Seismic wave field separation method based on polarization

Abstract

The invention relates to the technical field of geoscience seismic wave field separation, and discloses a polarization-based seismic wave field separation method, which comprises the following steps of: acquiring three-component seismic data by utilizing forward modeling; preprocessing seismic data obtained by forward modeling to obtain sgy or binary data; carrying out polarization analysis processing on the obtained sgy or binary data to obtain three parameter values, and selecting a threshold value; a selected threshold value is substituted into the whole wave field for calculation, and the wave field is separated into linear body wave and nonlinear surface wave components; according to the seismic wave field separation method based on polarization, seismic records are subjected to wave field separation by utilizing the polarization characteristic of a wave field and adopting a polarization analysis method, a selected threshold value is substituted into the whole wave field for calculation, seismic signals of the whole wave field are separated into relatively pure body waves and surface waves, and then frequency dispersion energy is further calculated. And frequency dispersion energy spectrums under different wave field conditions are obtained, and a clearer frequency dispersion curve is extracted.

Description

Technical Field

[0001] The invention relates to the technical field of geology seismic wave field separation, and in particular to a seismic wave field separation method based on polarization. Background Art

[0002] Seismic wavefield separation is an important technology in seismic data processing. It aims to separate useful wavefield information from complex seismic data, such as P waves, S waves, surface waves and other types of wavefields. This is of great significance for subsequent seismic data interpretation, such as velocity analysis, seismic imaging, lithology analysis, etc. In recent years, in-depth research has been conducted on wavefield separation technology at home and abroad, and the corresponding methods have been widely used in wave-type seismic data processing processes, such as elastic wave reverse time migration, elastic wave least squares migration, elastic wave waveform inversion, etc.

[0003] Common seismic wave field separation methods are mainly divided into the following categories:

[0004] 1. Frequency-based separation method, frequency-wavenumber domain analysis (FK analysis): convert seismic data into the frequency-wavenumber domain and separate them according to the frequency and wavenumber characteristics of different wave fields; Hilbert-Huang transform (HHT): suitable for seismic wave field separation of nonlinear and non-stationary signals.

[0005] 2. Time-based separation method, first arrival wave picking: by setting the threshold value, automatically or manually picking the P wave arrival time, and then separating the P wave; time-offset domain analysis: using the offset technology to convert the common shot point data to the offset domain to separate different wave fields.

[0006] 3. Attribute-based separation method, coherence volume technology: separate different wave fields by calculating the coherence attributes of seismic data; cluster analysis: use the clustering characteristics of seismic attributes to divide the wave fields into different categories.

[0007] 4. Separation method based on waveform difference: waveform fitting method: separate the target wave field by fitting the seismic waveform; pattern recognition method: use machine learning method to classify according to the characteristic differences of waveform.

[0008] 5. Separation method based on wave equation, reverse time migration (RTM): accurate separation of wave field is achieved through the reverse time migration process of wave equation; wave field decomposition technology: decompose the wave field into different types of waves, such as P wave and S wave, according to the wave equation.

[0009] 6. Artificial intelligence-based methods, deep learning: using deep learning techniques such as convolutional neural networks (CNN) to perform wave field identification and separation.

[0010] Polarization analysis is a spatial filtering signal processing method based on the polarization characteristics of waves. It simplifies information extraction by measuring the polarization properties of various waves (polarization degree, polarization direction, etc.). In seismic signal processing, polarization analysis is a method of obtaining underground medium information by analyzing the polarization characteristics of seismic waveforms. Polarization analysis is a technology used in seismic signal processing that can help us understand the physical characteristics of underground media.

[0011] With the development of science and technology in recent years, polarization analysis technology has also developed in a diversified and complex direction. At present, polarization analysis methods have achieved some research results in the fields of electromagnetism, optics and seismology. In the field of geophysics, polarization analysis methods are mainly used to identify and separate specific wave types, suppress various types of noise, study shear wave splitting, and identify multi-wave seismic phases and determine wave arrival times.

[0012] Although the existing seismic wave field separation methods can achieve the separation of body waves and surface waves to a certain extent, they still have some shortcomings: strong dependence on initial models and parameters, complex calculation process, and high requirements on computing resources; affected by the complexity of seismic wave fields and the interference of seismic multi-wave fields, the wave field separation effect is not good. Summary of the invention

[0013] In order to solve the problems that the above existing seismic wave field separation methods have strong dependence on initial models and parameters, complex calculation process, high requirements on computing resources, and poor wave field separation effect due to the complexity of seismic wave fields and interference of seismic multi-wave fields, the present invention is implemented through the following technical solutions: A seismic wave field separation method based on polarization, including the following methods:

[0014] S1. Obtain three-component seismic data using forward modeling;

[0015] S2, preprocessing the seismic data obtained by the forward simulation to obtain sgy or binary data;

[0016] S3, performing polarization analysis on the obtained sgy or binary data to obtain three parameter values ​​and select thresholds;

[0017] S4, using the selected threshold to bring the entire wave field into calculation, and separating the wave field into linear body wave and nonlinear surface wave components;

[0018] S5, calculating the dispersion energy spectrum of sgy or binary data under different thresholds;

[0019] S6. Extracting a dispersion curve according to the distribution of dispersion energy.

[0020] Furthermore, the seismic data are seismic records obtained in a laboratory environment, specifically:

[0021] In a laboratory environment, given the formation model parameters, numerical simulation is performed to obtain seismic records as seismic data.

[0022] Furthermore, the forward modeling is used to obtain seismic single shot records in sgy format or binary data with the head track removed in a laboratory environment.

[0023] Furthermore, the FK method is used in S5 to calculate the dispersion energy spectrum of sgy or binary data under different thresholds, where F is the frequency and K is the wave number, specifically:

[0024] S51. Data preprocessing

[0025] S511. Resample the seismic data to ensure the uniformity of the time series;

[0026] S512, applying a bandpass filter to retain signals within a target frequency range;

[0027] S513, remove instrument response and noise;

[0028] S52, FK transformation

[0029] S521, performing Fourier transform on the data of each time window to obtain frequency domain data;

[0030] S522, performing wave number transformation on the frequency domain data to obtain FK spectrum, that is, dispersion energy spectrum.

[0031] Furthermore, the extraction of the dispersion curve is specifically as follows:

[0032] S61, dispersion curve extraction

[0033] S611. In the FK spectrum, energy clusters are identified, with the center points corresponding to the phase velocity and group velocity;

[0034] S612. Track the dispersion curve by analyzing the position changes of energy clusters at different frequencies;

[0035] S62. Optimization and verification of dispersion curves

[0036] S621. Smoothing the extracted frequency dispersion points to eliminate random errors.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The polarization-based seismic wave field separation method utilizes the polarization characteristics of the wave field and adopts the polarization analysis method to separate the wave field of the seismic record, uses the selected threshold to bring the entire wave field into calculation, separates the seismic signal of the entire wave field into relatively pure body waves and surface waves, and then further calculates the dispersion energy to obtain the dispersion energy spectrum under different wave field conditions, and extracts clearer dispersion curves. This method is based on the polarization characteristics, the calculation is faster, and no data annotation is required, and a polarization-based seismic wave field separation space vector end change diagram can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of a three-component single shot recording of a single-layer model in an embodiment of the present invention;

[0040] Figure 2 The three parameter value diagram and the vector diagram corresponding to the three component records in the embodiment of the present invention;

[0041] Figure 3 Schematic diagram of three-component waveforms before and after polarization analysis in an embodiment of the present invention;

[0042] Figure 4 This is a comparison diagram of single shot records before and after polarization analysis in an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of a single shot record of a multi-layer model numerical simulation in an embodiment of the present invention;

[0044] Figure 6 A diagram showing three parameter values ​​corresponding to three component records in an embodiment of the present invention;

[0045] Figure 7 : is a dispersion energy diagram corresponding to different thresholds in an embodiment of the present invention;

[0046] Figure 8 The figure is a flow chart of the seismic wave field separation method of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] See also Figure 1-Figure 8 , an embodiment of the polarization-based seismic wave field separation method is as follows:

[0049] A polarization-based seismic wave field separation method, comprising the following method:

[0050] S1. Using forward modeling to obtain three-component seismic data, in a laboratory environment, given formation model parameters, the formation model parameters include shear wave velocity, longitudinal wave velocity, grid size, sampling rate, number of channels, number of sample points, number of shots, numerical simulation is performed to obtain seismic records as seismic data;

[0051] S2. Preprocess the seismic data obtained by the forward modeling to obtain sgy or binary data, and use the forward modeling to obtain seismic single shot records in sgy format or binary data without header files in a laboratory environment;

[0052] S3, performing polarization analysis on the obtained sgy or binary data to obtain three parameter values, which are azimuth, incident angle and ellipticity, using the threshold formula to select different thresholds for actual wave field analysis, and selecting a suitable threshold according to the result;

[0053] S4, using quality control steps such as threshold and power spectrum density to separate the wave field into linear body wave and nonlinear surface wave components;

[0054] The selected threshold is used to bring the entire wave field into calculation. The body wave component of the single shot record greater than the threshold occupies the main position, and the surface wave component less than the threshold occupies the main position. Therefore, the wave field can be well separated into linear body wave and nonlinear surface wave components.

[0055] S5. Using the FK method, calculate the dispersion energy spectrum of sgy or binary single shot records under different thresholds, where F is the frequency and K is the wave number;

[0056] Specifically:

[0057] Step 1: Data preprocessing

[0058] Resample the seismic data to ensure the uniformity of the time series;

[0059] Apply a bandpass filter to retain signals within the target frequency range;

[0060] Remove instrument response and noise;

[0061] Step 2: FK transformation

[0062] Perform Fourier transform on the data of each time window to obtain frequency domain data;

[0063] Perform wave number transformation on the frequency domain data to obtain the FK spectrum, i.e., the dispersion energy spectrum;

[0064] S6. extracting a dispersion curve according to the distribution of dispersion energy;

[0065] Dispersion curve extraction

[0066] In the FK spectrum, energy clusters are identified, with the central points corresponding to the phase velocity and group velocity;

[0067] By analyzing the position changes of energy clusters at different frequencies, the dispersion curve is tracked;

[0068] Optimization and verification of dispersion curves

[0069] The extracted frequency dispersion points are smoothed to eliminate random errors.

[0070] In order to study the feasibility of polarization analysis in seismic wave field separation, two numerical simulations were implemented in a laboratory environment to verify the feasibility of polarization analysis in wave field separation, and the results before and after processing were compared for analysis. First, a single-layer model simulation was performed, and the main parameters are shown in Table 1:

[0071] Table 1

[0072] Serial number Parameter name Parameter Value 1 Grid size x×y×z (450×3)×(500×3)×(450×3)m 2 Number of channels × number of sample points 150×800 3 Sampling rate 0.5ms 4 Shear wave velocity Vs 1150m / s 5 P-wave velocity Vp 2000m / s 6 Number of guns 1

[0073] Through forward numerical simulation based on finite difference, the seismic single shot records of each component are obtained as follows: Figure 1 As shown, Figure 1 (a), (b), and (c) are the single shot records of the X component, Y component, and Z component, respectively. It can be seen that the direct P wave and direct S wave are clear.

[0074] The sgy or binary single shot records (in Figure 1 The first one in (a) is shown as an example. Figure 2 (a)) to perform polarization analysis and obtain a three-parameter value diagram (azimuth, incident angle, ellipticity, as shown in Figure 2 (b)) and the vector end diagram of the three directional eigenvalues, as shown in Figure 2 As shown in (c).

[0075] Select the threshold R = 0.2 and obtain Figure 3 The numerical simulation waveform shown is a comparison of the waveform before and after polarization, where: Figure 3 (a) is the original numerical simulation three-component waveform. Figure 3 (b) is the waveform after polarization analysis. Figure 3 It can be seen from (a) and (b) that when the threshold R=0.2 is selected, the two wave fields can be well separated.

[0076] like Figure 4 As shown, Figure 4 (a), (b), and (c) are the three-component waveforms before polarization processing. Figure 4(d), (e), and (f) are the three-component waveforms after polarization processing. It can be seen that when the threshold R is 0.2, the two wave fields can be well separated.

[0077] The above are the polarization analysis results of the single-layer model. In order to verify the universality of the method, the numerical simulation of the multi-layer model is continued, and the polarization analysis is performed according to the above steps.

[0078] The main parameters of the multi-layer model simulation are shown in Table 2:

[0079] Table 2

[0080] Serial number Parameter name Parameter Value 1 Grid size x×y×z 200m×500m×200m 2 Number of channels × number of sample points 100×1000 3 Sampling rate 1ms 4 Vs1 289m / s 5 Vs2 462m / s 6 Vs3 578m / s 7 Vp1 500m / s 8 Vp2 800m / s 9 Vp3 1000m / s 10 Number of guns 1

[0081] Among them, Vs1, Vs2, and Vs3 represent the shear wave velocities of the 1st, 2nd, and 3rd layers respectively, and Vp1, Vp2, and Vp3 represent the longitudinal wave velocities of the 1st, 2nd, and 3rd layers respectively.

[0082] like Figure 5 As shown, Figure 5 (a) is a multi-layer model. Figure 5 (b), (c), and (d) are the single shot records corresponding to the numerical simulation of the multi-layer model.

[0083] like Figure 6 As shown, it can be seen that under the multi-layer model, the three parameters begin to become complex and show obvious fluctuations. Among them, the incident angle changes the least, the azimuth angle is second, and the ellipticity is the largest. Figure 2 , Figure 3 , Figure 6 In the diagram, N is north, E is east, and Z is perpendicular to N and E.

[0084] In order to obtain the best imaging effect, the threshold range was set to 0.05-0.5, and the corresponding dispersion energy map was obtained, such as Figure 7 As shown. It can be seen that various wave fields are mixed in the original dispersion, and no dispersion characteristics can be seen. From 0.05-0.08, the dispersion energy begins to increase continuously, and the continuity of the dispersion curve is also increasing. From 0.08-0.30, the dispersion energy is slightly weakened, but the high-order energy is still clearly visible, the continuity of the fundamental dispersion curve is good, and the continuity of the high-order dispersion curve is reduced, but the reduction is not large.

[0085] 0.30-0.50, the dispersion energy is significantly weakened, and the continuity of the fundamental and high-order decreases rapidly. The high-order decreases more and the continuity becomes worse. When the threshold is 0.50, the fundamental-order energy is visible, but the high-order energy is invisible, and a large amount of body wave interference occurs.

[0086] Through overall analysis and comparison, it can be seen that under the single-layer model, by selecting a suitable threshold, the polarization-based seismic wave field separation method can well separate P waves and S waves; under the complex wave field of the multi-layer model, selecting a suitable threshold can not only separate body waves and surface waves, but also enhance the dispersion energy and continuity of dispersion curves for high-order surface wave components that are difficult to extract, which is of great significance for subsequent high-order surface wave imaging research.

[0087] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A polarization-based seismic wave field separation method, characterized in that: The following methods are included: S1. Obtain three-component seismic data using forward modeling; S2, preprocessing the seismic data obtained by the forward simulation to obtain sgy or binary data; S3, performing polarization analysis on the obtained sgy or binary data to obtain three parameter values ​​and select thresholds; S4, using the selected threshold to bring the entire wave field into calculation, and separating the wave field into linear body wave and nonlinear surface wave components; S5, calculating the dispersion energy spectrum of sgy or binary data under different thresholds; S6. Extracting a dispersion curve according to the distribution of dispersion energy.

2. The polarization-based seismic wave field separation method according to claim 1, characterized in that: The seismic data are seismic records obtained in a laboratory environment, specifically: In a laboratory environment, given the formation model parameters, numerical simulation is performed to obtain seismic records as seismic data.

3. The polarization-based seismic wave field separation method according to claim 2, characterized in that: The forward modeling is used to obtain seismic single shot records in sgy format or binary data with the head track removed in a laboratory environment.

4. The polarization-based seismic wave field separation method according to claim 1, characterized in that: In S5, the FK method is used to calculate the dispersion energy spectrum of sgy or binary data under different thresholds, where F is the frequency and K is the wave number, specifically: S51. Data preprocessing S511. Resample the seismic data to ensure the uniformity of the time series; S512, applying a bandpass filter to retain signals within a target frequency range; S513, remove instrument response and noise; S52, FK transformation S521, performing Fourier transform on the data of each time window to obtain frequency domain data; S522, performing wave number transformation on the frequency domain data to obtain FK spectrum, that is, dispersion energy spectrum.

5. The polarization-based seismic wave field separation method according to claim 4, characterized in that: The extraction of the dispersion curve is specifically as follows: S61, dispersion curve extraction S611. In the FK spectrum, energy clusters are identified, with the center points corresponding to the phase velocity and group velocity; S612. Track the dispersion curve by analyzing the position changes of energy clusters at different frequencies; S62. Optimization and verification of dispersion curves S621. Smoothing the extracted frequency dispersion points to eliminate random errors.

Images