Method for protecting low-frequency signals of seismic data

Abstract

The present invention provides a method for protecting low-frequency signals of seismic data, comprising: step 1, according to the seismic single shot record to be processed, using modified S transform to perform synchronous time-frequency analysis on the seismic single shot records of different offset distances; step 2, synchronously and automatically identifying the distribution time, distribution frequency, distribution range and distribution energy of surface wave noise in the time-frequency-offset domain spectrum; step 3, transforming the time-offset domain data into the time-apparent velocity domain; step 4, performing modified S transform on the surface wave data after radial trace transformation; step 5, performing modified S inverse transform and radial trace inverse transform on the extracted surface wave noise to obtain the surface wave noise in the time-offset domain. The method for protecting low-frequency signals of seismic data can give full play to the differences between surface waves and low-frequency effective signals in terms of frequency and apparent velocity, and subtract the surface wave noise from the original seismic single shot record, thereby realizing the suppression of surface wave noise and the protection of low-frequency signals.

Description

Technical Field

[0001] The invention relates to the field of oil and gas geophysical seismic data processing, and in particular to a method for protecting low-frequency signals of seismic data. Background Art

[0002] With the continuous deepening of oil and gas exploration and development, the geological problems faced are becoming more and more complex, and higher requirements are placed on the fine processing of seismic data. Single-point high-density seismic technology is gradually promoted and applied. The frequency band of seismic data is not only widened to the high frequency band, but also the low-frequency band information is becoming more and more abundant. The low-frequency effective signal in the seismic data is submerged in the surface wave interference, which seriously affects the quality of the seismic data. Surface wave suppression is a crucial step in the seismic data processing link. As a regular interference, seismic surface waves are commonly found in onshore seismic data. They have the characteristics of low frequency, strong energy, and low speed. The surface wave energy is about ten times the effective signal energy. It is distributed approximately linearly on the shot gather. The surface wave and the low-frequency effective signal are intertwined and overlapped. Low-frequency signals can maintain a wider effective frequency band, improve seismic resolution, and are conducive to establishing an initial velocity model and improving deep imaging quality. The protection of low-frequency signals is becoming more and more important.

[0003] Surface waves have the characteristics of low frequency, strong energy, and low apparent velocity. The surface wave suppression methods mainly include:

[0004] The bandpass filtering method based on frequency difference is to design a high-pass filter according to the frequency distribution range of surface waves in seismic data, set the low cutoff frequency of the filter, filter the seismic data, and filter out the components below the low cutoff frequency, so as to realize surface wave processing. The frequency-space domain coherent noise attenuation method is performed on the data in the frequency range of surface waves. Based on the minimum square error criterion, the surface waves are estimated at each frequency. The area where surface waves appear is restricted by the two parameters of velocity and frequency, which has little effect on seismic signals outside the range. The frequency-wavenumber domain surface wave suppression method based on apparent velocity difference performs a two-dimensional Fourier transform on the seismic data to obtain a frequency-wavenumber spectrum. Since there is a difference in apparent velocity between surface waves and effective signals, the low frequency of surface waves is separated from the effective signals in the frequency-wavenumber spectrum. The surface wave area is cut off, and then a two-dimensional Fourier inverse transform is performed to obtain seismic data without surface waves. The adaptive surface roll attenuation method based on frequency and apparent velocity transforms the seismic data into the time-frequency domain. According to the differences in frequency, energy and spatial distribution range between surface rolls and effective signals, the distribution range of surface rolls in time and space is first detected. Then, according to the frequency band range of surface rolls, frequency division and weighted suppression of surface rolls are performed.

[0005] At present, the surface wave suppression method has two main shortcomings. On the one hand, the surface wave identification is inaccurate, and the data without surface waves are also processed, destroying the effective signal; on the other hand, the surface wave suppression loses the effective signal in the low-frequency band. In the process of suppressing surface waves, the method of drawing time windows in the single-shot record in the time-offset range domain has the shortcomings of large workload of time window picking and large influence of human factors. Usually, all data are processed, and the data without surface waves loses the effective signal in the processing process. Through the frequency band processing method, radial trace transformation method or frequency-wavenumber domain surface wave suppression method, the difference characteristics of surface waves and effective signals in frequency and apparent velocity can be brought into play, but the effective signal in the low-frequency band is still lost, which is not conducive to subsequent data interpretation and reservoir inversion.

[0006] Patent CN104345341A proposes a frequency-band energy seismic surface wave processing method based on regional constraints. It uses wavelet transform to perform frequency division processing on seismic data, calculates the average amplitude value in different time windows and different frequency bands, and uses the cumulative superposition statistical analysis method to set the amplitude threshold value. The new amplitude value calculated by the suppression factor formula replaces the original amplitude value, thereby achieving surface wave suppression.

[0007] Patent CN104614769B proposes a beam filtering method for suppressing seismic surface waves. The method analyzes the spatial, temporal, and amplitude distribution of each frequency of the Rayleigh surface waves recorded in the seismic record, obtains the propagation characteristics of the Rayleigh surface waves through methods such as dispersion curve extraction, and then estimates and merges the results of the Fourier transform of the pulse sequence of the surface waves. A surface wave beam filtering method is constructed using determined parameters, and the effective signal after filtering can be obtained by taking the inverse Fourier transform of the filtered output data.

[0008] Patent CN105319591A proposes an SVD adaptive surface roll suppression method based on radial track transformation, that is, firstly, radial track transformation is used to move the surface roll frequency band to a lower frequency range, then low-pass filtering is used to obtain low-frequency surface roll signals in the radial track domain, and then the singular value decomposition method is used to separate the effective low-frequency signals in the surface roll, and finally the surface roll is subtracted from the original data.

[0009] Patent CN104375185A proposes a method and device for removing surface rolls from seismic records, which uses wavelet frequency division to obtain seismic records of different frequency bands, performs radial trace transformation on the seismic records containing surface rolls, and then performs high-pass filtering, and finally performs wavelet reconstruction to obtain the seismic records with surface rolls removed.

[0010] The above existing technologies are all significantly different from the present invention and fail to solve the technical problem we want to solve. Therefore, we have invented a new method for protecting low-frequency signals of seismic data. Summary of the invention

[0011] The purpose of the present invention is to provide a method for protecting low-frequency signals of seismic data, which can suppress surface wave noise and protect low-frequency signals.

[0012] The purpose of the present invention can be achieved by the following technical measures: a method for protecting low-frequency signals of seismic data, the method comprising:

[0013] Step 1, according to the seismic single shot record to be processed, using modified S transform to perform synchronous time-frequency analysis on the seismic single shot records of different offset distances;

[0014] Step 2, synchronously and automatically identify the distribution time, distribution frequency, distribution range, and distribution energy of surface wave noise in the time-frequency-offset domain spectrum;

[0015] Step 3, transform the time-offset domain data to the time-apparent velocity domain;

[0016] Step 4, performing a modified S transform on the surface wave data after the radial track transform;

[0017] Step 5: Perform modified S inverse transform and radial trace inverse transform on the extracted surface roll noise to obtain the surface roll noise in the time-offset domain.

[0018] The purpose of the present invention can also be achieved by the following technical measures:

[0019] Step 1: Based on the seismic single shot records to be processed, the modified S transform is used to perform synchronous time-frequency analysis on the seismic single shot records with different offsets. The time-frequency-offset domain spectrum calculation formula of the seismic single shot records is as follows:

[0020]

[0021] Where: s(t,offset) is the original seismic single shot record with different offset distances, t is time in ms, offset is offset in m; ST P (α,fre,offset) is the time-frequency-offset domain spectrum of s(t,offset), α is the time position of the window function in ms, fre is the frequency in Hz; parameters β and λ are adjustment parameters of the modified S transform.

[0022] Step 2: Perform synchronous automatic identification of the surface roll noise distribution time, distribution frequency, distribution range, and distribution energy in the time-frequency-offset domain spectrum. The four-synchronous automatic identification formulas of the surface roll noise distribution time, distribution frequency, distribution range, and distribution energy are as follows:

[0023]

[0024] Among them: RE(t,fre,w,eng) is the four-synchronous recognition result of surface wave noise, t is time, unit ms, fre is frequency, unit Hz, w is range, unit m, eng is energy, dimensionless unit; when the energy difference |ST P (t i ,fre,w)|-|ST P (t i-1 ,fre,w)|When the recognition condition is met, t1=t i , i is the number of sampling points, t2 is the single shot recording length, unit ms; when the energy difference |ST P (t,fre j ,w)|-|ST P (t,fre j-1 ,w)|When the recognition condition is met, fre1 is the starting frequency value, fre2=fre j , j is the time spectrum frequency sampling interval, unit Hz; when the energy difference |ST P (t,fre,w k )|-|ST P (t,fre,w k-1 ) When the recognition conditions are met, w1 is the starting offset, w2 = w k , k is the offset interval, unit is m; eng1 and eng2 are the distribution energy identification ranges, which are calculated based on the energy of the surface wave and effective signal in the single shot record.

[0025] Step 2: Decompose the single shot record s(t, offset) to be processed into the surface wave data s N (t, offset) and does not contain surface wave data s M (t, offset), only for the data within the surface wave distribution time, distribution frequency, distribution range, and distribution energy s N (t, offset) is used to suppress noise, where t is time in ms and offset is offset in m, thereby protecting the effective signal in the low-frequency band from loss.

[0026] Step 3: For the data s containing surface waves N (t, offset) is used to transform the time-offset domain data into the time-apparent velocity domain. The radial trace transformation formula containing surface wave data is as follows:

[0027]

[0028] Where: RT[] is the radial track transformation, s N (t, offset) contains surface wave data, t is time, unit is ms, offset is offset distance, unit is m; It is the data after radial track transformation, t0 is the starting time, in ms, and offset0 is the starting offset, in m.

[0029] Step 4: Perform a modified S transform on the surface wave data after the radial trace transform. The time-frequency-apparent velocity domain spectrum calculation formula is as follows:

[0030]

[0031] Where: RT[s N (t, offset)] is the surface wave data after radial trace transformation, t is time, unit is ms, offset is offset, unit is m; ST RT (α,fre,offset) is RT[s N The time-frequency-apparent velocity domain spectrum of the window function is (t, offset)], α is the time position of the window function, unit is ms, fre is the frequency, unit is Hz; parameters β and λ are the adjustment parameters of the modified S transform.

[0032] Step 4, using radial trace transformation and modified S transform, the two-dimensional time-apparent velocity domain is converted into a three-dimensional time-frequency-apparent velocity domain. Surface roll has the characteristics of low frequency and low velocity value, and effective signal has the characteristics of low frequency and high velocity value. In the time-frequency-apparent velocity domain, the differences between surface roll and low-frequency effective signal in frequency and apparent velocity are fully utilized to better distinguish surface roll and low-frequency effective signal, so as to extract surface roll noise.

[0033] Step 5, the extracted surface wave noise is subjected to modified S inverse transform and radial trace inverse transform to obtain the surface wave noise in the time-offset domain, and the surface wave noise is subtracted from the original seismic single shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency signals.

[0034] The low-frequency signal protection method of seismic data in the present invention uses modified S transform to perform time-frequency analysis on seismic single-shot records of different offset distances, and performs synchronous automatic identification of surface wave distribution time, distribution frequency, distribution range, and distribution energy in the time-frequency-offset domain. Through the four-synchronous automatic identification processing of surface wave distribution time, distribution frequency, distribution range, and distribution energy, the single-shot record is divided into data containing noise and data not containing noise. Only the data containing noise is subjected to noise suppression, and the data not containing surface wave is not processed, thereby protecting the effective signal from loss. In the process of surface wave suppression, radial trace transformation is performed on the data containing surface waves, and the time-offset domain data is transformed into the time-apparent velocity domain. The surface wave frequency moves to a lower frequency band. On this basis, the data after radial trace transformation is subjected to modified S transformation. In the time-frequency-apparent velocity domain, the differences between surface waves and low-frequency effective signals in terms of frequency and apparent velocity can be fully utilized to extract the surface wave noise, perform modified S inverse transformation and radial trace inverse transformation, and subtract the surface wave noise from the original seismic single-shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of a specific embodiment of the method for protecting low-frequency signals of seismic data of the present invention;

[0036] Figure 2 A schematic diagram of processing a pre-seismic single shot record in a specific embodiment of the present invention;

[0037] Figure 3 It is a schematic diagram of the time-frequency-offset domain data of a seismic single shot record in a specific embodiment of the present invention;

[0038] Figure 4 It is a schematic diagram of time-apparent velocity domain data containing surface wave data after radial trace transformation in a specific embodiment of the present invention;

[0039] Figure 5 It is a schematic diagram of the time-frequency-apparent velocity domain data of a seismic single shot record after the modified S-transform in a specific embodiment of the present invention;

[0040] Figure 6 It is a schematic diagram of the radial channel surface wave noise suppressed in a specific embodiment of the present invention;

[0041] Figure 7 A schematic diagram of suppressed surface wave noise interference in a specific embodiment of the present invention;

[0042] Figure 8 It is a schematic diagram of a seismic single shot record after suppressing surface waves in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0043] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0044] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0045] The invention discloses a method for protecting low-frequency signals of seismic data. Firstly, a modified S transform is used to perform time-frequency analysis on seismic single-shot records of different offsets, and synchronous automatic identification of surface wave distribution time, distribution frequency, distribution range and distribution energy is performed in a time-frequency-offset domain, and the seismic single-shot records are divided into data containing noise and data not containing noise, and noise suppression is performed only on the data containing noise, thereby protecting effective signals from loss. Secondly, a radial trace transform is performed on the data containing surface waves, and the time-offset domain data is transformed into the time-apparent velocity domain, and the surface wave frequency moves to a lower frequency band. Then, a modified S transform is performed on the data after the radial trace transform, and the difference in frequency and apparent velocity between the surface wave and the low-frequency effective signal can be fully utilized in the time-frequency-apparent velocity domain. Finally, the surface wave noise is extracted, and a modified S inverse transform and a radial trace inverse transform are performed, and the surface wave noise is subtracted from the original seismic single-shot record, thereby realizing the suppression of surface wave noise and the protection of low-frequency signals.

[0046] Embodiment 1:

[0047] In a specific embodiment 1 of the present invention, Figure 1 As shown, Figure 1 The flowchart of the low-frequency signal protection method of seismic data of the present invention includes the following steps:

[0048] Step 1: Import the seismic single shot records to be processed ( Figure 2 ), using modified S transform to perform synchronous time-frequency analysis on seismic single shot records with different offsets, the time-frequency-offset domain spectrum of seismic single shot records ( Figure 3 ) is calculated as follows:

[0049]

[0050] Where: s(t,offset) is the original seismic single shot record with different offset distances, t is time in ms, offset is offset in m; ST P (α,fre,offset) is the time-frequency-offset domain spectrum of s(t,offset), α is the time position of the window function in ms, fre is the frequency in Hz; parameters β and λ are adjustment parameters of the modified S transform.

[0051] Step 2: Perform synchronous automatic identification of the surface wave noise distribution time, distribution frequency, distribution range, and distribution energy in the time-frequency-offset domain spectrum. The formula for the four synchronous automatic identification of the surface wave noise distribution time, distribution frequency, distribution range, and distribution energy is as follows:

[0052]

[0053] Among them: RE(t,fre,w,eng) is the four-synchronous recognition result of surface wave noise, t is time, unit ms, fre is frequency, unit Hz, w is range, unit m, eng is energy, dimensionless unit; when the energy difference |ST P (t i ,fre,w)|-|ST P (t i-1 ,fre,w)|When the recognition condition is met, t1=t i , i is the number of sampling points, t2 is the single shot recording length, unit ms; when the energy difference |ST P (t,fre j ,w)|-|ST P (t,fre j -1,w) meets the recognition condition, fre1 is the starting frequency value, fre2 = fre j , j is the time spectrum frequency sampling interval, unit Hz; when the energy difference |ST P (t,fre,w k )|-|ST P (t,fre,w k-1 ) When the recognition conditions are met, w1 is the starting offset, w2 = w k , k is the offset interval, unit is m; eng1 and eng2 are the distribution energy identification ranges, which are calculated based on the energy of the surface wave and effective signal in the single shot record.

[0054] Decompose the single shot record s(t, offset) to be processed into the surface wave data s N (t, offset) and does not contain surface wave data s M (t, offset), only for the data within the surface wave distribution time, distribution frequency, distribution range, and distribution energy sN (t, offset) is used to suppress noise, where t is time in ms and offset is offset in m, thereby protecting the effective signal in the low-frequency band from loss.

[0055] Step 3: For the data containing surface waves N (t, offset) is used to transform the time-offset domain data into the time-apparent velocity domain. The radial trace transformation formula containing surface wave data is as follows:

[0056]

[0057] Where: RT[] is the radial track transformation, s N (t, offset) contains surface wave data, t is time, unit is ms, offset is offset distance, unit is m; It is the data after radial track transformation, t0 is the starting time, in ms, and offset0 is the starting offset, in m.

[0058] Figure 4 The radial trace transformation is the time-apparent velocity domain data containing surface roll data. The radial trace transformation makes the surface roll frequency move to a lower frequency band, which is beneficial to suppressing surface roll noise and protecting the effective signal in the low frequency band.

[0059] Step 4: Perform a modified S transform on the surface wave data after radial trace transformation. The time-frequency-apparent velocity domain spectrum calculation formula is as follows:

[0060]

[0061] Where: RT[s N (t, offset)] is the surface wave data after radial trace transformation, t is time, unit is ms, offset is offset, unit is m; ST RT (α,fre,offset) is RT[s N The time-frequency-apparent velocity domain spectrum of the window function is (t, offset)], α is the time position of the window function, unit is ms, fre is the frequency, unit is Hz; parameters β and λ are the adjustment parameters of the modified S transform.

[0062] Figure 5To correct the time-frequency-apparent velocity domain data of seismic single shot records after S transformation. Using radial trace transformation and modified S transformation, the two-dimensional time-apparent velocity domain is converted into a three-dimensional time-frequency-apparent velocity domain. Surface waves have the characteristics of low frequency and low velocity value, and effective signals have the characteristics of low frequency and high velocity value. In the time-frequency-apparent velocity domain, the differences between surface waves and low-frequency effective signals in terms of frequency and apparent velocity can be fully utilized, and surface waves and low-frequency effective signals can be better distinguished, thereby extracting surface wave noise.

[0063] Step 5: Perform the modified S inverse transform and radial trace inverse transform on the extracted surface roll noise to obtain the surface roll noise in the time-offset domain. Figure 6 To suppress the surface wave noise interference, the surface wave noise is subtracted from the original seismic single shot record. Figure 7 To suppress the seismic single shot record after the surface wave, thus achieving the suppression of surface wave noise and the protection of low-frequency signals, Figure 8 This is a single-shot seismic record after surface wave suppression.

[0064] Embodiment 2:

[0065] In the specific embodiment 2 of the present invention, the following steps are included:

[0066] Step 1: Import the seismic single shot record to be processed, and use the modified S transform to perform synchronous time-frequency analysis on the seismic single shot records of different offsets. The time-frequency-offset domain spectrum calculation formula of the seismic single shot record is as follows:

[0067]

[0068] Where: s(t,offset) is the original seismic single shot record with different offset distances, t is time in ms, offset is offset in m; ST P (α,fre,offset) is the time-frequency-offset domain spectrum of s(t,offset), α is the time position of the window function, unit is ms, fre is the frequency, unit is Hz; parameters β and λ are the adjustment parameters of the modified S transform. The corresponding adjustment parameter size can be set according to the characteristics of different seismic data. Here, the values ​​are β=2.0 and λ=1.0.

[0069] Step 2: Perform automatic and synchronous identification of the distribution time, distribution frequency, distribution range, and distribution energy of surface wave noise in the time-frequency-offset domain spectrum. Decompose the single shot record s(t, offset) to be processed into the surface wave data s N (t, offset) and does not contain surface wave data s M (t, offset), only for the data within the surface wave distribution time, distribution frequency, distribution range, and distribution energy sN (t, offset) is used to suppress noise, where t is time in ms and offset is offset in m, thereby protecting the effective signal in the low-frequency band from loss.

[0070] Step 3: For the data containing surface waves N (t,offset) performs radial trace transformation to transform the time-offset domain data into the time-apparent velocity domain.

[0071] Step 4: Perform a modified S transform on the surface wave data after radial trace transformation. The time-frequency-apparent velocity domain spectrum calculation formula is as follows:

[0072]

[0073] Where: RT[s N (t, offset)] is the surface wave data after radial trace transformation, t is time, unit is ms, offset is offset, unit is m; ST RT (α,fre,offset) is RT[s N The time-frequency-apparent velocity domain spectrum of [t, offset] is shown in Figure 5. α is the time position of the window function, in ms, and fre is the frequency, in Hz. The parameters β and λ are the adjustment parameters of the modified S transform. The corresponding adjustment parameters can be set according to the characteristics of different seismic data. Here, the values ​​are β = 2.0 and λ = 1.0.

[0074] The two-dimensional time-apparent velocity domain is converted into a three-dimensional time-frequency-apparent velocity domain by using radial trace transform and modified S transform. Surface roll has the characteristics of low frequency and low velocity value, and effective signal has the characteristics of low frequency and high velocity value. In the time-frequency-apparent velocity domain, the differences between surface roll and low-frequency effective signal in frequency and apparent velocity can be fully utilized to better distinguish surface roll and low-frequency effective signal, so as to extract surface roll noise.

[0075] Step 5: Perform modified S inverse transform and radial trace inverse transform on the extracted surface wave noise to obtain the surface wave noise in the time-offset domain, and subtract the surface wave noise from the original seismic single shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency signals.

[0076] Embodiment 3:

[0077] In the specific embodiment 3 of the present invention, the following steps are included:

[0078] Step 1: Import the seismic single shot record to be processed. The minimum offset distance of a single shot record is 25m, the maximum offset distance is 4000m, and the offset interval between two traces is 25m. Use the modified S transform to perform synchronous time-frequency analysis on seismic single shot records with different offset distances.

[0079] Step 2: Perform synchronous automatic identification of the surface wave noise distribution time, distribution frequency, distribution range, and distribution energy in the time-frequency-offset domain spectrum. The formula for the four synchronous automatic identification of the surface wave noise distribution time, distribution frequency, distribution range, and distribution energy is as follows:

[0080]

[0081] Among them: RE (t, fre, w, eng) is the four-synchronous identification result of surface wave noise, t is time, unit ms, fre is frequency, unit Hz, w is range, unit m, eng is energy, dimensionless unit. The identification of distribution time is to analyze each data, starting from time 0, using the energy ratio method to identify the appearance time of the surface wave. When the energy difference meets the conditions, the appearance time of the surface wave is identified. The identification of distribution frequency is to identify the surface wave from the frequency domain. When the energy difference meets the conditions, the maximum frequency value of the surface wave is identified. The identification of distribution range is to identify the surface wave from the offset domain. The surface wave is mainly distributed in the near offset. When the energy difference meets the conditions, the maximum offset of the surface wave distribution is identified.

[0082] Step 3: The minimum offset distance of a single shot record is 25m, the maximum offset distance is 4000m, and the offset distance interval between two tracks is 25m. The data containing surface waves are transformed into radial tracks to transform the time-offset domain data into the time-apparent velocity domain.

[0083] Step 4: Perform a modified S transform on the surface roll data after radial trace transformation. In the time-frequency-apparent velocity domain, the differences in frequency and apparent velocity between surface rolls and low-frequency effective signals can be fully utilized to better distinguish surface rolls from low-frequency effective signals.

[0084] Step 5: Perform modified S inverse transform and radial trace inverse transform on the extracted surface wave noise to obtain the surface wave noise in the time-offset domain, and subtract the surface wave noise from the original seismic single shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency signals.

[0085] The present invention has two advantages. On the one hand, the modified S transform is used to perform time-frequency analysis on seismic single-shot records of different offsets, and the surface wave distribution time, distribution frequency, distribution range, and distribution energy are synchronously and automatically identified in the time-frequency-offset domain. Through the four synchronous automatic identification processes of the surface wave distribution time, distribution frequency, distribution range, and distribution energy, the single-shot record is divided into data containing noise and data not containing noise. Only the data containing noise is subjected to noise suppression, and the data not containing surface waves is not processed, thereby protecting the effective signal from loss. On the other hand, in the process of surface wave suppression, the data containing surface waves is subjected to radial trace transformation, and the time-offset domain data is transformed into the time-apparent velocity domain, and the surface wave frequency moves to a lower frequency band. On this basis, the data after the radial trace transformation is subjected to modified S transform. In the time-frequency-apparent velocity domain, the difference between the surface wave and the low-frequency effective signal in terms of frequency and apparent velocity can be fully utilized, and the surface wave noise can be extracted. The modified S inverse transform and the radial trace inverse transform are performed, and the surface wave noise is subtracted from the original seismic single-shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency band signals.

[0086] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0087] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.

Claims

1. A method for protecting low-frequency signals of seismic data, characterized in that: The seismic data low-frequency signal protection method includes: Step 1: Based on the seismic single shot records to be processed, the modified S transform is used to perform synchronous time-frequency analysis on the seismic single shot records with different offsets. The time-frequency-offset domain spectrum calculation formula of the seismic single shot records is as follows: Where: s(t,offset) is the original seismic single shot record with different offset distances, t is time in ms, offset is offset in m; ST P (α,fre,offset) is the time-frequency-offset domain spectrum of s(t,offset), α is the time position of the window function, in ms, fre is the frequency, in Hz; parameters β and λ are the adjustment parameters of the modified S transform; Step 2, synchronously and automatically identify the distribution time, distribution frequency, distribution range, and distribution energy of surface wave noise in the time-frequency-offset domain spectrum; Step 3, performing radial trace transformation on the data containing surface waves, transforming the time-offset domain data into the time-apparent velocity domain; Step 4, performing the modified S transform on the surface wave data after the radial track transform; Step 5: Perform modified S inverse transform and radial trace inverse transform on the extracted surface roll noise to obtain the surface roll noise in the time-offset domain.

2. The method for protecting low-frequency signals of seismic data according to claim 1, characterized in that: Step 2: Perform synchronous automatic identification of the surface roll noise distribution time, distribution frequency, distribution range, and distribution energy in the time-frequency-offset domain spectrum. The four-synchronous automatic identification formulas of the surface roll noise distribution time, distribution frequency, distribution range, and distribution energy are as follows: Among them: RE(t,fre,w,eng) is the four-synchronous recognition result of surface wave noise, t is time, unit ms, fre is frequency, unit Hz, w is range, unit m, eng is energy, dimensionless unit; when the energy difference |ST P (t i ,fre,w)|-|ST P (t i-1 ,fre,w)|When the recognition condition is met, t1=t i , i is the number of sampling points, t2 is the single shot recording length, unit ms; when the energy difference |ST P (t,fre j ,w)|-|ST P (t,fre j-1 ,w)|When the recognition condition is met, fre1 is the starting frequency value, fre2=fre j , j is the time spectrum frequency sampling interval, unit Hz; when the energy difference |ST P (t,fre,w k )|-|ST P (t,fre,w k-1 )|When the recognition conditions are met, w1 is the starting offset, w2=w k , k is the offset interval, unit is m; eng1 and eng2 are the distribution energy identification ranges, which are calculated based on the energy of the surface wave and effective signal in the single shot record.

3. The method for protecting low-frequency signals of seismic data according to claim 2, characterized in that: Step 2: Decompose the single shot record s(t, offset) to be processed into the surface wave data s N (t, offset) and does not contain surface wave data s M (t, offset), only for the data within the surface wave distribution time, distribution frequency, distribution range, and distribution energy s N (t, offset) is used to suppress noise, where t is time in ms and offset is offset in m, thereby protecting the effective signal in the low-frequency band from loss.

4. The method for protecting low-frequency signals of seismic data according to claim 1, characterized in that: Step 3: For the data s containing surface waves N (t, offset) is used to transform the time-offset domain data into the time-apparent velocity domain. The radial trace transformation formula containing surface wave data is as follows: Where: RT[] is the radial track transformation, s N (t, offset) contains surface wave data, t is time, unit is ms, offset is offset distance, unit is m; It is the data after radial track transformation, t0 is the starting time, in ms, and offset0 is the starting offset, in m.

5. The method for protecting low-frequency signals of seismic data according to claim 1, characterized in that: Step 4: Perform a modified S transform on the surface wave data after the radial trace transform. The time-frequency-apparent velocity domain spectrum calculation formula is as follows: Where: RT[s N (t, offset)] is the surface wave data after radial trace transformation, t is time, unit is ms, offset is offset, unit is m; ST RT (α,fre,offset) is RT[s N The time-frequency-apparent velocity domain spectrum of the window function is (t, offset)], α is the time position of the window function, unit is ms, fre is the frequency, unit is Hz; parameters β and λ are the adjustment parameters of the modified S transform.

6. The method for protecting low-frequency signals of seismic data according to claim 5, characterized in that: Step 4, using radial trace transformation and modified S transform, the two-dimensional time-apparent velocity domain is converted into a three-dimensional time-frequency-apparent velocity domain. Surface roll has the characteristics of low frequency and low velocity value, and effective signal has the characteristics of low frequency and high velocity value. In the time-frequency-apparent velocity domain, the differences between surface roll and low-frequency effective signal in frequency and apparent velocity are fully utilized to better distinguish surface roll and low-frequency effective signal, so as to extract surface roll noise.

7. The method for protecting low-frequency signals of seismic data according to claim 1, characterized in that: Step 5, the extracted surface wave noise is subjected to modified S inverse transform and radial trace inverse transform to obtain the surface wave noise in the time-offset domain, and the surface wave noise is subtracted from the original seismic single shot record, thereby achieving the suppression of surface wave noise and the protection of low-frequency signals.

Images