A method and device for separating longitudinal and shear waves of distributed optical fiber seismic data

Abstract

The present invention relates to a method and apparatus for separating P- and S-waves from distributed fiber-optic seismic data. The method comprises: performing denoising on distributed fiber-optic seismic data; estimating the local elastic wavefield polarization vector of a wellbore detector using dispersion relations based on P- and S-wave logging velocities; performing Fourier transform on the elastic wavefield polarization vector to obtain a convolution filter operator in the time-space domain; filtering the distributed fiber-optic vertical seismic profile data using the convolution operator in the time-space domain, and performing P- and S-wave separation on the single-component distributed fiber-optic seismic data through polarization projection based on Helmholtz decomposition; and, to improve separation accuracy, calculating corresponding polarization projection filter operators on common shot point and common receiver point gathers, and applying polarization projection filtering to obtain the final P- and S-wave separation results. Compared with existing technologies, the present invention can flexibly and accurately separate P- and S-wave data.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic longitudinal and shear wave separation, and in particular to a method and device for separating longitudinal and shear waves of distributed optical fiber seismic data. Background Art

[0002] Distributed acoustic sensing (DAS) is a new, low-cost seismic data observation tool. It is typically used to detect phase changes in the backscattered Rayleigh scattering of laser pulses, obtain axial strain or strain rate on the optical fiber, and then obtain the local elastic vibration signal of the seismic wave field. Compared with conventional seismic three-component geophones or other types of sensors, observations using DAS sensors can provide long-term, long-distance, high-resolution continuous measurement results. Therefore, it is particularly suitable for real-time active source seismic observations and long-term passive source seismic monitoring, such as hydraulic fracturing, near-surface imaging, and carbon capture, utilization and storage (CCUS) monitoring. In addition, because optical fiber installation is usually non-invasive and low-cost, DAS acquisition systems have been widely used in deep seismic exploration such as vertical seismic profiles (VSP) in recent years to obtain high-resolution underground imaging.

[0003] In conventional well VSP observations, geophones are typically used to record three-component seismic data. Due to the widespread heterogeneity of the subsurface medium, three-component VSP records often contain complex P-wave and S-wave fields. Utilizing this elastic wavefield information can effectively improve the imaging accuracy of gas cloud regions, enhance the success rate of lithologic characterization, and estimate fracture and stress characteristics. However, these processes often rely heavily on effective elastic wavefield separation techniques. Only by obtaining P / S wave data with accurate amplitude and phase relationships can subsequent elastic seismic imaging be performed.

[0004] Current P / S wavefield separation methods primarily rely on the physical properties of P and S waves, such as polarization direction and propagation velocity. Based on the Helmholtz decomposition theory, polarization projection methods can project the vector elastic wavefield onto the polarization direction of the P or S wave, thereby obtaining pure P-wave or pure S-wave data. The polarization direction at the detector point can be estimated first, and then polarization rotation can be used to separate the P- and S-waves. However, when the data signal-to-noise ratio is low or P / S cross-aliasing occurs, accurately estimating the polarization direction is often difficult, and P / S separation still faces severe amplitude distortion and mode leakage. Another approach is to achieve wavefield separation by filtering in the transform domain based on the apparent velocity difference between the P and S waves through various mathematical transformations, such as the FK transform, linear Radon transform, and parabolic Radon transform. However, in practice, P and S wave data often overlap in energy in the transform domain, making it difficult to obtain satisfactory P and S wave separation, especially when the wavefield is complex. In addition, existing technologies use wave field extension based on elastic wave equations for separation, but such methods are highly dependent on accurate formation elastic parameters. In areas with complex structures, where formation parameters vary dramatically or are difficult to obtain, P / S will still be affected by pattern leakage.

[0005] In summary, most existing P / S separation methods require multi-component VSP data as input. In DAS-VSP acquisition systems, although DAS sensors can simultaneously record P and S waves, they can only measure strain or strain rate along the fiber tangent, meaning only a single component of data is available. Therefore, due to the lack of sufficient information to describe the vector elastic wavefield, P / S wavefield separation in DAS-VSP data is extremely challenging. Currently, the industry has not yet proposed an effective solution to this problem. Therefore, finding a stable and reliable method for elastic wavefield separation in DAS-VSP data to obtain accurate P / S wave data is an urgent issue to improve DAS-VSP data utilization and obtain elastic imaging results for subsurface media. Summary of the Invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method and device for separating longitudinal and shear waves of distributed optical fiber seismic data, which can flexibly and accurately separate longitudinal and shear wave data.

[0007] The purpose of the present invention can be achieved by the following technical solution: A method for separating longitudinal and shear waves of distributed optical fiber seismic data, comprising the following steps:

[0008] S1. Obtain the VSP recorded data collected by the DAS sensor, then pre-process it and convert it according to the relationship between the axial strain wave field and the displacement wave field to obtain the axial wave field displacement vibration signal;

[0009] S2. Acquire acoustic logging velocity information and convert it into layer velocity within the seismic frequency band, thereby obtaining the P-wave and S-wave velocities at each geophone point in the wellbore and calculating the local wavefield polarization direction of each geophone.

[0010] S3. transforming the calculated wave field polarization direction into the time-space domain to obtain the convolution operator in the time-space domain;

[0011] S4. Using the convolution operator in the time and space domain, the axial wave field displacement vibration signal is convolved in the time and space domain, and the initial longitudinal and transverse wave separation results of the gun domain are obtained through polarization projection filtering;

[0012] S5. Re-sort the initial P-wave and S-wave separation results to obtain common detection point gathers;

[0013] S6. In the common detection point gather, according to the wave field reciprocity principle, the ground shot points are regarded as detection points. Based on the dispersion relationship between the emission slowness in the frequency wavenumber domain and the vertical and horizontal wavenumbers, the local wavefield polarization direction of the common detection point gather is calculated. Then, the process returns to execute steps S3 to S4 to obtain the final P-wave and S-wave separation results.

[0014] Furthermore, the preprocessing process in step S1 includes but is not limited to denoising, removing bubbles, and removing coupling noise.

[0015] Furthermore, the step S2 specifically includes the following steps:

[0016] S21. Calculate the interval velocity of the formation based on the acoustic time difference data collected by acoustic logging, then convert the high-frequency logging velocity into interval velocity within the seismic frequency band, and thus obtain the P-wave and S-wave velocities at each detection point in the wellbore;

[0017] S22. On the common shot gather, the P- and S-wave velocities at the detector point are regarded as the exit slowness of the elastic P- and S-wave fields. In the frequency-wavenumber domain, the local wavefield polarization direction of each detector is calculated based on the dispersion relationship between the exit slowness and the vertical and horizontal wavenumbers.

[0018] Furthermore, the step S21 specifically adopts the Backus averaging method to smooth the converted layer velocity of the formation, thereby obtaining the P-wave and S-wave layer velocity within the seismic frequency band.

[0019] Furthermore, the specific process of step S22 is as follows: the P-wave and S-wave velocities at the receiver point in the common shot gather are regarded as the exit slowness of the elastic wave field. Since the optical fiber sensors are arranged vertically or approximately vertically in the DAS-VSP observation, the vertical wave number and frequency information recorded by the VSP are obtained by two-dimensional fast Fourier transform. Then, according to the wave field propagation theory, the exit slowness and vertical wave number are used to calculate the horizontal wave number information of the P-wave and S-wave wave fields through the dispersion relation constraint, that is, k px With k sx Based on this, the Helmholtz decomposition theory is used to construct the polarization projection direction of the longitudinal and transverse wave fields at each detection point:

[0020]

[0021]

[0022]

[0023]

[0024] in, represents the vector displacement wave field in the wavenumber domain, and Represent the separated P-wave and S-wave vector displacement wave fields respectively.

[0025] Furthermore, the specific process of step S3 is as follows: since the P- and S-wave velocities of the strata at different detector positions vary dramatically in the real earth medium, the slowness vector in the dispersion relation needs to be adjusted according to the spatial position. Then, according to the Fourier transform principle, the polarization projection filter operator in the frequency-wavenumber domain is equivalent to the convolution operator in the time-space domain. Thus, the frequency-wavenumber domain polarization filter of the P- and S-wave fields is transformed into the time-space domain to obtain the convolution operator in the time-space domain.

[0026] Furthermore, the specific process of step S4 is as follows: the DAS-VSP data only has the axial component of the elastic displacement field. When the wellbore is vertical or approximately vertical, it can be regarded as the vertical component of the elastic displacement field. According to the Helmholtz decomposition principle, the longitudinal wave is obtained using the following formula:

[0027]

[0028] in, is an inverse Fourier transform operator, and the convolution operator in the time-space domain obtained in step S3 is used to apply non-stationary phase filtering to the DAS-VSP axial displacement component to obtain the longitudinal wave data component;

[0029] For the shear wave component, the subtraction method is used to separate it:

[0030] us =u z -u p

[0031] Among them, u p is the scalar P wave field in the time and space domain after separation, u s is the scalar shear wave component.

[0032] Furthermore, the specific process of step S6 is as follows: on the common receiver gather, since the horizontal arrangement of the ground shot points can obtain horizontal wave number information, the dispersion relationship is used to calculate the vertical wave number of the local elastic wave field; at the same time, the time-space-varying equivalent apparent velocity information is used to construct a time-space domain convolution operator on the common receiver gather, and polarization projection filtering is applied in a non-stationary phase filtering manner to finally obtain a DAS-VSP seismic record with P- and S-wave separation, wherein the equivalent apparent velocity depends on the wave field propagation path, specifically, it is estimated based on the actual positions of the shot points and the receiver points and using a double square root moveout operator.

[0033] A distributed optical fiber seismic data longitudinal and shear wave separation device includes a data acquisition sensor and an input device respectively connected to a CPU processor, the CPU processor is also connected to a display, and the data acquisition sensor is used to collect VSP recording data;

[0034] The input device is used to store convolution operator data in the time and space domain;

[0035] The CPU processor is used to pre-process the VSP recorded data, convert the axial strain wave field and the displacement wave field according to the relationship between them, and perform the longitudinal and transverse wave separation process in combination with the convolution operator in the time and space domains;

[0036] The display is used to display the final longitudinal and shear wave separation results obtained by the CPU processor output.

[0037] Furthermore, the input device includes but is not limited to computer memory and hard disk.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] First, the present invention preprocesses distributed fiber-optic seismic data and converts it based on the relationship between the axial strain wavefield and the displacement wavefield to obtain an axial wavefield displacement vibration signal. It also estimates the local elastic wavefield polarization vector of the wellbore detector using the dispersion relationship based on the P-wave and S-wave logging velocities. Then, it performs a Fourier transform on the elastic wavefield polarization vector to obtain a time-space domain convolution filter operator. This convolution operator can thus be used to filter the distributed fiber-optic vertical seismic profile data in the time-space domain, performing P-wave and S-wave separation on the single-component distributed fiber-optic seismic data based on Helmholtz decomposition and polarization projection. To improve separation accuracy, corresponding polarization projection filter operators are calculated on the common shot point and common receiver point gathers, and the aforementioned polarization projection filtering is applied to obtain the final P-wave and S-wave separation results, ensuring the accuracy of the P-wave and S-wave separation results. Furthermore, the above P-wave and S-wave separation processes can be combined or repeated in any order, depending on the actual data situation. This enhances the flexibility of the P-wave and S-wave separation method of the present invention and overcomes the drawback that single-component seismic data is difficult to effectively separate P-wave and S-wave.

[0040] 2. Taking into account the different signal characteristics of the P-wave and S-wave wave fields on the common shot point gather and the common detection point gather, the present invention is designed to perform P-wave and S-wave separation on the common shot point and the common detection point gather respectively. On the one hand, the time-space domain convolution operator is used on the common shot point gather to perform time-space domain convolution on the axial displacement wave field vibration signal, and polarization projection filtering is implemented to obtain the P-wave and S-wave separation results in the shot domain. In this way, after one wave field separation in the common shot domain, the overall P-wave and S-wave wave field energy is better separated. On the other hand, the initial P-wave and S-wave separation results are re-sorted and transformed to the common detection point gather. The local wave field polarization direction of the common detection point gather is recalculated, and the wave field polarization direction that meets the characteristics of the common detection point gather is transformed to the time-space domain to obtain the corresponding convolution filter operator, and the convolution operator is used in the time-space domain to implement polarization projection filtering to further remove mode leakage, thereby obtaining more accurate P-wave and S-wave separation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Schematic diagram of the method flow of the present invention;

[0042] Figure 2 Schematic diagram of the device structure of the present invention;

[0043] Figure 3 Schematic diagram of the application process of the embodiment;

[0044] Figure 4a Schematic diagram of P-wave and S-wave separation of DAS-VSP data on common shot gathers;

[0045] Figure 4b Schematic diagram of P-wave and S-wave separation of DAS-VSP data on common receiver gathers;

[0046] Figure 5a is the elastic theory longitudinal wave velocity model used to synthesize DAS-VSP data in the embodiment;

[0047] Figure 5b is the elastic theory shear wave velocity model used to synthesize DAS-VSP data in the embodiment;

[0048] Figure 6 a~6f are the theoretically synthesized DAS-VSP data and the accurately separated P- and S-wave results in the examples;

[0049] Figure 7 a~7d are the results of P / S separation in the gun area in the embodiment;

[0050] Figure 8 a~8d are the P / S separation results of the co-detection point domain in the embodiment;

[0051] Figure 9 a~9b are actual DAS-VSP data of a region in the East China Sea in the examples;

[0052] Figure 10 a~10d are the actual P / S separation results of DAS-VSP data in a certain area of ​​the East China Sea in the embodiment;

[0053] Explanation of the marks in the figure: 1. Data acquisition sensor, 2. CPU processor, 3. Input device, 4. Display. DETAILED DESCRIPTION

[0054] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0055] Example

[0056] like Figure 1 As shown, a method for separating longitudinal and shear waves of distributed optical fiber seismic data includes the following steps:

[0057] S1. Obtain the VSP recorded data collected by the DAS sensor, then pre-process it and convert it according to the relationship between the axial strain wave field and the displacement wave field to obtain the axial wave field displacement vibration signal;

[0058] S2. Acquire acoustic logging velocity information and convert it into layer velocity within the seismic frequency band, thereby obtaining the P-wave and S-wave velocities at each geophone point in the wellbore and calculating the local wavefield polarization direction of each geophone.

[0059] S3. transforming the calculated wave field polarization direction into the time-space domain to obtain the convolution operator in the time-space domain;

[0060] S4. Using the convolution operator in the time and space domain, the axial wave field displacement vibration signal is convolved in the time and space domain, and the initial longitudinal and transverse wave separation results of the gun domain are obtained through polarization projection filtering;

[0061] S5. Re-sort the initial P-wave and S-wave separation results to obtain common detection point gathers;

[0062] S6. In the common detection point gather, according to the wave field reciprocity principle, the ground shot points are regarded as detection points. Based on the dispersion relationship between the emission slowness in the frequency wavenumber domain and the vertical and horizontal wavenumbers, the local wavefield polarization direction of the common detection point gather is calculated. Then, the process returns to execute steps S3 to S4 to obtain the final P-wave and S-wave separation results.

[0063] Based on the above method, a corresponding distributed optical fiber seismic data longitudinal and shear wave separation device is constructed, such as Figure 2 As shown, it includes a data acquisition sensor 1, an input device 3, and a display 4 respectively connected to a CPU processor 2, wherein the data acquisition sensor 1 is used to collect VSP recorded data;

[0064] The input device 3 is used to store the convolution operator data in the time and space domains. In practical applications, the input device 3 can be a computer memory, a hard disk, or other devices with data storage functions;

[0065] CPU processor 2 is used to pre-process the VSP recorded data, convert it according to the relationship between the axial strain wave field and the displacement wave field, and perform the longitudinal and transverse wave separation process in combination with the convolution operator in the time and space domains;

[0066] The display 4 is used to display the final longitudinal and transverse wave separation results obtained by the CPU processor 2.

[0067] This embodiment applies this technical solution, such as Figure 3 As shown, the main contents include:

[0068] 1) The vertical seismic profile (VSP) data collected by the distributed optical fiber (DAS) sensor are preprocessed according to the conventional seismic data processing process and converted according to the relationship between the axial strain wavefield and the displacement wavefield to obtain the axial wavefield displacement vibration signal with a high signal-to-noise ratio, which serves as the input data for the subsequent P-wave and S-wave separation operation.

[0069] 2) Based on the acoustic logging records, the interval velocity of the formation is converted using the acoustic time difference. Then, the high-frequency logging velocity is converted into interval velocity within the seismic frequency band with the help of Backus averaging, and the P-wave and S-wave velocities at each detection point in the wellbore are obtained.

[0070] 3) On the common shot gather, the P- and S-wave velocities at the detector point are regarded as the exit slowness of the elastic P- and S-wave fields. In the frequency-wavenumber domain, the local wavefield polarization direction of each detector is calculated based on the dispersion relationship between the exit slowness and the vertical and horizontal wavenumbers.

[0071] 4) The polarization direction in the frequency-wavenumber domain calculated in step 3) is transformed into the time-space domain to obtain a convolution operator that varies with time and space, and input it into the processor.

[0072] 5) Using the convolution operator in step 4), the DAS axial displacement wave field vibration signal in step 1) is convolved in the time and space domains, and polarization projection filtering is implemented to obtain the longitudinal and shear wave separation results of the gun domain.

[0073] 6) Re-sort the P-wave and S-wave separation results in step 5) and transform them into common detection point gathers.

[0074] 7) In the common receiver gather, according to the principle of wavefield reciprocity, the ground shot points can be regarded as receiver points. Based on the formation velocity in step 2), the dispersion relationship between the frequency-wavenumber domain exit slowness and the vertical and horizontal wavenumbers is used to recalculate the local wavefield polarization direction of the common receiver gather. Steps 3) to 5) are then repeated to further separate the P- and S-waves of the DAS-VSP records to obtain the final P- and S-wave separation results.

[0075] Specifically, in step 1), the VSP records collected by the distributed optical fiber sensor are preprocessed according to the conventional seismic data processing process to remove conventional noise and the unique coupling noise in DAS observations. Then, according to the relationship between the axial strain wave field and the displacement wave field, a high signal-to-noise ratio axial wave field displacement vibration signal is obtained. This is used as the input data for the subsequent steps. According to the Marmousi-II model ( Figure 5a and Figure 5b As shown in Figure 2, DAS-VSP data can be synthesized by forward modeling the elastic wave field, and the DAS-VSP responses corresponding to the reference longitudinal and transverse wave fields are generated by Helmholtz decomposition during wave field propagation. The above original wave field and the DAS-VSP responses corresponding to the longitudinal and transverse waves are converted into the axial displacement wave field, as shown in Figure 2. Figure 6 a~6f, where Figure 6 a~6c are the original common shot gathers of the 100th shot, which are DAS-VSP records and accurately separated P-wave and S-wave data records, respectively. Figure 6 d to 6f are the common detection point gathers of the 200th detector, which are the original DAS-VSP records and the accurately separated P-wave and S-wave data records respectively.

[0076] In step 2), the interval velocity of the formation is converted from the acoustic wave time difference according to the acoustic logging record, and the high-frequency logging velocity is converted into the interval velocity within the seismic frequency band by means of Backus averaging, thereby obtaining the P-wave and S-wave velocities at each detection point in the wellbore. In this embodiment, since the P-wave and S-wave velocities in the theoretical model are known (e.g. Figure 5a and 5b As shown in Figure 2), the P-wave and S-wave velocities of each detection point in the wellbore can be directly obtained.

[0077] In step 3), if Figure 4a As shown in the figure, on the common shot gather, the P-wave and S-wave velocities at the receiver point can be regarded as the exit slowness of the elastic P-wave and S-wave fields. Since the fiber optic sensors are usually arranged vertically or approximately vertically in DAS-VSP observations, the vertical wavenumber and frequency information recorded by the VSP can be obtained by two-dimensional fast Fourier transform. Then, according to the wave field propagation theory, the exit slowness and vertical wavenumber are used to calculate the horizontal wavenumber information of the P-wave and S-wave fields through the dispersion relation constraint, that is, k px With k sx Based on this, the Helmholtz decomposition theory is used to construct the polarization projection direction of the longitudinal and transverse wave fields at each detection point:

[0078]

[0079] in, represents the vector displacement wave field in the wavenumber domain, and represents the separated P-wave and S-wave vector displacement wavefield,

[0080] That is, in the frequency-wavenumber domain, the local wavefield polarization direction of each detector is calculated based on the dispersion relationship between the exit slowness and the vertical and horizontal wavenumbers. In this embodiment, the horizontal wavenumbers of the exiting P- and S-wave fields on the common shot gather are first calculated in the frequency-wavenumber domain based on the P- and S-wave velocities obtained in step 2). These are then combined with the existing vertical wavenumbers to construct the local P- and S-wave polarization directions.

[0081] In step 4), since the P- and S-wave velocities of the formations at different detector positions vary dramatically in the real earth medium, the slowness vector in the dispersion relation needs to be adjusted according to the spatial position. Then, according to the Fourier transform principle, the polarization projection filter operator in the frequency-wavenumber domain is equivalent to the convolution operator in the time-space domain. The frequency-wavenumber domain polarization filter of the P- and S-wave fields is transformed into the time-space domain to obtain the convolution operator in the time-space domain.

[0082] In step 5), the time-space convolution operator from step 4 is used to perform time-space convolution on the axial displacement wavefield vibration signal, and polarization projection filtering is applied to obtain the separation of longitudinal and transverse waves in the shot area. DAS-VSP data only contains the axial component of the elastic displacement field. When the wellbore is vertical or nearly vertical, this component can be considered the vertical component of the elastic displacement field. Based on the Helmholtz decomposition principle, the longitudinal wave can be approximately obtained using the following formula:

[0083]

[0084] in is the inverse Fourier transform operator, u p is the separated scalar P wave field in the time and space domain. Using the time and space domain convolution operator obtained in step 4), the DAS-VSP axial displacement component is subjected to non-stationary phase filtering to obtain the longitudinal wave data component. s ), the separation will be performed by subtraction:

[0085] u s =u z -u p

[0086] like Figure 7 As shown in a~7d, after a wave field separation in the common shot domain, the overall longitudinal and shear wave field energies are well separated, but there is still a small amount of mode leakage, among which, Figure 7 a and 7b are the P / S wave data of the common shot point of the 100th shot, Figure 7 c and 7d are the P / S wave data of the 200th common detection point, respectively. The arrows indicate that some mode leakage still exists.

[0087] In step 6), in order to further remove the mode leakage of the P-wave and S-wave field separation, the separation result in step 5) is re-sorted and transformed into a common detection point gather, and the P-wave and S-wave signal characteristics in the common detection point gather are used for separation.

[0088] In step 7), if Figure 4b As shown in , in the common detection point gathers, according to the wave field reciprocity principle, the shot points on the ground can be regarded as detection points. Based on the dispersion relationship between the emission slowness and the vertical and horizontal wave numbers in the frequency wavenumber domain, the local wavefield polarization direction of the common detection point gathers is recalculated, and the wavefield polarization direction that meets the characteristics of the common detection point gathers is transformed into the time and space domain to obtain the corresponding convolution filter operator. The convolution operator is used in the time and space domain to implement polarization projection filtering, further removing mode leakage and obtaining more accurate P-wave and S-wave separation results, as shown in . Figure 8 As shown in a~8d, Figure 8 a and 8b are the P / S wave data of the common shot point of the 100th shot, Figure 8Figures c and d are the P / S wave data at the 200th co-detection point, respectively. The results show that mode leakage is significantly improved after two P- and S-wave separations.

[0089] That is, on the common-receiver gathers, since the horizontal arrangement of the ground shot points can obtain horizontal wavenumber information, the dispersion relationship is used to calculate the vertical wavenumber of the local elastic wavefield. At the same time, the equivalent apparent velocity information that varies with time and space is used (the equivalent apparent velocity depends on the wavefield propagation path and can be estimated using the double square root moving average operator based on the actual positions of the shot points and the receiver points). A time-space domain convolution operator is constructed on the common-receiver gathers, and polarization projection filtering is applied in a non-stationary phase filtering manner to ultimately obtain DAS-VSP seismic records with separated P- and S-waves.

[0090] It should be noted that in the above steps 3) to 7), the separation operation of P-wave and S-wave is very flexible and can be performed in any combination or multi-stage repeated operation in any order according to the actual data situation, thereby realizing a flexible P-wave and S-wave field separation scheme, overcoming the defect that single-component seismic data is difficult to effectively separate P-wave and S-wave.

[0091] To verify the effectiveness of this technical solution, this embodiment conducts a P-wave and S-wave separation experiment on actual data from a certain area in the East China Sea. First, the acoustic time difference data obtained by actual acoustic logging is converted into P-wave and S-wave layer velocity data, and the data are smoothed using the Backus averaging method with half the seismic wavelength as the radius to obtain one-dimensional P-wave and S-wave layer velocities within the seismic frequency band.

[0092] Afterwards, the actual DAS-VSP data collected at sea are pre-processed with conventional denoising, bubble removal, and coupling noise removal, and the data are converted based on the relationship between the axial strain and the displacement strain, and the relationship between the axial strain wave field and the displacement wave field to obtain the axial wave field displacement vibration signal with a high signal-to-noise ratio, such as Figure 8 As indicated by the arrows in Figures 8a and 8b, there are obvious shear wave components in the actual pre-processed DAS-VSP data. It is very necessary to separate the P- and S-wave fields of the above data before subsequent velocity modeling and imaging.

[0093] Based on the above steps 3) to 7), the actual DAS-VSP is subjected to polarization projection filtering, and the P-wave and S-wave field separation is performed on the common shot point gather and the common receiver point gather in turn, and finally the separated DAS-VSP P-wave and S-wave data are obtained, as shown in the figure. Figure 10 a~10d, where Figure 10 a and 10b are the P and S wave data of the common shot point of the 80th shot, respectively. Figure 10 c and 10d are the P and S wave data of the 180th common detection point, respectively.

[0094] In summary, this technical solution first denoises distributed fiber-optic seismic data. Based on the P-wave and S-wave logging velocities, the dispersion relationship is used to estimate the local elastic wavefield polarization vector at the borehole detector. The elastic wavefield polarization vector is Fourier transformed to obtain a convolution filter operator in the time-space domain. The distributed fiber-optic vertical seismic profile data is then filtered using the convolution operator in the time-space domain. P-wave and S-wave separation is performed on the single-component distributed fiber-optic seismic data through polarization projection based on Helmholtz decomposition. To improve separation accuracy, the corresponding polarization projection filter operators are calculated on the common-shot and common-receiver gathers, and polarization projection filtering is applied to obtain the final P-wave and S-wave separation results. This results in a stable and reliable elastic wavefield separation scheme for DAS-VSP data, enabling efficient acquisition of accurate P / S wave data.

Claims

1. A method for separating longitudinal and shear waves of distributed optical fiber seismic data, characterized in that: The following steps are involved: S1. Obtain the VSP recorded data collected by the DAS sensor, then pre-process it and convert it according to the relationship between the axial strain wave field and the displacement wave field to obtain the axial wave field displacement vibration signal; S2. Acquire acoustic logging velocity information and convert it into layer velocity within the seismic frequency band, thereby obtaining the P-wave and S-wave velocities at each geophone point in the wellbore and calculating the local wavefield polarization direction of each geophone. S3. transforming the calculated wave field polarization direction into the time-space domain to obtain the convolution operator in the time-space domain; S4. Using the convolution operator in the time and space domain, the axial wave field displacement vibration signal is convolved in the time and space domain, and the initial longitudinal and transverse wave separation results of the gun domain are obtained through polarization projection filtering; S5. Re-sort the initial P-wave and S-wave separation results to obtain common detection point gathers; S6. In the common detection point gather, according to the wave field reciprocity principle, the ground shot points are regarded as detection points. Based on the dispersion relationship between the emission slowness in the frequency wavenumber domain and the vertical and horizontal wavenumbers, the local wavefield polarization direction of the common detection point gather is calculated. Then, the process returns to execute steps S3 to S4 to obtain the final P-wave and S-wave separation results.

2. A method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 1, characterized in that: The pre-processing process in step S1 includes but is not limited to denoising, removing bubbles, and removing coupling noise.

3. The method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 1, characterized in that: The step S2 specifically includes the following steps: S21. Calculate the interval velocity of the formation based on the acoustic time difference data collected by acoustic logging, then convert the high-frequency logging velocity into interval velocity within the seismic frequency band, and thus obtain the P-wave and S-wave velocities at each detection point in the wellbore; S22. On the common shot gather, the P- and S-wave velocities at the detector point are regarded as the exit slowness of the elastic P- and S-wave fields. In the frequency-wavenumber domain, the local wavefield polarization direction of each detector is calculated based on the dispersion relationship between the exit slowness and the vertical and horizontal wavenumbers.

4. A method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 3, characterized in that: The step S21 specifically adopts the Backus averaging method to smooth the converted layer velocity of the formation, thereby obtaining the P-wave and S-wave layer velocity within the seismic frequency band.

5. The method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 3, characterized in that: The specific process of step S22 is as follows: the longitudinal and transverse wave velocities at the detection point in the common shot gather are regarded as the exit slowness of the elastic wave field. Since the optical fiber sensors are arranged vertically or approximately vertically in the DAS-VSP observation, the vertical wave number and frequency information recorded by the VSP are obtained by two-dimensional fast Fourier transform. Then, according to the wave field propagation theory, the exit slowness and vertical wave number are used to calculate the horizontal wave number information of the longitudinal and transverse wave fields through the dispersion relation constraint, that is, k px With k sx Based on this, the Helmholtz decomposition theory is used to construct the polarization projection direction of the longitudinal and transverse wave fields at each detection point: in, represents the vector displacement wave field in the wavenumber domain, and Represent the separated P-wave and S-wave vector displacement wave fields respectively.

6. A method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 5, characterized in that: The specific process of step S3 is as follows: since the P- and S-wave velocities of the formations at different detector positions vary dramatically in the real earth medium, the slowness vector in the dispersion relation needs to be adjusted according to the spatial position. Then, according to the Fourier transform principle, the polarization projection filter operator in the frequency-wavenumber domain is equivalent to the convolution operator in the time-space domain. Thus, the frequency-wavenumber domain polarization filter of the P- and S-wave fields is transformed into the time-space domain to obtain the convolution operator in the time-space domain.

7. A method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 6, characterized in that: The specific process of step S4 is as follows: DAS-VSP data only has the axial component of the elastic displacement field. When the wellbore is vertical or approximately vertical, it can be regarded as the vertical component of the elastic displacement field. According to the Helmholtz decomposition principle, the longitudinal wave is obtained using the following formula: in, is an inverse Fourier transform operator, and the convolution operator in the time-space domain obtained in step S3 is used to apply non-stationary phase filtering to the DAS-VSP axial displacement component to obtain the longitudinal wave data component; For the shear wave component, the subtraction method is used to separate it: in s =in zp Among them, u p is the scalar P wave field in the time and space domain after separation, u s is the scalar shear wave component.

8. The method for separating longitudinal and shear waves of distributed optical fiber seismic data according to claim 1, characterized in that: The specific process of step S6 is as follows: on the common receiver gather, since the horizontal arrangement of the ground shot points can obtain horizontal wave number information, the dispersion relationship is used to calculate the vertical wave number of the local elastic wave field; at the same time, the time-space-varying equivalent apparent velocity information is used to construct a time-space domain convolution operator on the common receiver gather, and polarization projection filtering is applied in a non-stationary phase filtering manner to finally obtain a DAS-VSP seismic record with separated longitudinal and transverse waves. The equivalent apparent velocity depends on the wave field propagation path, specifically, it is estimated based on the actual positions of the shot points and the receiver points using a double square root moveout operator.

9. A distributed optical fiber seismic data longitudinal and shear wave separation device using the distributed optical fiber seismic data longitudinal and shear wave separation method according to claim 1, characterized in that: The device comprises a data acquisition sensor (1) and an input device (3) respectively connected to a CPU processor (2). The CPU processor (2) is also connected to a display (4). The data acquisition sensor (1) is used to collect VSP recorded data. The input device (3) is used to store convolution operator data in the time-space domain; The CPU processor (2) is used to pre-process the VSP recorded data, convert the data according to the relationship between the axial strain wave field and the displacement wave field, and perform a longitudinal and transverse wave separation process in combination with a convolution operator in the time and space domain; The display (4) is used to display the final longitudinal and transverse wave separation results output by the CPU processor (2).

10. The distributed optical fiber seismic data longitudinal and shear wave separation device according to claim 9, characterized in that: The input device (3) includes but is not limited to computer memory and hard disk.

Images