A pre-stack amplitude compensation method suitable for loess plateau region

Abstract

The present application belongs to the field of geophysical exploration technology, and relates to a prestack amplitude compensation method suitable for loess tableland area, which comprises the following steps: denoising the seismic data, stacking the denoised data, partitioning or not partitioning the data according to the quality of the stacked profile, performing spectrum analysis on the stacked profile of each partitioned data, dividing the frequency band according to the spectrum analysis result of the stacked profile of each partitioned data, determining the low-frequency compensation factor of each partition, calculating the compensation factor increment of each frequency segment of the high frequency of each partition, performing frequency division amplitude compensation on the data of each partitioned region and frequency division, and reconstructing the data. The present application is convenient for effectively widening the frequency band and improving the resolution through deconvolution.

Description

Technical Field

[0001] This invention belongs to the field of geophysical exploration technology and relates to seismic data processing technology. Specifically, it is a pre-stack amplitude compensation method applicable to the Loess Plateau region. Background Technology

[0002] The Ordos Basin's Loess Plateau region is largely covered by thick Quaternary loess. The loose and dry nature of this loess severely attenuates seismic waves, resulting in narrow bandwidth and low signal-to-noise ratio (SNR) in seismic exploration data. Improving the frequency response of seismic data in the Loess Plateau region primarily relies on pre-stack deconvolution. Li Qingzhong, in his article "The Relationship Between Seismic Exploration Resolution and SNR Spectrum," points out that "if the SNR spectrum is ideal, i.e., the effective bandwidth (i.e., the frequency components with an SNR greater than 1) is wide, then the apparent period of the wavelet can be compressed into a near-narrow pulse through deconvolution, thus improving resolution." However, currently, the low SNR and narrow bandwidth of the original data from the Ordos Loess Plateau region makes it difficult to obtain good deconvolution results (see attached). Figure 1 Therefore, it is necessary to perform pre-stack amplitude compensation on the data to increase the energy of effective high-frequency reflection and broaden the frequency band.

[0003] Conventional amplitude compensation methods utilize full-band single-shot analysis for quality control of compensation effectiveness. However, full-band single-shot analysis primarily reflects the energy changes in the time domain of low-frequency information, neglecting the characteristic that loess absorbs and attenuates high-frequency components more strongly than low-frequency components, and that high-frequency attenuation is much faster than low-frequency attenuation. This often results in good low-frequency compensation but severely insufficient high-frequency compensation (as shown in the attached figure). Figure 1 b. (Insufficient energy compensation occurs in the deep layers of the profile above 25Hz), making it difficult to broaden the effective frequency band and resulting in unsatisfactory deconvolution effects. Therefore, amplitude processing with frequency division is required.

[0004] Currently, publicly available literature and reports on amplitude compensation methods related to time-frequency-amplitude correlation mainly include inverse Q filtering, time-frequency analysis, and inverse Q-shifting. However, these methods have not yielded ideal results in the Loess Plateau region because:

[0005] Inverse Q filtering and inverse Q migration require a high-precision Q value. Currently, in the Ordos Basin, an attempt is being made to establish the VQ field (velocity and Q value) by combining surface Q (point: dual-well micrologging, surface: tomographic inversion velocity field) with deeper Q (superimposed velocity field constrained by well data). However, the construction cost of ultra-deep dual-well micrologging is high, and due to the influence of acquisition technology (coupling conditions between receiving equipment and wellbore), some micrologging data have poor first-arrival morphology and cannot be used for micrologging interpretation. Furthermore, micrologging interpretation is highly ambiguous, and the interpreted velocity curve will affect the accuracy of the Q value. At the same time, the accuracy of the near-surface velocity field obtained by single-shot first-arrival tomographic static correction inversion is insufficient, all of which will affect the accuracy of the Q value in the work area.

[0006] Time-frequency analysis methods, particularly the commonly used frequency-division amplitude compensation, typically employ a cosine function to construct a frequency-division filter function in the frequency domain to divide the data into frequencies. Then, polynomial fitting is performed on the absorption attenuation function of each frequency band to reconstruct the data and obtain the final compensation output. However, this method uses equal-interval frequency division, which may result in the grouping of low-frequency bands with less absorption and attenuation with high-frequency bands with strong absorption and attenuation, leading to statistical inaccuracies. Furthermore, the automatic extraction of compensation factors for the frequency-division data is problematic, as random noise can introduce inaccuracies into the calculation of compensation factors for low signal-to-noise ratio data from the Loess Plateau. Spectral whitening can also be performed using frequency-amplitude processing, but its main method involves time-varying gain processing of the desired frequency band while suppressing other bands, resulting in poor fidelity and amplitude preservation. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention aims to provide a pre-stack amplitude compensation method suitable for the Loess Plateau region, thereby effectively broadening the frequency band and improving resolution through deconvolution.

[0008] To achieve the above objectives, the technical solution adopted by this invention is as follows: A pre-stack amplitude compensation method suitable for the Loess Plateau region, comprising the following steps:

[0009] Step 1: Denoise the seismic data;

[0010] Step 2: Overlay the denoised data;

[0011] Step 3: Partition the data or leave it unpartitioned based on the quality of the overlay profile;

[0012] Step 4: Perform spectral analysis on the overlay profile of data from each partition;

[0013] Step 5: Divide the frequency bands based on the results of the overlay profile spectral analysis of the data from each partition;

[0014] Step 6: Determine the low-frequency compensation factor for each zone;

[0015] Step 7: Calculate the compensation factor increment for each high-frequency sub-band in each partition;

[0016] Step 8: Perform amplitude compensation for each sub-region and frequency-divided data;

[0017] Step 9: Data Reconstruction.

[0018] As a limitation of the present invention, in step 3: based on the resolution of the superimposed profile and the three-dimensional data, combined with the attribute map and near-surface structure map extracted from the profile, the data is partitioned.

[0019] As a further limitation of the present invention, in step 4:

[0020] The analysis time window is fixed at (t2-t1), and (t2-t1)>300Ms;

[0021] Using (t2-t1) as a propagation period, and employing Formula 1, the value of the attenuation coefficient COEF(t2-t1) over the period of (t2-t1) is calculated.

[0022]

[0023] In the formula, n2 is the compensation factor at time t2, and n1 is the compensation factor at time t1.

[0024] As a further limitation of the present invention, in step 5: after performing spectrum analysis in a fixed time window (t2-t1), the low-frequency range in which the energy in the spectrum is in a state of maintenance is taken as the standard for no absorption attenuation, and the segment from which the energy begins to decrease is taken as the segment for which compensation is to be added.

[0025] While performing amplitude compensation, compensation control is applied to the high-frequency band. The high-pass filter scan of the superimposed profile is used to estimate the maximum effective frequency at which energy recovery can be achieved, and no compensation is performed for frequency bands higher than this frequency.

[0026] As a further limitation of the present invention, in step 5:

[0027] When dividing the frequency band, each fixed decrease in decibel β on the spectrum is divided into a frequency band;

[0028] Calculate β according to Formula 2:

[0029]

[0030] A1 and A2 are the amplitude values ​​of adjacent wave peaks, and their calculation formula is as follows:

[0031] A2 = A1·e -αr ·ω(t)————————————————Formula 3

[0032] In the formula, r is the propagation distance, α is the absorption attenuation coefficient, and ω(t) is the wave function; if it is a spherical wave, multiply it by 1 / r before A1.

[0033] As a further limitation of the present invention, β = 10 dB is used in the Loess Mountain area.

[0034] As a further limitation of the present invention, in step 6: the low-frequency part separated as the compensation standard is tested by using absorption attenuation compensation formula 1 and combining the energy curve of profile or single shot compensation to determine the low-frequency basic compensation factor n0.

[0035] As a further limitation of the present invention, in step 7: (n1-n0) = 0.6, (n2-n1) = 0.6, (n3-n2) = 0.6.

[0036] As another limitation of the present invention, the formulas for steps 5 to 9 are as follows:

[0037] Xi(t)=x(t)×Hi(t) ——————————————— Formula 4

[0038]

[0039]

[0040] Hi(t) is the filtering operator of the i-th filter, x(t) is the original record, Xi(t) is the frequency-divided data, and w i , where is the weighting factor, Zi(t) is the compensated frequency-divided data, COEF(Xi(t)) is the expression for the absorption attenuation coefficient of the frequency-divided data amplitude as a function of time, t is the time window size for spectral analysis, nxi is the absorption attenuation compensation parameter corresponding to each frequency-divided data, and Y(t) is the reconstructed record after compensation.

[0041] By adopting the above-described technical solution, the beneficial effects achieved by this invention compared to the prior art are as follows:

[0042] (1) This invention addresses the characteristics of rapid lateral variation and severe absorption and attenuation of high-frequency components in seismic data in the Loess Plateau region (see attached). Figure 3 This paper proposes a pre-stack amplitude compensation technique based on the difference in absorption attenuation between high and low frequencies. This technique is tailored to the characteristic that the absorption attenuation at low frequencies is relatively small (as shown in the attached diagram). Figure 3 A good energy compensation effect can be obtained by simply compensating for low frequencies in the time domain (as shown in the attached figure). Figure 2 (Low-frequency compensation effect below 10Hz) can be achieved by using low frequencies as a model standard without near-surface absorption and attenuation. High-frequency components are compensated for energy in the direction of this standard, which can solve the problem of near-surface absorption and attenuation of high frequencies. This effectively broadens the frequency band while maintaining fidelity and amplitude, and improves resolution through deconvolution.

[0043] (2) This invention can effectively compensate for the amplitude of different frequency bands of Loess Plateau data before stacking, with high fidelity and good amplitude preservation. It can effectively broaden the frequency band and increase the energy of high-frequency reflections. Through deconvolution and frequency enhancement, a high-resolution seismic profile is finally obtained (as shown in the appendix). Figure 8 (9,10); It can be time-varying and space-varying, and is suitable for the characteristics of rapid quality changes of single shots on the Loess Plateau. By utilizing the difference in absorption and attenuation between high and low frequencies, it can realize the calculation of compensation parameters and quality control; It has the advantages of low cost, simple operation and short calculation time.

[0044] (3) This invention utilizes the difference in absorption and attenuation characteristics between high and low frequencies in the Loess Plateau region. Taking advantage of the smaller absorption and attenuation at low frequencies, spectral analysis can be performed using the denoised and superimposed data profile to determine the relative attenuation in decibels of high frequencies compared to low frequencies in different frequency bands. This value can be converted to an absorption and attenuation compensation factor, thus determining the increase in compensation for high frequencies compared to low frequencies, achieving the purpose of frequency band compensation. It fully utilizes its own data to achieve one-to-one compensation. Attached Figure Description

[0045] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0046] Figure 1 This is a superimposed profile of different deconvolution data for a certain survey line segment in the Loess Mountain area. Figure 1 a represents the superimposed profile after conventional amplitude compensation. Figure 1 b represents the overlay profile of the predicted deconvolution data at a prediction step size of 24. Figure 1 c represents the overlay profile of the predicted deconvolution data when the prediction step size is 16. Figure 1 d represents the overlay profile of the predicted deconvolution data when the prediction step size is 8;

[0047] Figure 2 This is a map showing the current status of seismic data compensation in the Loess Mountain area. Figure 2 a is a superimposed profile (broadband profile) of conventional compensation plus surface uniform amplitude compensation; Figure 2 b is the superimposed profile of conventional compensation plus surface uniform amplitude compensation (Lp < 10Hz); Figure 2 c represents the superimposed profile of conventional compensation plus surface uniform amplitude compensation (Hp > 25Hz); Figure 2 d represents the superimposed profile of conventional compensation plus surface uniform amplitude compensation (Hp > 35Hz); Figure 2 e represents the superimposed profile of conventional compensation plus surface uniform amplitude compensation (Hp > 45Hz); Figure 2 f is the superimposed profile of conventional compensation plus surface uniform amplitude compensation (Hp > 55Hz); Figure 2 g represents the frequency division and energy curve (broadband profile) of conventional compensation plus surface uniform amplitude compensation; Figure 2 h represents the frequency division and energy curve of conventional compensation plus surface uniform amplitude compensation (Lp < 10 Hz); Figure 2 i represents the frequency division and energy curve of conventional compensation plus surface uniform amplitude compensation (Hp > 25Hz); Figure 2 j represents the frequency division and energy curve of conventional compensation plus surface uniform amplitude compensation (Hp > 35Hz); Figure 2 k represents the frequency division and energy curve of conventional compensation plus surface uniform amplitude compensation (Hp > 45Hz); Figure 2 l represents the frequency division and energy curve of conventional compensation plus surface uniform amplitude compensation (Hp > 55Hz);

[0048] Figure 3 To illustrate the differences in high- and low-frequency absorption attenuation in seismic data from the Loess Mountain region, among which... Figure 3 a represents the original data (Lp < 14Hz). Figure 3 b represents the original data (Bp: 25-40Hz). Figure 3 c represents the original data (Bp: 45-60Hz). Figure 3 d represents the spectrum analysis of the original data. Figure 3 The area circled in section d represents the segment where energy remains stable and absorption attenuation is minimal.

[0049] Figure 4 This is the micro-logging record of an ultra-deep dual-well system in the Ordos Basin, as described in Embodiment 1 of the present invention. Figure 4 a represents the micrologging record of the ultra-deep dual-well system – the record is generated at the bottom of one well and received in the other well. Figure 4 b is the spectral analysis diagram of micrologging in ultra-deep dual wells.

[0050] Figure 5 This is a schematic diagram of data partitioning for a three-dimensional data set in the Loess Hilly region of the Ordos Basin, according to Embodiment 1 of the present invention. Figure 5 a is a schematic diagram of the cross-sectional partition. Figure 5 b is the spectral analysis diagram of section 1. Figure 5 c represents the spectral analysis diagram of section 2. Figure 5 d is a three-dimensional thickness map of a low-velocity layer in the Ordos Basin;

[0051] Figure 6 This is a spectral analysis diagram of three-dimensional data from the Ordos Loess Mountain region in Embodiment 1 of the present invention, wherein... Figure 6 a is the cross-sectional spectral analysis diagram. Figure 6 b is the high-pass filtered scan analysis graph for Hp > 45Hz;

[0052] Figure 7 This is a schematic diagram of the frequency band division in Embodiment 1 of the present invention;

[0053] Figure 8 This invention provides a comparison of conventional amplitude compensation for a survey line in the Ordos Basin in Embodiment 1 with the pre-stacked deconvolution profile compensated in Embodiment 1. Figure 8 'a' represents conventional amplitude compensation. Figure 8 b represents compensation in Example 1;

[0054] Figure 9 This invention provides a comparative spectral analysis of pre-stacked deconvolution profiles from a survey line in the Ordos Basin, obtained using conventional amplitude compensation and the method of the invention, as described in Embodiment 1 of this invention. Figure 9'a' represents conventional amplitude compensation. Figure 9 b represents compensation in Example 1;

[0055] Figure 10 This invention provides a comparison of high-pass filtering (high-pass filtering > 45Hz) of pre-stacked deconvolution profiles from a survey line in the Ordos Basin using conventional amplitude compensation and the method of the invention. The comparison is shown in Example 1 of this invention. Figure 10 'a' represents conventional amplitude compensation. Figure 10 b represents compensation in Example 1;

[0056] Figure 11 This is a demonstration of the noise reduction effect of a two-dimensional survey line processing example in the eastern Ordos Basin of Embodiment 2 of the present invention. Figure 11 'a' represents the original data. Figure 11 b represents the noise suppression result;

[0057] Figure 12 Example of processing a two-dimensional survey line in the eastern Ordos Basin in Embodiment 2 of the present invention - noise reduction and overlay profile analysis;

[0058] Figure 13 This is an example of two-dimensional survey line processing in the eastern Ordos Basin of Embodiment 2 of the present invention, wherein... Figure 13 a is the spectral analysis diagram of the denoised overlay profile. Figure 13 b represents the frequency division compensation scheme developed based on spectrum analysis;

[0059] Figure 14 This is a profile of a two-dimensional survey line in the eastern Ordos Basin of the present invention, processed using conventional amplitude compensation methods. (Example 2 of the present invention)

[0060] Figure 15 This is a two-dimensional survey line compensation processing profile in the eastern Ordos Basin of Embodiment 2 of the present invention. Detailed Implementation

[0061] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustrative and understanding purposes only and are not intended to limit the scope of the invention.

[0062] Example 1: A pre-stack amplitude compensation method applicable to the Loess Plateau region

[0063] This embodiment describes a pre-stack amplitude compensation method for the Loess Mountain region of the Ordos Basin, including the following steps:

[0064] Step 1: Denoise the seismic data, especially low-frequency surface waves and regular interference, to highlight the energy of the effective low-frequency signal and improve the signal-to-noise ratio of the reflector layer.

[0065] Step 2: Overlay the denoised data. This improves the signal-to-noise ratio of the reflector layer and suppresses the influence of noise (higher accuracy than single-shot analysis), which is beneficial for improving the accuracy of subsequent data partitioning and spectral analysis.

[0066] Step 3: Divide the data into partitions based on the quality of the overlay profile.

[0067] Explanation of data partitioning: There are two main reasons why the seismic data in the loess mountains varies rapidly in the lateral direction. First, the surface structure of the loess mountains varies greatly in the lateral direction (thickness of the low-velocity layer), resulting in differences in the absorption and attenuation of the seismic data. Second, the seismic excitation factors change (changes in the amount of excitation charge when encountering surface obstacles). The data differences caused by these two factors can be well reflected in the stacked profile (differences in resolution). Based on the profile quality, the seismic data can be partitioned, and within each partition, the data can be divided into frequency bands. This allows single shots with the same absorption and attenuation level to be grouped together for subsequent compensation processing.

[0068] Based on the quality of the stacked seismic profiles, seismic data can be zoned, primarily according to resolution and 3D data. Due to the large data volume, attribute maps extracted from the profiles, such as dominant frequency and signal-to-noise ratio maps, can be used. Combining this with near-surface structure maps reveals a correlation between seismic profiles and the thickness of low-velocity layers in the Loess Plateau region. A reasonable zoning of individual shots can be achieved by combining multiple map types (e.g.,...). Figure 5 In the figure, there is a significant difference in resolution between section 1 and section 2. This difference can be clearly reflected in the profile spectrum. Section 2 has a higher resolution and a wider corresponding bandwidth. The difference in resolution is related to the thickness of the low-velocity layer on the surface. The area where section 2 is located has a lower thickness of the low-velocity layer, and the absorption and attenuation of the surface layer are relatively small, resulting in a higher profile resolution and bandwidth.

[0069] Step 4: Perform spectral analysis on the overlay profile of the data from each partition.

[0070] The analysis time window is fixed at (t2-t1), and (t2-t1)>300Ms.

[0071] Using (t2-t1) as a propagation period, and employing Formula 1, the value of the attenuation coefficient COEF(t2-t1) over the period of (t2-t1) is calculated.

[0072]

[0073] In the formula, n2 is the compensation factor at time t2, and n1 is the compensation factor at time t1.

[0074] The spectrum analysis time window (t2-t1) can be determined according to the actual situation of the data. In this embodiment, a spectrum analysis time window of 1 second length is used in the Loess Mountain area of ​​the Ordos Basin.

[0075] Step 5: Divide the frequency bands based on the results of the superimposed profile spectrum analysis of the data from each partition.

[0076] Frequency band division and high-frequency compensation with low frequency as reference: Based on the theory that "assuming the underground is a perfectly elastic medium with no absorption or attenuation in the strata, the difference between deep and shallow reflections is only the phase difference; the amplitude and spectral characteristics are indistinguishable—Zhang Gulan"; and based on the example of a single shot and spectral analysis of bottom-hole excitation-well-reception (approximately assuming no near-surface absorption or attenuation) in an ultra-deep dual-well micro-logging system in the Loess Plateau region, as shown in the attached diagram. Figure 4 As shown, Figure 4 The data in the image represents ultra-deep dual-well micrologging records from the Ordos Basin. The drilling depth reaches the old strata (high-velocity layer). A dual-well configuration with a 5-meter well spacing is typically used, with one well generating the data and the other receiving it (using fiber optics, downhole geophones, etc.). This approach can approximately eliminate the absorption and attenuation effects near the surface. Single-shot records show minimal amplitude variation over time, and spectral analysis reveals that the energy at different frequencies remains essentially within a decibel level, which corroborates the underlying theory. This indicates the absence of formation absorption attenuation, with the data spectrum maintaining a relatively high dB level in the low-to-high frequency range. Since low-frequency absorption attenuation is weak in the Loess Plateau region, simple compensation (basic compensation factor n0) can be considered as eliminating absorption attenuation. Spectral analysis can be performed by dividing the frequency range based on energy decrease, with high-frequency components compensated for using low-frequency energy as the target.

[0077] After performing spectral analysis within a fixed time window (t2-t1), the low-frequency range where the energy in the spectrum remains stable is used as the standard for no absorption attenuation. In other words, the low-frequency range where the energy in the spectrum remains relatively stable is used as the standard for no absorption attenuation, as shown in the attached figure. Figure 6 Using this as a reference standard for high-frequency compensation, the maximum frequency range for compensation can be determined by scanning with a high-pass filter, and the range where the energy begins to decrease is taken as the range where compensation should be increased.

[0078] While performing amplitude compensation, noise amplification is inevitable, especially in the high-frequency band where compensation control is necessary. For high-pass filtered scanning of the superimposed profile, the maximum effective frequency for energy recovery can be estimated (for scan segments with very poor resolution). For frequencies above this range, little or no compensation is needed to reduce the introduction of random noise (see attached). Figure 6 ).

[0079] When dividing the frequency band, each fixed decrease in decibel β on the spectrum is divided into a frequency band;

[0080] Calculate β according to Formula 2:

[0081]

[0082] A1 and A2 are the amplitude values ​​of adjacent wave peaks, and their calculation formula is as follows:

[0083] A2 = A1·e -αr ·ω(t)————————————————Formula 3

[0084] In the formula, r is the propagation distance, α is the absorption attenuation coefficient, and ω(t) is the wave function; if it is a spherical wave, multiply it by 1 / r before A1.

[0085] In this embodiment, after the spectrum analysis time window is fixed at 1 second, the result is obtained from formula 3:

[0086]

[0087] If n2-n1 is obtained using formulas 2 and 7, then during frequency band allocation, each fixed decrease in the frequency spectrum by a certain decibel β must be designated as a frequency band (as shown in the appendix). Figure 7 In this embodiment, β = 10 dB is used for the Loess Mountain region of the Ordos Basin. The β value can be determined according to the actual situation of the data. When the absorption attenuation is more severe, the β value can be appropriately reduced.

[0088] Step 6: Determine the low-frequency compensation factor for each zone.

[0089] For the low-frequency component separated as the compensation standard, the low-frequency basic compensation factor n0 is tested by using absorption attenuation compensation formula 1 and combining the energy curve of profile or single-shot compensation.

[0090] Step 7: Calculate the compensation factor increment for each high-frequency sub-band of each partition.

[0091] Formulas 8 and 2 show that n2-n1 equals 0.78 (i.e., the difference in compensation factors between two adjacent frequency bands with an energy difference of 10 dB). However, considering that the underground medium is a non-homogeneous viscoelastic medium, n2-n1 should be slightly less than 0.78. Therefore, a difference in adjacent frequency bands less than 0.78 can be used for compensation. The reconstructed data are then superimposed, and the superimposed profile spectrum of the reconstructed data is analyzed. Using a standard where the high-frequency energy is slightly lower than the low-frequency energy, subsequent extensive research has shown that a value of around 0.6 generally yields the best results. Therefore, (n1-n0) = 0.6, (n2-n1) = 0.6, (n3-n2) = 0.6.

[0092] Step 8: Perform amplitude compensation for each sub-region and frequency-divided data.

[0093] Step 9: Data Reconstruction.

[0094] The formulas for steps 5 to 9 are as follows:

[0095] Xi(t)=x(t)×Hi(t) ——————————————— Formula 4

[0096]

[0097]

[0098] Hi(t) is the filtering operator of the i-th filter, x(t) is the original record, Xi(t) is the frequency-divided data, and w i , where is the weighting factor, Zi(t) is the compensated frequency-divided data, COEF(Xi(t)) is the expression for the absorption attenuation coefficient of the frequency-divided data amplitude as a function of time, t is the time window size for spectral analysis, nxi is the absorption attenuation compensation parameter corresponding to each frequency-divided data, and Y(t) is the reconstructed record after compensation.

[0099] This embodiment obtains a high-resolution seismic profile through deconvolution frequency enhancement, as shown in the attached figure. Figure 8-10 From the appendix Figure 8 As clearly shown by the arrow, the weak reflection can be effectively identified after compensation using the method of this embodiment, and the resolution of the profile is significantly improved; (from the attached...) Figure 9 It can be seen that the bandwidth of the seismic profile is significantly broadened after compensation using the method of this embodiment; as shown in the attached... Figure 10 It can be seen that the high-frequency effective reflection energy of the seismic profile is effectively compensated after the invention method is used, and the high-frequency signal-to-noise ratio is improved.

[0100] Example 2: A pre-stack amplitude compensation method applicable to the Loess Plateau region

[0101] This embodiment demonstrates a two-dimensional survey line processing example in the eastern Ordos Basin, including the following steps:

[0102] Step 1: Denoise the seismic data, such as surface waves and refracted waves. Figure 11 As shown.

[0103] Step 2: Overlay the denoised data.

[0104] Step 3: Perform data without partitioning based on the quality of the overlay profile.

[0105] Based on the denoised stacked profiles and surface structure analysis of seismic data, such as Figure 12 As shown, since the low-velocity area is relatively thin overall and the seismic acquisition excitation parameters are basically consistent, this embodiment does not adopt a zoning approach.

[0106] Step 4: Perform spectral analysis on the overlay profile of the data from each partition.

[0107] Using a 1-second spectral analysis window, the spectral analysis results are as follows: Figure 13 As shown in a.

[0108] Step 5: Divide the frequency bands based on the results of the superimposed profile spectrum analysis of the data from each partition.

[0109] In the Loess Mountain region, β = 10 is used. The frequency division method in this embodiment is as follows: Figure 13 As shown in b.

[0110] Step 6: Determine the low-frequency compensation factor for each zone.

[0111] Based on the test results of the low-frequency compensation factor, the basic compensation factor n0 = 1 is selected.

[0112] Step 7: Calculate the compensation factor increment for each high-frequency sub-band of each partition.

[0113] We obtain n1 = 1.6, n2 = 2.2, and n3 = 2.8.

[0114] Step 8: Perform amplitude compensation for each sub-region and frequency-divided data.

[0115] Step 9: Data Reconstruction.

[0116] This embodiment achieved excellent results, significantly improving upon conventionally compensated profiles. The energy and continuity of each reflection layer were enhanced, effectively improving the resolution of seismic data (e.g., Figure 14 , 15 (As shown).

[0117] It should be noted that the above description is merely 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 above embodiments, those skilled in the art can still modify the technical solutions described in the above embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A pre-stack amplitude compensation method applicable to the Loess Plateau region, characterized in that... Includes the following steps: Step 1: Denoise the seismic data; Step 2: Overlay the denoised data; Step 3: Divide the data into partitions based on the quality of the overlay profile; Step 4: Perform spectral analysis on the overlay profiles of data from each partition; fix the profile analysis time window. ,and >300Ms; Will As a propagation cycle, the amplitude is calculated using Formula 1. The expression for the attenuation coefficient The value, ——————————————Official 1 In the formula, n2 is the compensation factor at time t2, and n1 is the compensation factor at time t1; Step 5: Divide the frequency bands based on the overlay profile spectral analysis results of each partition's data; fix the profile analysis time window. After performing spectral analysis, the low-frequency range where the energy in the spectrum remains constant is used as the standard for no absorption attenuation, and the segment where the energy begins to decrease is used as the segment for which compensation needs to be added. While performing amplitude compensation, compensation control is performed on the high-frequency band. The high-pass filter scan of the superimposed profile is performed to estimate the maximum effective frequency for energy recovery, and no compensation is performed for frequency bands higher than this frequency. Step 6: Determine the low-frequency compensation factor for each zone; Step 7: Calculate the compensation factor increment for each high-frequency sub-band in each partition; Step 8: Perform amplitude compensation for each sub-region and frequency-divided data; Step 9: Data Reconstruction.

2. The pre-stack amplitude compensation method applicable to the Loess Plateau region according to claim 1, characterized in that... In step 3: Based on the resolution of the overlay profile and the 3D data, combined with the attribute map and near-surface structure map extracted from the profile, the data is partitioned.

3. The pre-stack amplitude compensation method applicable to the Loess Plateau region according to claim 1, characterized in that... In step 5: When dividing the frequency band, each fixed decrease in decibel β on the spectrum is divided into a frequency band; Calculate β according to Formula 2: ————————————————Official 2 The amplitude value of adjacent wave crests is calculated using the following formula: —————————————————Official 3 In the formula Where α is the propagation distance and α is the absorption attenuation coefficient. For wave functions; if it is a spherical wave, in Multiply by 1 / r.

4. The pre-stack amplitude compensation method applicable to the Loess Plateau region according to claim 1, characterized in that... In step 5: In the Loess Mountain area, β = 10 dB is taken.

5. A pre-stack amplitude compensation method applicable to the Loess Plateau region according to any one of claims 1-4, characterized in that... In step 6: For the low-frequency part separated as the compensation standard, the low-frequency basic compensation factor n0 is tested by using absorption attenuation compensation formula 1 and combining the energy curve of profile or single-shot compensation.

6. A pre-stack amplitude compensation method applicable to the Loess Plateau region according to claim 5, characterized in that... In step 7: , , .

7. A pre-stack amplitude compensation method applicable to the Loess Plateau region according to any one of claims 1-4 and 6, characterized in that... The formulas for steps 5 to 9 are as follows: ———————————————Official 4 ————————————Official 5 ————————————Official 6 Let i be the filtering operator for the i-th filter. For the original record, For frequency division data, As a weighting factor, For the compensated frequency division data, Here, is the expression for the absorption attenuation coefficient of the frequency-divided data amplitude as a function of time, where t is the size of the time window for spectral analysis, and nxi is the absorption attenuation compensation parameter corresponding to each frequency-divided data. This is the reconstructed record after compensation.

Images