallow stratum electrical model.

[0035] Optionally, step S104: obtaining the depth-domain apparent resistivity profile based on the shallow formation electrical model specifically includes:

[0036] Step S1041: Conduct electromagnetic forward modeling based on shallow electrical model to clarify the frequency domain electromagnetic response characteristics of the overlying strata.

[0037] Step S1042: Extract low-frequency signals based on the forward modeling results, establish an initial electromagnetic inversion model with shallow electrical model constraints and deep grid uniform partitioning, carry out low-frequency electromagnetic inversion for deep layers, iterate multiple times until the fitting difference is minimized, and obtain the apparent resistivity profile in the depth domain.

[0038] Optionally, step S5: characterizing the distribution of deep fractures based on the depth-domain apparent resistivity profile specifically includes:

[0039] Step S105: Depth-domain apparent resistivity profile analysis to delineate the low-resistivity distribution range and characterize the deep fracture distribution.

[0040] This invention provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. The method includes: Step S1: acquiring electromagnetic data and preprocessing it to obtain preprocessed electromagnetic data; Step S2: obtaining the apparent resistivity and phase in the frequency domain based on the preprocessed electromagnetic data; Step S3: establishing a shallow stratum electrical model; Step S4: obtaining a depth domain apparent resistivity profile based on the shallow stratum electrical model; Step S5: characterizing the distribution of deep and large faults based on the depth domain apparent resistivity profile. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults.

[0041] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1A flowchart of a deep fracture characterization method based on low-frequency electromagnetic information recovery enhancement provided in an embodiment of the present invention;

[0044] Figure 2 This is a diagram showing the electromagnetic signal denoising effect based on a combination of nonlinear filters and singular spectral decomposition provided by the present invention.

[0045] Figure 3 This is a diagram of the formation electrical properties provided by the present invention;

[0046] Figure 4 This invention provides frequency-domain electromagnetic apparent resistivity curves based on forward modeling of formation electrical properties.

[0047] Figure 5 This is a deep fracture characterization diagram based on electromagnetic inversion provided by the present invention. Detailed Implementation

[0048] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0049] The terms "comprising" and "having," and any variations thereof, in the specification, embodiments, claims, and drawings of this invention are intended to cover non-exclusive inclusion, such as including a series of steps or units.

[0050] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0051] Example 1

[0052] like Figure 1 As shown, Example 1 provides a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement, which specifically includes the following steps:

[0053] Step S101: Suppress electromagnetic time series noise based on a combination of nonlinear filters and singular spectrum decomposition to obtain electromagnetic data with high signal-to-noise ratio;

[0054] Step S102: Time-frequency conversion and power spectrum estimation to obtain the apparent resistivity and phase in the frequency domain;

[0055] Step S103: Establish the electrical properties model of shallow strata to provide an inversion constraint model;

[0056] Step S104: Low-frequency electromagnetic inversion based on formation electrical model constraints to obtain the depth domain apparent resistivity profile;

[0057] Step S105: Depth-domain apparent resistivity profile analysis to characterize the extent of deep fractures.

[0058] Example 1 provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults. It can be widely applied in the field of geothermal energy exploration.

[0059] Example 2

[0060] Example 2 provides a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement, which specifically includes the following steps:

[0061] Step S101: Suppress electromagnetic time series noise using a combination of nonlinear filters and singular spectral decomposition to obtain electromagnetic data with a high signal-to-noise ratio. Figure 2 Specifically, this includes:

[0062] Step S1011: Denoise the acquired one-dimensional electromagnetic time series signal using a nonlinear filter;

[0063] Step S1012: Construct a trajectory matrix from the filtered one-dimensional electromagnetic time series signal and decompose it, construct a covariance matrix, calculate the eigenvalues ​​and eigenvectors of the covariance matrix, and obtain the singular values ​​of the matrix.

[0064] Step S1013: Singular vector grouping, separating the electromagnetic time series signal from the noise signal;

[0065] Step S1014: Singular vector diagonal averaging, converting the grouped matrix into an electromagnetic time series signal of a certain time length, and reconstructing the denoised electromagnetic time series signal.

[0066] Step S102: Time-frequency conversion and power spectrum estimation to obtain the apparent resistivity and phase in the frequency domain;

[0067] Step S103: Establish the electrical properties model of shallow strata to provide an inversion constraint model;

[0068] Step S104: Low-frequency electromagnetic inversion based on formation electrical model constraints to obtain the depth domain apparent resistivity profile;

[0069] Step S105: Depth-domain apparent resistivity profile analysis to characterize the extent of deep fractures.

[0070] Example 2 provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults. It can be widely applied in the field of geothermal energy exploration.

[0071] Example 3

[0072] Example 3 provides a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement, which specifically includes the following steps:

[0073] Step S101: Suppress electromagnetic time series noise based on a combination of nonlinear filters and singular spectrum decomposition to obtain electromagnetic data with high signal-to-noise ratio;

[0074] Step S102: Time-frequency conversion and power spectrum estimation to obtain the apparent resistivity and phase in the frequency domain;

[0075] Step S103: Establishment of shallow strata electrical model ( Figure 3 ), providing an inversion constraint model;

[0076] Step S104: Low-frequency electromagnetic inversion based on formation electrical model constraints to obtain the depth domain apparent resistivity profile;

[0077] Specifically, it includes the following sub-steps:

[0078] Step S1041: Conduct electromagnetic forward modeling based on shallow electrical model to clarify the frequency domain electromagnetic response characteristics of the overlying strata.

[0079] Step S1042: Extract low-frequency signals based on forward modeling results, establish an initial electromagnetic inversion model with shallow electrical model constraints and deep grid uniform partitioning, carry out low-frequency electromagnetic inversion for deep layers, iterate multiple times until the fitting difference is minimized, and obtain the apparent resistivity profile in the depth domain.

[0080] Step S105: Depth-domain apparent resistivity profile analysis to characterize the extent of deep fractures.

[0081] Example 3 provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults. It can be widely applied in the field of geothermal energy exploration.

[0082] Example 4

[0083] Example 4 provides a method for characterizing deep and large fractures based on low-frequency electromagnetic information recovery enhancement, which specifically includes the following steps:

[0084] Step S101: Suppress electromagnetic time series noise using a combination of nonlinear filters and singular spectral decomposition to obtain electromagnetic data with a high signal-to-noise ratio; specifically including:

[0085] Step S1011: Denoise the acquired one-dimensional electromagnetic time series signal using a nonlinear filter;

[0086] Step S1012: Construct a trajectory matrix from the filtered one-dimensional electromagnetic time series signal and decompose it, construct a covariance matrix, calculate the eigenvalues ​​and eigenvectors of the covariance matrix, and obtain the singular values ​​of the matrix.

[0087] Step S1013: Singular vector grouping, separating the electromagnetic time series signal from the noise signal;

[0088] Step S1014: Singular vector diagonal averaging, converting the grouped matrix into an electromagnetic time series signal of a certain time length, and reconstructing the denoised electromagnetic time series signal.

[0089] Step S102: Time-frequency conversion and power spectrum estimation to obtain frequency domain apparent resistivity and phase; the frequency domain electromagnetic apparent resistivity curve based on the forward model of the formation electrical properties is shown below. Figure 4 As shown;

[0090] Step S103: Establish the electrical properties model of shallow strata to provide an inversion constraint model;

[0091] Step S104: Low-frequency electromagnetic inversion based on formation electrical model constraints to obtain the depth domain apparent resistivity profile;

[0092] Specifically, it includes the following sub-steps:

[0093] Step S1041: Conduct electromagnetic forward modeling based on shallow electrical model to clarify the frequency domain electromagnetic response characteristics of the overlying strata. The forward modeling results show that...

[0094] Step S1042: Extract low-frequency signals based on forward modeling results, establish an initial electromagnetic inversion model with shallow electrical model constraints and deep grid uniform partitioning, carry out low-frequency electromagnetic inversion for deep layers, iterate multiple times until the fitting difference is minimized, and obtain the apparent resistivity profile in the depth domain.

[0095] Step S105: Depth-domain apparent resistivity profile analysis to characterize the extent of deep fractures ( Figure 5 ).

[0096] Example 4 provides a method for characterizing deep and large faults based on the recovery and enhancement of low-frequency electromagnetic information. This method enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults. It is widely used in the field of geothermal energy exploration.

[0097] Beneficial Effects: The deep and large fault characterization method based on low-frequency electromagnetic signal enhancement provided by this invention enhances low-frequency electromagnetic signals, improves the inversion accuracy of electromagnetic data, effectively recovers low-frequency geological information, and can effectively characterize deep and large faults. It can be widely applied in the field of geothermal energy exploration. Using the method provided by this invention, deep and large fault characterization based on electromagnetic inversion has been carried out in multiple areas, including the Yihezhuang area, Chenjiazhuang area, and Wangpanzhen area of ​​the Jiyang Depression, providing geophysical evidence for geothermal resource exploration.

[0098] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for depicting deep and large faults based on low-frequency electromagnetic information recovery enhancement, characterized in that, The characterization method includes: Step S1: Collect electromagnetic data and perform preprocessing to obtain preprocessed electromagnetic data, specifically including: Step S1011: Use a nonlinear filter to denoise the acquired one-dimensional electromagnetic time series signal; Step S1012: Construct a trajectory matrix from the filtered electromagnetic time series signal and decompose it to construct a covariance matrix. Calculate the eigenvalues ​​and eigenvectors of the covariance matrix to obtain the singular values ​​of the matrix. Step S1013: Perform singular vector grouping based on the singular values ​​of the matrix to separate the electromagnetic time series signal and the noise signal; Step S1014: Singular vector diagonal averaging, converting the grouped matrix into an electromagnetic time series signal of a certain time length, and reconstructing the denoised electromagnetic time series signal; Step S2: Based on the preprocessed electromagnetic data, obtain the apparent resistivity and phase in the frequency domain; Step S3: Establish a shallow formation electrical model; Step S4: Based on the shallow formation electrical model, obtain the depth-domain apparent resistivity profile, specifically including: Step S1041: Conduct electromagnetic forward modeling based on shallow electrical model to clarify the frequency domain electromagnetic response characteristics of the overlying strata. Step S1042: Extract low-frequency signals based on forward modeling results, establish an initial electromagnetic inversion model with shallow electrical model constraints and deep grid uniform partitioning, carry out low-frequency electromagnetic inversion for deep layers, iterate multiple times until the fitting difference is minimized, and obtain the apparent resistivity profile in the depth domain. Step S5: Based on the depth domain apparent resistivity profile, characterize the distribution of deep fractures.

2. The method according to claim 1, wherein, Step S2: Obtaining the apparent resistivity and phase in the frequency domain based on the preprocessed electromagnetic data specifically includes: Step S102: Perform time-frequency conversion and power spectrum estimation on the preprocessed electromagnetic data to obtain the apparent resistivity and phase in the frequency domain.

3. The method according to claim 1, wherein, Step S3: Establishing a shallow formation electrical model specifically includes: Step S103: Establish a shallow stratum electrical property model and provide an inversion constraint model.

4. The method according to claim 1, wherein, Step S5: Characterizing the distribution of deep fractures based on the depth domain apparent resistivity profile specifically includes: Step S105: Depth-domain apparent resistivity profile analysis to delineate the low-resistivity distribution range and characterize the deep fracture distribution.

Images