Method and system for suppressing low-frequency reverberation in shallow sea seismic data bubble envelope

Abstract

The invention provides a method and a system for suppressing low-frequency reverberation in a shallow sea seismic data bubble envelope. The suppressing method comprises the steps of S1, acquiring a shallow water time domain seismic signal; s2, converting the time domain seismic signal into a time lag-logarithm domain; s3, designing a time delay-logarithm domain filter; s4, suppressing low-frequency reverberation in the bubble envelope by time delay-logarithmic domain filtering; and S5, returning to a time domain to obtain seismic data after noise suppression. And bubbles and low-frequency reverberation in bubble envelope are effectively suppressed.

Description

Technical Field

[0001] The present invention relates to the field of shallow sea seismic exploration, and in particular to a method and system for suppressing low-frequency reverberation in a bubble envelope of shallow sea seismic data. Background Art

[0002] In shallow-water seismic exploration, bubble effects and reverberation are two common types of seismic noise. Airguns are the most commonly used seismic source in marine seismic exploration. Parkes et al. have well described the physical process of bubble oscillation after excitation from an airgun source. Upon excitation, the airgun releases high-pressure gas into the water, where bubbles form and undergo periodic damped expansion and contraction, generating source wavelets that propagate through the water layer and into the subsurface, generating seismic waves. This bubble effect is often accompanied by the bubble effect in marine seismic exploration, manifesting in seismic data as low-frequency events below the seismic reflection wave, which significantly reduces the resolution and signal-to-noise ratio of seismic data. Furthermore, in shallow-water exploration, when the seafloor is relatively flat and has a relatively stable reflection coefficient, energy entering the water layer can be reflected multiple times between the water surface and the seafloor, causing resonance in the water layer and generating reverberation or reverberation. Reverberation is a strong interfering wave in marine seismic exploration, similar to a sine oscillation, which can contaminate seismic data from shallow to deep depths and reduce the signal-to-noise ratio. Therefore, suppressing bubble effects and reverberation noise has become a key task in shallow-water seismic exploration and seismic data processing. In the practice of suppressing these two types of noise, deterministic deconvolution based on far-field wavelets, predictive deconvolution and some multiple wave prediction techniques are often used tools.

[0003] In shallow-water seismic exploration, the interaction between bubbles and reverberation can affect the amplitude, frequency, and attenuation of noise. In seismic data from a certain area, the combined effects of reverberation and bubbles result in the emergence of a very high-energy, low-frequency reverberation, whose amplitude variations exhibit a distinct bubble envelope trend. This strong noise obscures the effective seismic signal, not only reducing the signal-to-noise ratio of the seismic data, creating a strong notch effect and lowering the resolution of the seismic data, but also affecting subsequent ghost wave suppression. Driven by the demand for broadband data, it is widely recognized that the low-frequency components of seismic signals are crucial for optimizing full waveform inversion (FWI) and improving deep-level target imaging. Conventional predictive deconvolution and deterministic deconvolution using far-field wavelets are not ideal for suppressing low-frequency reverberation in the bubble envelope. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide a method and system for suppressing low-frequency reverberation in the bubble envelope of shallow-water seismic data, which overcomes the above problems or at least partially solves the above problems.

[0005] According to one aspect of the present invention, a method for suppressing low-frequency reverberation in a bubble envelope of shallow-water seismic data is provided, the method comprising:

[0006] Step S1: Acquire shallow water time domain seismic signals;

[0007] Step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain;

[0008] Step S3: Designing a time-delay-logarithmic domain filter;

[0009] Step S4: time-delay-logarithmic domain filtering suppresses low-frequency reverberation in the bubble envelope;

[0010] Step S5: Return to the time domain to obtain the noise-suppressed seismic data.

[0011] Optionally, the step S1: obtaining time domain seismic signals in shallow water seismic data specifically includes: in shallow water seismic data, under the combined effect of reverberation and bubbles, low-frequency reverberation appears in the seismic data, and its amplitude strength changes show an obvious bubble envelope trend.

[0012] Optionally, the step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain specifically includes:

[0013] An algorithm for decomposing deconvolution filters in the time-delay-logarithmic domain using the Kolmogorov spectral decomposition method;

[0014] By parameterizing the logarithm of the time-delayed-logarithmic signal spectrum and selecting parameters along the inverse frequency axis, a filter is designed to suppress the bubble effect and ringing noise.

[0015] Under the premise of maintaining the minimum phase characteristics of the data, the boundary parameters of the filter are defined in the parameterized time-delay-logarithmic domain to suppress the low-frequency reverberation in the bubble envelope;

[0016] The seismic data containing low-frequency reverberation in the bubble envelope can be expressed as follows:

[0017] s(n)=b(n)*m(n-τ1)*q(n-τ2)

[0018] Where: s(n) is the observed seismic signal, b(n) is the noise-free seismic signal, m(n-τ1) is the reverberation noise, τ1 is the time delay of the reverberation with respect to the effective signal, representing the reverberation period, q(t-τ2) is the bubble noise, τ2 is the time delay of the bubble noise, representing the expansion and contraction period of the bubble.

[0019] Optionally, the step S2 of transforming the time-domain seismic signal into the time-delay-logarithmic domain further includes:

[0020] Transform the data into the time-delay-logarithmic domain to obtain the inverse spectrum of the data,

[0021] c(τ)=Z -1 [lnS(Z)]=Z-1 [lnB(Z)]+Z -1 [lnM(Z)]+Z -1 [lnQ(Z)]

[0022] Optionally, the step S2 of transforming the time-domain seismic signal into the time-delay-logarithmic domain further includes:

[0023] The process from the time domain to the time-delay-logarithmic domain is: s(n)→S(Z)→lnS(Z)→c(τ).

[0024] Optionally, the step S3: designing a time-delay-logarithmic domain filter specifically includes:

[0025] Define a filter F(Z), let lnF(Z)=U(Z), then F(Z)=e U(Z) ;

[0026] In order to realize the time-delay-logarithmic domain filtering, a time function u is defined τ ,u τ The value of is used to parameterize the time-delay-logarithmic domain filter;

[0027] By using the Kolmokorov method, U(z) and u τ Associating, we get the polynomial:

[0028]

[0029] Where: τ defines the time-delay-logarithmic domain horizontal axis, also called the inverse frequency axis;

[0030] According to a property of the exponential function exp(A+B+C)=exp(A)exp(B)exp(C), when U(z) is used to find the power, It is also decomposed into small time lag term, medium time lag term and large time lag term;

[0031] A filtering operator is decomposed into three parts, each of which has its fixed marine seismological meaning;

[0032]

[0033] When Clearbout introduced a similar equation, he used the second term on the right side of the equation as the zero-phasing operator.

[0034] Adjust the second term to maintain the minimum phase characteristic of the seismic data, so that Represents a continuous earthquake signal, which is also a small time lag signal. is the noise with medium time delay, representing the reverberation noise, is the large time-delay noise, representing the bubble noise;

[0035] Where a and b are boundary parameters, which are used to define the range of effective signal, reverberation and bubble noise in the time-delay-logarithmic domain;

[0036] By setting the coefficients of the parameterized partial time-delay signal specified in the Kolmogorov filter to zero, a low-frequency reverberation filter that suppresses the bubble envelope is constructed.

[0037] Optionally, the filter coefficient setting specifically includes:

[0038] The first one uses the medium time delay filter coefficient (1,0,0…0,u a+1 ,u a+2 …u b ,0,0…) to suppress reverberation, and then use a large time-delay filter coefficient Filtering is performed to suppress the bubble effect. By applying these two filters, the low-frequency reverberation in the bubble envelope that is difficult to separate in the time domain can be suppressed.

[0039] The second method is to set the filter coefficients to take into account that the time-delayed logarithmic domain filtering does not affect the continuous signal of τ = (0 ~ a). The low-frequency reverberation in the bubble envelope is suppressed by filtering out the medium-delay and large-delay signals simultaneously.

[0040] Optionally, step S4: suppressing low-frequency reverberation in the bubble envelope by filtering in the time-delay-logarithmic domain specifically includes: converting the time signal into the time-delay-logarithmic domain, so that the signal and noise that were originally mixed together in the time domain and difficult to separate become "additive" signals in the time-delay-logarithmic domain, and suppressing the noise using a designed filter to obtain a valid signal in the time-delay-logarithmic domain without noise:

[0041] c o (t) = Z -1 [lnB(Z)].

[0042] Optionally, step S5: returning to the time domain to obtain the noise-suppressed seismic data specifically includes: performing signal-noise separation in the time-lag-logarithmic domain to obtain data in which low-frequency reverberation in the bubble envelope is suppressed;

[0043] Transform the low-frequency reverberation data back to the time domain to obtain the seismic data with the low-frequency reverberation in the bubble envelope suppressed, b(n) = Z -1 [expC o (Z)].

[0044] The present invention also provides a system for suppressing low-frequency reverberation in a bubble envelope of shallow-water seismic data, which uses the above-mentioned method for suppressing low-frequency reverberation in a bubble envelope of shallow-water seismic data. The suppression system comprises:

[0045] Seismic signal acquisition module, which acquires shallow water time-domain seismic signals;

[0046] A seismic signal conversion module, configured to convert the time-domain seismic signal into a time-delay-logarithmic domain;

[0047] Filter design module, used to design time-delay-logarithmic domain filters;

[0048] Low-frequency reverberation suppression module, used for time-delay-logarithmic domain filtering to suppress low-frequency reverberation in bubble envelope;

[0049] The suppressed seismic data module is used to return to the time domain and obtain the seismic data after noise suppression.

[0050] The present invention provides a method and system for suppressing low-frequency reverberation in bubble envelopes of shallow-water seismic data. The method comprises: step S1: acquiring shallow-water time-domain seismic signals; step S2: transforming the time-domain seismic signals into a time-delayed-logarithmic domain; step S3: designing a time-delayed-logarithmic domain filter; step S4: suppressing low-frequency reverberation in the bubble envelopes through time-delayed-logarithmic filtering; and step S5: returning to the time domain to obtain de-noised seismic data. This method effectively suppresses low-frequency reverberation in bubbles and bubble envelopes.

[0051] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 Seismic data and its spectrum and cepstrum provided by the embodiment of the present invention;

[0054] Figure 2 A schematic diagram of the effect of suppressing low-frequency reverberation using different filtering parameters provided by an embodiment of the present invention;

[0055] Figure 3 The embodiments of the present invention provide different methods for suppressing low-frequency reverberation effects. DETAILED DESCRIPTION

[0056] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0057] The terms "comprises" and "comprising" and any variations thereof in the description, embodiments, claims and drawings of the present invention are intended to cover non-exclusive inclusions, for example, including a series of steps or units.

[0058] The technical solution of the present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0059] In shallow-water seismic exploration, bubble effects and reverberation are two common types of seismic noise. Deterministic deconvolution and predictive deconvolution based on far-field wavelets are commonly used to suppress these two types of noise.

[0060] In the seismic data of a certain water area, under the combined effect of reverberation and bubbles, a low-frequency reverberation with strong energy appears in the seismic data, and its amplitude changes show an obvious bubble envelope trend, such as Figure 3 As shown in a. The low-frequency reverberation in this bubble envelope masks the effective seismic signal, which not only reduces the signal-to-noise ratio of the seismic data, forms a strong notch effect, and reduces the resolution of the seismic data, but also affects the subsequent ghost wave suppression. Driven by the demand for broadband data, people generally recognize that the low-frequency components of seismic signals are very important for optimizing full waveform inversion and improving deep target imaging. Therefore, how to suppress this low-frequency reverberation while protecting the low-frequency signal of the seismic data has brought new challenges to seismic data processing. Conventional predictive deconvolution and deterministic deconvolution using far-field wavelets are not ideal when suppressing low-frequency reverberation in the bubble envelope.

[0061] Based on the Ricker wavelet deconvolution proposed by Claerbout, the content of the time-delay signal in the time-delay-logarithmic domain was redefined. Parameters were selected along the "inverse frequency" axis, and a filter was designed to suppress the low-frequency reverberation in the bubble envelope, achieving good results in actual seismic data processing.

[0062] 2. Transform the time domain seismic signal to the time-delay-logarithmic domain

[0063] Claerbout's "Rake wavelet deconvolution" algorithm utilizes Kolmogorov spectral decomposition to decompose the deconvolution filter in the time-delayed-logarithmic domain. By parameterizing the time-delayed-logarithmic signal and selecting parameters along the "reciprocal frequency" axis, the filter design effectively suppresses the bubble effect and wavelet zero-phase processing. Our method avoids the debate over zero-phase versus minimum-phase. Instead, we define a filter in the parameterized time-delayed-logarithmic domain while maintaining the minimum-phase characteristic of the data, suppressing low-frequency reverberation in the bubble envelope.

[0064] The seismic data containing low-frequency reverberation in the bubble envelope can be expressed as follows:

[0065] s(n)=b(n)*m(n-τ1)*q(n-τ2)

[0066] Where: s(n) is the observed seismic signal, b(n) is the noise-free seismic signal, m(n-τ1) is the reverberation noise, τ1 is the time delay of the reverberation relative to the effective signal, representing the reverberation period, q(t-τ2) is the bubble noise, τ2 is the time delay of the bubble noise, representing the expansion and contraction period of the bubble. The data is transformed into the time-delay-logarithmic domain to obtain the inverse spectrum of the data,

[0067] c(τ)=Z- 1 [lnS(Z)]

[0068] =Z -1 [lnB(Z)]+Z -1 [lnM(Z)]+Z- 1 [lnQ(Z)].

[0069] Figure 1 Where a represents the seismic trace, b represents the power spectrum, and c represents the inverse spectrum.

[0070] The process from the time domain to the time-delay-logarithmic domain is described as follows:

[0071] s(n)→S(Z)→>lnS(Z)→c(τ)

[0072] Designing Time-Delay-Log Domain Filters

[0073] Define a filter F(Z), let lnF(Z)=U(Z), then F(Z)=e U(Z) In order to realize the time-delay-logarithmic domain filtering, we define a time function u τ ,u τ The value of is used to parameterize the time-delay-logarithmic domain filter. U(z) and u are transformed into τ Associating, we get the polynomial:

[0074] Where: τ defines the time-lag-logarithmic domain horizontal axis, also called the inverse frequency axis; according to a property of the exponential function exp(A+B+C)=exp(A)exp(B)exp(C), when U(z) is raised to the power, It is also decomposed into small lag terms, medium lag terms and large lag terms.

[0075] A filter operator is decomposed into three parts, and each part has its fixed marine seismological meaning.

[0076]

[0077] When Clearbout introduced a similar equation, he used the second term on the right side of the equation as a zero-phase operator. In the present invention, the second term is adjusted to maintain the minimum phase characteristics of the seismic data, so that Represents a continuous earthquake signal, which is also a small time lag signal. is the noise with medium time delay, representing the reverberation noise, is a large time-delay noise, representing bubble noise.

[0078] A complex seismic trace containing a mixture of effective signal, reverberation noise, and bubble noise is parameterized. a and b are boundary parameters that define the range of the effective signal, reverberation, and bubble noise in the time-delay-logarithmic domain. By setting the coefficients of the parameterized portion of the time-delay signal specified in the Kolmogorov filter to zero, a filter is constructed to suppress the low-frequency reverberation in the bubble envelope. There are two ways to set the coefficients of this filter:

[0079] The first one uses the medium time delay filter coefficient (1,0,0…0,u a+1 ,u a+2 …u b ,0,0…) to suppress reverberation, and then use a large time-delay filter coefficient Filtering is performed to suppress the bubble effect. By applying these two filters, the low-frequency reverberation in the bubble envelope that is difficult to separate in the time domain can be suppressed.

[0080] The second method is to set the filter coefficients to take into account that the time-delay-logarithmic domain filtering does not affect the continuous signal of τ = (0 ~ a). The low-frequency reverberation in the bubble envelope is suppressed by filtering out the medium-delay and large-delay signals simultaneously.

[0081] Suppressing low-frequency reverberation in bubble envelopes by time-delay-logarithmic filtering

[0082] By converting the time signal into the time-delay-logarithmic domain, the signal and noise that were originally mixed together in the time domain and difficult to separate become "additive" signals in the time-delay-logarithmic domain. The present invention uses the filter designed in the third step to suppress the noise and obtain an effective signal without noise in the time-delay-logarithmic domain: c o (τ) = Z -1 [lnB(Z)].

[0083] Return to the time domain to obtain the noise-suppressed seismic data

[0084] By performing signal-noise separation in the time-delay-logarithmic domain, we can obtain data that suppresses the low-frequency reverberation in the bubble envelope. By transforming this data back to the time domain, we can obtain seismic data that suppresses the low-frequency reverberation in the bubble envelope. b(n) = Z -1 [expC o (Z)]

[0085] The process from the time-delay-logarithmic domain to the time domain is described as follows:

[0086]

[0087] Example 2

[0088] The present invention provides a technology for suppressing low-frequency reverberation in bubble envelopes in shallow-water seismic exploration. The technology separates the signal and noise by transforming the effective signal and noise, which are difficult to separate in the time domain, into the time-delay-logarithmic domain. The original noisy signal can be regarded as the convolution of the effective signal, bubble noise and reverberation noise. The present invention converts the convolution signal into an "additive" signal by transforming the time domain signal into the time-delay-logarithmic domain. According to the different distribution ranges of the effective signal, reverberation noise and bubble noise in the time-delay-logarithmic domain, they are defined as small time-delay signals, medium time-delay signals and large time-delay signals, and a filter is designed to suppress the low-frequency reverberation in the bubble envelope in the time-delay-logarithmic domain. The specific implementation process is as follows:

[0089] Step S1: Perform Z transform on the original signal to obtain a frequency domain expression, that is, obtain the spectrum of the original signal.

[0090] Step S2: Calculate the absolute value of the spectrum obtained in step S1 to obtain the power spectrum, such as Figure 1 As shown in b.

[0091] Step S3: Calculate the logarithm of the power spectrum obtained in step S2.

[0092] Step S4: Perform an inverse Z transform on the logarithm obtained in step S3 to obtain the expression of the data in the time-delay-logarithmic domain, that is, the inverse spectrum of the data, such as Figure 1 As shown in c.

[0093] Step S5: Design a filter based on the distribution range of the effective signal, reverberation, and bubble noise in the time-delay-logarithmic domain.

[0094] Step S6: Apply the filter obtained in step S5 to the time-delayed-logarithmic domain signal obtained in step 4 to suppress reverberation noise and bubble noise.

[0095] Step S7: performing a Z transform on the time-delayed-logarithmic domain data obtained in step S6 after suppressing reverberation and bubble noise, to obtain the logarithm of the frequency domain signal.

[0096] Step S8: exponentiate the data obtained in step S7.

[0097] Step S9: Perform an inverse Z transform on the data obtained in step S8 to obtain a time domain signal of the low-frequency reverberation suppressed in the bubble envelope.

[0098] In step S5, different filtering parameters are selected to achieve different noise suppression effects. In actual processing, the time-delay-logarithmic domain intercept length is used to define the retained data range. The second coefficient setting method for designing the time-delay-logarithmic domain filter is used as described above. The boundary parameter a is obtained by dividing the intercept length by the data sampling rate. Figure 1 Taking earthquake data as an example, Figure 1 The time domain signal of a is transformed into Figure 1 c in the time-delay-logarithmic domain. The time interval between the continuous signal and the noise, 0.064 s, is used as the starting point for noise suppression. This corresponds to a 4 ms sampling rate for seismic data. When setting the filter coefficient, a = 15. Selecting an appropriate filter coefficient suppresses low-frequency reverberation in the bubble envelope through time-delay-logarithmic filtering.

[0099] Figure 2 This is a schematic diagram of the effect of suppressing low-frequency reverberation with different filter parameters.

[0100] Figure 3 The comparison of the recordings before and after using these four methods to suppress the low-frequency reverberation in the bubble envelope is shown in Figure 2. Figure 3 There are a lot of low-frequency reverberations in the original data shown in a. The amplitude changes of these low-frequency reverberations are obviously affected by bubbles, and have a "strong-weak-strong" change trend from shallow to deep. The effect of using far-field wavelets to suppress the low-frequency reverberations in the bubble envelope is shown in Figure 1. Figure 3 As shown in b, although the noise has been suppressed to a certain extent, there are still a lot of low-frequency reverberations in the record; the record after suppressing the low-frequency reverberations in the bubble envelope using predictive deconvolution is as follows Figure 3 As shown in c, the noise is well suppressed, but residual noise can still be observed in the recording; Figure 3 d is a record after the low-frequency reverberation of the bubble envelope is suppressed by the present invention. After being processed by the present invention, the low-frequency reverberation with the bubble envelope amplitude variation characteristic is well suppressed.

[0101] Beneficial effect: The present invention transforms noisy seismic data into the time-lag-logarithmic domain, utilizes the different distribution ranges of signals and noise in the time-lag-logarithmic domain, designs a Kolmogorov filter, and filters the data to ultimately suppress noise and improve the wave field characteristics of seismic data.

[0102] The above specific implementation methods further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation methods 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 in the scope of protection of the present invention.

Claims

1. A method for suppressing low-frequency reverberation in bubble envelopes of shallow-water seismic data, characterized in that: The pressing method comprises: Step S1: Acquire shallow water time domain seismic signals; Step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain; Step S3: Designing a time-delay-logarithmic domain filter; Step S4: time-delay-logarithmic domain filtering suppresses low-frequency reverberation in the bubble envelope; Step S5: Return to the time domain to obtain the noise-suppressed seismic data.

2. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S1: obtaining time domain seismic signals from shallow water seismic data specifically includes: in shallow water seismic data, under the combined effect of reverberation and bubbles, low-frequency reverberation appears in the seismic data, and its amplitude strength changes show an obvious bubble envelope trend.

3. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain specifically includes: An algorithm for decomposing deconvolution filters in the time-delay-logarithmic domain using the Kolmogorov spectral decomposition method; By parameterizing the time-delayed-logarithmic domain signal and selecting parameters along the inverse frequency axis, a filter is designed to suppress the bubble effect and ringing noise. Under the premise of maintaining the minimum phase characteristics of the data, the boundary parameters of the filter are defined in the parameterized time-delay-logarithmic domain to suppress the low-frequency reverberation in the bubble envelope; The seismic data containing low-frequency reverberation in the bubble envelope can be expressed as follows: s(n)=b(n)*m(n-τ1)*q(n-τ2) Where: s(n) is the observed seismic signal, b(n) is the noise-free seismic signal, m(n-τ1) is the reverberation noise, τ1 is the time delay of the reverberation with respect to the effective signal, representing the reverberation period, q(t-τ2) is the bubble noise, τ2 is the time delay of the bubble noise, representing the expansion and contraction period of the bubble.

4. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain further comprises: Transform the data into the time-delay-logarithmic domain to obtain the inverse spectrum of the data, c(T)=Z -1 [lnS(Z)]=Z -1 [lnB(Z)]+Z -1 [1nM(Z)]+Z -1 [lnQ(Z)] 5. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S2: transforming the time-domain seismic signal into the time-delay-logarithmic domain further comprises: The process from the time domain to the time-delay-logarithmic domain is: s(n)→S(Z)→lnS(Z)→c(τ).

6. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S3: designing a time-delay-logarithmic domain filter specifically includes: Define a filter F(Z), let lnF(Z)=U(Z), then F(Z)=e U(Z) ; In order to realize the time-delay-logarithmic domain filtering, a time function u is defined τ ,u τ The value of is used to parameterize the time-delay-logarithmic domain filter; By using the Kolmokorov method, U(z) and u τ Associating, we get the polynomial: Where: τ defines the time-delay-logarithmic domain horizontal axis, also called the inverse frequency axis; According to a property of the exponential function exp(A+B+C)=exp(A)exp(B)exp(C), when U(z) is used to find the power, It is also decomposed into small time lag term, medium time lag term and large time lag term; A filtering operator is decomposed into three parts, each of which has its fixed marine seismological meaning; When Clearbout introduced a similar equation, he used the second term on the right side of the equation as the zero-phasing operator. Adjust the second term to maintain the minimum phase characteristic of the seismic data, so that Represents a continuous earthquake signal, which is also a small time lag signal. is the noise with medium time delay, representing the reverberation noise, is the large time-delay noise, representing the bubble noise; Where a and b are boundary parameters, which are used to define the range of effective signal, reverberation and bubble noise in the time-delay-logarithmic domain; By setting the coefficients of the parameterized partial time-delay signal specified in the Kolmogorov filter to zero, a low-frequency reverberation filter that suppresses the bubble envelope is constructed.

7. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 6, characterized in that: The coefficient setting of the filter specifically includes: The first one uses the medium time delay filter coefficient (1, 0, 0...0, u a+1 ,u a+2 …u b , 0, 0...) to design a filter to suppress reverberation, and then use a large time-delay filter coefficient Filtering is performed to suppress the bubble effect. By applying these two filters, the low-frequency reverberation in the bubble envelope that is difficult to separate in the time domain can be suppressed. The second method is to set the filter coefficients to take into account that the time-delayed logarithmic domain filtering does not affect the continuous signal of τ = (0 ~ a). The low-frequency reverberation in the bubble envelope is suppressed by filtering out the medium-delay and large-delay signals simultaneously.

8. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S4: suppressing low-frequency reverberation in the bubble envelope by filtering in the time-delay-logarithmic domain specifically includes: converting the time signal into the time-delay-logarithmic domain, so that the signal and noise that were originally mixed together in the time domain and difficult to separate become "additive" signals in the time-delay-logarithmic domain, and suppressing the noise by using the designed filter to obtain an effective signal without noise in the time-delay-logarithmic domain: c o (τ) = Z -1 [lnB(Z)].

9. The method for suppressing low-frequency reverberation in bubble envelope of shallow-water seismic data according to claim 1, characterized in that: The step S5: returning to the time domain to obtain the noise-suppressed seismic data specifically includes: performing signal-noise separation in the time-lag-logarithmic domain to obtain data that suppresses the low-frequency reverberation in the bubble envelope; Transform the low-frequency reverberation data back to the time domain to obtain the seismic data with the low-frequency reverberation in the bubble envelope suppressed, b(n) = Z -1 [expC o (Z)].

10. A system for suppressing low-frequency reverberation in bubble envelopes of shallow-water seismic data, using the method for suppressing low-frequency reverberation in bubble envelopes of shallow-water seismic data according to any one of claims 1 to 9, characterized in that: The pressing method comprises: Seismic signal acquisition module, used to obtain time-domain seismic signals from shallow water seismic data; A seismic signal conversion module, configured to convert the time-domain seismic signal into a time-delay-logarithmic domain; Filter design module, used to design time-delay-logarithmic domain filters; Low-frequency reverberation suppression module, used for time-delay-logarithmic domain filtering to suppress low-frequency reverberation in bubble envelope; The suppressed seismic data module is used to return to the time domain and obtain the seismic data after noise suppression.

Images