A quantitative analysis and evaluation method for seismic result data absorption compensation

Abstract

The present application relates to a kind of seismic result data absorption compensation quantitative analysis evaluation method, it includes: step one, the amplitude spectrum coefficient a of the calculation seismic result data;Step two, the energy attenuation slope b of the calculation seismic result data;Step three, the correlation coefficient c of the calculation seismic result data;Step four, the signal-to-noise ratio SNR of the calculation seismic result data;Step five, the absorption compensation evaluation value of the calculation seismic result data ε, through the size of evaluation value ε, the rationality of absorption compensation is evaluated, if ε<0, it indicates that absorption compensation is under-compensation, there is insufficient compensation situation, when 0<ɛ<0.5, it indicates that absorption compensation is over-compensation, there is excessive compensation situation, when ɛ>0.5 or equal to 0.5, it indicates that absorption compensation is reasonable.The absorption compensation evaluation value of the present application is simple to calculate, can be quickly applied to the quantitative analysis evaluation of poststack result data absorption compensation.

Description

Technical fields:

[0001] This invention relates to the field of reflection seismic data processing technology in seismic exploration, specifically to a quantitative analysis and evaluation method for seismic data absorption compensation. Background technology:

[0002] Subsurface media are viscoelastic, and seismic waves undergo energy absorption and attenuation as they propagate through them. This manifests as weakened wave amplitude and thickened phase axes, leading to a significant reduction in the resolution of seismic data and severely limiting its application in oil and gas exploration and development. To address the energy absorption and attenuation problem in seismic data, numerous absorption compensation methods have been developed, such as inverse Q-filtering and Q-migration. These methods can compensate for the absorption of seismic data to a certain extent. However, the rationality of this compensation is currently assessed qualitatively using conventional methods such as profile comparison and spectral analysis. Existing methods employ limited evaluation indicators, and the results are susceptible to external factors such as signal-to-noise ratio and time window selection, exhibiting significant instability. Currently, there are few methods available for quantitative analysis and evaluation of seismic data absorption compensation. Summary of the Invention:

[0003] The purpose of this invention is to provide a quantitative analysis and evaluation method for the absorption compensation of earthquake data. This method addresses the problems of existing methods, such as single evaluation indicators, unstable results, and inability to conduct multi-factor comprehensive and quantitative analysis and evaluation of the absorption compensation of earthquake data.

[0004] The technical solution adopted by this invention to solve its technical problem is: a quantitative analysis and evaluation method for seismic data absorption compensation.

[0005] Step 1: Calculate the amplitude spectral coefficient 'a' of the seismic data;

[0006] Normalized amplitude spectrum analysis was performed on the same time window before and after absorption compensation of the seismic data. Based on the spectrum shape, high and low frequency distribution characteristics, dominant frequency and bandwidth information, the rationality of absorption compensation was determined from the perspective of spectrum. a is the product of the dominant frequency shift value f and the bandwidth band, i.e. a = f × band, where f is the dominant frequency shift value before and after absorption compensation of the seismic data, band is the bandwidth after absorption compensation, and band is the difference between high and low frequencies when the vertical axis of the amplitude spectrum is 0.2.

[0007] Step 2: Calculate the energy attenuation slope b of the seismic data.

[0008] A generalized S-transform is used for time-frequency analysis of seismic traces from shallow to deep. The time-frequency spectrum is analyzed horizontally to assess high-frequency energy compensation and vertically to assess mid-to-deep energy recovery. 'b' is defined as the energy attenuation slope. The seismic trace length is defined as T seconds, and the sampling interval as sam. Therefore, the number of sampling points n = T / sam. For the u-th sampling point of a seismic trace (u∈[0,n)), its time depth is represented by t. u = u×sam, and the corresponding energy attenuation slope is expressed as the dominant frequency F of the time spectrum at that sampling point. u The dominant frequency F of the time spectrum corresponding to the next adjacent sampling point u+1 The difference is denoted as b. u =F u+1 -F u The energy attenuation slope of the seismic trace is expressed as: When there is no absorption attenuation in the seismic data, b=1; when there is complete absorption attenuation, b=0; when there is undercompensation, it is a negative value; otherwise, it is a positive value.

[0009] Step 3: Calculate the correlation coefficient c of the seismic data.

[0010] The dominant frequency of shallow layers in statistical seismic data is f. main Using a main frequency of f main The Reich wavelet is used to generate a synthetic seismic record. The seismic data after absorption compensation is cross-correlated with the synthetic seismic record, and the correlation coefficient is obtained and denoted as c.

[0011] Step 4: Calculate the signal-to-noise ratio (SNR) of the seismic data.

[0012] Assume the signal-to-noise ratio calculation time window is represented as [w ij ] M×N Within this time window, there are N seismic data channels, each with M seismic sampling points, i∈M,j∈N,w ij Let represent the amplitude value corresponding to the i-th point on the j-th channel. Because seismic data is in phase and noise is random, let the effective signal be represented as signal and the noise as noise, then we have

[0013] w ij =signal ij +noise ij

[0014]

[0015] valid signal i It is the result of horizontal superposition of N seismic traces, so the total energy of the N effective signals is expressed as:

[0016]

[0017] The total energy of N signals is expressed as:

[0018]

[0019] And E = E signal +E noise

[0020] Therefore, the signal-to-noise ratio (SNR) is obtained as follows:

[0021]

[0022] Step 5: Calculate the earthquake data absorption compensation evaluation value ε. The rationality of the absorption compensation is evaluated by the magnitude of the evaluation value ε. If ε < 0, it means that the absorption compensation is under-compensated and there is insufficient compensation. If 0 < ε < 0.5, it means that the absorption compensation is over-compensated and there is excessive compensation. If ε > 0.5 or equal to 0.5, it means that the absorption compensation is reasonable.

[0023] The method for calculating the ε value is as follows: Substitute the a, b, c values ​​obtained from steps one to four and the SNR value into the following formula for calculating the absorption compensation evaluation value.

[0024] ε=SNR×(0.3a+0.4abs(b)+0.3c) / max(a,b,c)×sign(b)

[0025] Here, abs(b) is the operation to find the absolute value of b, max(a,b,c) is the operation to find the largest absolute value among a,b,c, and sign(b) is the operation to take the sign of b. If b is negative, then sign(b) = -1, otherwise sign(b) = 1.

[0026] The present invention has the following beneficial effects:

[0027] 1. This invention uses the signal-to-noise ratio of seismic data as a background value, considers the impact of the imaging quality of seismic data, and comprehensively and quantitatively determines the rationality of absorption compensation from different perspectives such as amplitude spectrum, time spectrum and correlation coefficient.

[0028] 2. The absorption compensation evaluation value of this invention is simple to calculate and can be quickly applied to the quantitative analysis and evaluation of absorption compensation of post-stack results data.

[0029] 3. This invention provides a reference for determining the effectiveness of absorption compensation treatment and offers an effective evaluation method for optimizing parameters in earthquake processing. Attached Figure Description

[0030] Figure 1 These are images showing the original seismic data and the data after different absorption compensation processing, using inverse Q filtering. Figure 1 In the middle, (1) is the original seismic result data. Inverse Q filtering is applied to (1), and three different absorption compensation processing data (2), (3) and (4) are obtained for different Q fields.

[0031] Figure 2 Figure 1 The time-frequency graphs corresponding to the four types of data.

[0032] Figure 3 Figure 1 The normalized amplitude spectrum diagrams corresponding to the four types of data. Detailed Implementation

[0033] The invention will be further described below with reference to the accompanying drawings:

[0034] This method for quantitative analysis and evaluation of seismic data absorption compensation:

[0035] Step 1: Calculate the amplitude spectral coefficient 'a' of the seismic data;

[0036] Normalized amplitude spectrum analysis was performed on the same time window before and after absorption compensation of the seismic data. Based on the spectrum shape, high and low frequency distribution characteristics, dominant frequency and bandwidth information, the rationality of absorption compensation was determined from the perspective of spectrum. a is the product of the dominant frequency shift value f and the bandwidth band, i.e. a = f × band, where f is the dominant frequency shift value before and after absorption compensation of the seismic data, band is the bandwidth after absorption compensation, and band is the difference between high and low frequencies when the vertical axis of the amplitude spectrum is 0.2.

[0037] Step 2: Calculate the energy attenuation slope b of the seismic data.

[0038] A generalized S-transform is used for time-frequency analysis of seismic traces from shallow to deep. The time-frequency spectrum is analyzed horizontally to assess high-frequency energy compensation and vertically to assess mid-to-deep energy recovery. 'b' is defined as the energy attenuation slope. The seismic trace length is defined as T seconds, and the sampling interval as sam. Therefore, the number of sampling points n = T / sam. For the u-th sampling point of a seismic trace (u∈[0,n)), its time depth is represented by t. u = u×sam, and the corresponding energy attenuation slope is expressed as the dominant frequency F of the time spectrum at that sampling point. u The dominant frequency F of the time spectrum corresponding to the next adjacent sampling point u+1 The difference is denoted as b. u =F u+1 -F u The energy attenuation slope of the seismic trace is expressed as: When there is no absorption attenuation in the seismic data, b=1; when there is complete absorption attenuation, b=0; when there is undercompensation, it is a negative value; otherwise, it is a positive value.

[0039] Step 3: Calculate the correlation coefficient c of the seismic data.

[0040] The dominant frequency of shallow (within 0.8 seconds) seismic data is f. main Using a main frequency of f main The Reich wavelet is used to generate a synthetic seismic record. The seismic data after absorption compensation is cross-correlated with the synthetic seismic record, and the correlation coefficient is obtained and denoted as c.

[0041] Step 4: Calculate the signal-to-noise ratio (SNR) of the seismic data.

[0042] Assume the signal-to-noise ratio calculation time window is represented as [w ij ] M×N Within this time window, there are N seismic data channels, each with M seismic sampling points, i∈M,j∈N,w ij Let represent the amplitude value corresponding to the i-th point on the j-th channel. Because seismic data is in phase and noise is random, let the effective signal be represented as signal and the noise as noise, then we have

[0043] w ij =signal ij +noise ij

[0044] The amplitude value is equal to the sum of the effective signal and the noise;

[0045]

[0046] valid signal i It is the result of horizontal superposition of N seismic traces, so the total energy of the N effective signals is expressed as:

[0047]

[0048] The total energy of N signals can be expressed as:

[0049]

[0050] And E = E signal +E noise The total energy of N channels is the sum of the total energy of N effective channels and the total energy of N noise channels;

[0051] Therefore, the signal-to-noise ratio (SNR) is obtained as follows:

[0052]

[0053] Step 5: Calculate the earthquake data absorption compensation evaluation value ε. The rationality of the absorption compensation is evaluated by the magnitude of the evaluation value ε. If ε < 0, it means that the absorption compensation is under-compensated and there is insufficient compensation. If 0 < ε < 0.5, it means that the absorption compensation is over-compensated and there is excessive compensation. If ε > 0.5 or equal to 0.5, it means that the absorption compensation is reasonable.

[0054] The method for calculating the ε value is as follows: Substitute the a, b, c values ​​obtained from steps one to four and the SNR value into the following formula for calculating the absorption compensation evaluation value.

[0055] ε=SNR×(0.3a+0.4abs(b)+0.3c) / max(a,b,c)×sign(b)

[0056] Here, abs(b) is the operation to find the absolute value of b, max(a,b,c) is the operation to find the largest absolute value among a,b,c, and sign(b) is the operation to take the sign of b. If b is negative, then sign(b) = -1, otherwise sign(b) = 1.

[0057] Example 1:

[0058] A quantitative analysis and evaluation method for seismic data absorption compensation, taking model seismic data as an example, specifically includes the following steps:

[0059] Step 1, according to Figure 2 ,calculate Figure 1 The amplitude spectrum coefficients a of the three different absorption compensation data (2), (3) and (4) were obtained, and the values ​​a = 0.2 for (2), a = 0.9 for (3) and a = 0.18 for (4).

[0060] Step Two, according to Figure 3 ,calculate Figure 1 The time-spectral coefficients b of the three different absorption compensation processed data (2), (3) and (4) were obtained, and the values ​​b for (2) were -0.3, b for (3) were 0.86 and b for (4) were 0.3.

[0061] Step 3: Calculation Figure 1 The correlation coefficients c of the three different absorption compensation processing data (2), (3) and (4) were obtained, and the c = 0.3 for (2), c = 0.82 for (3) and c = 0.4 for (4).

[0062] Step 4: Calculation Figure 1 The signal-to-noise ratios (SNR) of the three different absorption compensation processed data (2), (3) and (4) were compared, and the results showed that the SNR for (2) was 0.7, the SNR for (3) was 0.85, and the SNR for (4) was 0.4.

[0063] Step 5: Calculation Figure 1 The absorption compensation evaluation values ​​ε for three different absorption compensation processing data (2), (3) and (4) were obtained, and the ε corresponding to (2) was -0.63, the ε corresponding to (3) was 0.75, and the ε corresponding to (4) was 0.17.

[0064] Table 1 Figure 1 Calculation of absorption compensation evaluation value ε for data corresponding to (2), (3), and (4)

[0065]

[0066]

[0067] Table 1 shows... Figure 1 The calculation of the absorption compensation evaluation value ε for the corresponding data in (2), (3), and (4) shows that... Figure 1 The ε value of the corresponding data in (2) is -0.63, which is undercompensated, indicating insufficient compensation. Figure 1 The ε value of the data corresponding to (4) is 0.17, which is overcompensated, indicating excessive compensation. Figure 1 The ε value of the data corresponding to (3) is 0.75, which is a reasonable compensation, and the inverse Q filtering process can be considered appropriate.

Claims

1. A method for quantitative analysis and evaluation of seismic result data absorption compensation, characterized in that: Step one, calculating the amplitude spectrum coefficient a of the seismic result data; The same time window before and after the absorption compensation of the seismic result data is normalized amplitude spectrum analysis, based on the spectrum shape, high and low frequency distribution characteristics, main frequency and bandwidth information, from the perspective of spectrum to determine the rationality of absorption compensation, a is the product of the main frequency shift value f and the bandwidth band, that is, a = f x band, f is the main frequency shift value before and after the absorption compensation of the seismic result data, band is the bandwidth after the absorption compensation, band is the difference value of high frequency and low frequency when the amplitude spectrum vertical coordinate is 0.2; Step two, calculating the energy attenuation slope b of the seismic result data; The generalized S transform is used to analyze the seismic trace from shallow to deep, the compensation of high frequency energy is analyzed in the lateral direction of time-frequency spectrum, the recovery of middle-deep energy is analyzed in the longitudinal direction, b is defined as the energy attenuation slope, the length of seismic trace is defined as T seconds, the sampling interval is sam, the sampling point number n=T / sam, for the u-th sampling point of a seismic trace, u∈[0,n), its time-depth representation is t u =u×sam, the energy attenuation slope corresponding to the u-th sampling point is represented as the difference between the dominant frequency F u of the u-th sampling point and the dominant frequency F u+1 of the next adjacent sampling point, recorded as b u =F u+1 -F u , the energy attenuation slope of the seismic trace is represented as b=1 when the seismic data has no absorption attenuation, b=0 when the seismic data has complete absorption attenuation, b is negative when the compensation is insufficient, and b is positive in other cases. Step three, calculating the correlation coefficient c of the seismic result data; The main frequency of the statistical seismic result data is f main , and the artificial synthetic seismic record is generated by using the main frequency f main of the Rayleigh wave, the absorption compensated seismic result data is cross-correlated with the artificial synthetic seismic record, the correlation coefficient is calculated and is denoted as c; Step four, calculating the signal-to-noise ratio SNR of the seismic result data; Assume the signal-to-noise ratio calculation window is represented as [w ij ] M×N , in which there are N seismic data channels, each with M seismic sampling points, i∈M, j∈N, w ij represents the amplitude value corresponding to the i-th point on the j-th channel. Since the seismic data has the same phase and the noise is random, let the effective signal be represented as signal and the noise be represented as noise, then w ij = signal ij + noise ij effective signal signal i is the result of N-trace seismic trace horizontal stacking, so the total energy of N-trace effective signal is expressed as: The total energy of N channel signals is represented as and E = E signal + E noise Thus the signal-to-noise ratio SNR is Step five, calculating the absorption compensation evaluation value ε of the seismic result data, through the size of the evaluation value ε, the rationality of absorption compensation is evaluated, if ε < 0, it indicates that the absorption compensation is under compensation, there is insufficient compensation, when 0 < ε < 0.5, it indicates that the absorption compensation is over compensation, there is excessive compensation, when ε > 0.5 or equal to 0.5, it indicates that the absorption compensation is reasonable; The calculation method of ε value: the a, b, c and SNR value calculated in steps one to four are substituted into the following absorption compensation evaluation value calculation formula ε = SNR x (0.3a + 0.4abs(b) + 0.3c) / max(a, b, c) x sign(b) Where, abs(b) is the absolute value operation of b, max(a, b, c) is the maximum absolute value of a, b and c, sign(b) is the sign operation, if b is negative, sign(b) = -1, otherwise sign(b) = 1.

Images