A deconvolution method for seismic data

Abstract

The present invention provides a deconvolution method for seismic data, belonging to the technical field of seismic exploration. The method comprises: acquiring seismic data; performing spectrum analysis on each seismic data body in the seismic data to determine the main frequency of each seismic data body, performing statistical analysis on the main frequency of each seismic data body, and selecting the main frequency with the highest distribution probability as the main frequency of the seismic data; performing surface consistency prediction deconvolution on the seismic data to retain the main frequency characteristics of the seismic data; performing single-channel small step prediction deconvolution or single-channel pulse deconvolution on the seismic data after surface consistency prediction deconvolution to retain the local spectrum characteristics of the seismic data, and obtaining the final seismic data after deconvolution. The present invention includes two deconvolution processing processes when performing deconvolution processing on seismic data, which can retain both the overall spectrum characteristics of the original seismic data and the local spectrum characteristics of the original seismic data, and has a good deconvolution processing effect.

Description

Technical Field

[0001] The invention relates to a deconvolution method for seismic data and belongs to the technical field of seismic exploration. Background Art

[0002] Deconvolution is an especially important link in the process of seismic data processing, and plays a connecting role in the whole process. Deconvolution mainly improves the vertical resolution by compressing the basic seismic wavelets in the seismic records, suppressing reverberation and short-period multiple waves. At present, the main methods for deconvolution of seismic data are: surface consistency prediction deconvolution and surface consistency pulse deconvolution. The effect of deconvolution can directly determine the effect of subsequent processing and imaging. Since seismic data often contains multiple seismic data bodies, the spectral characteristics of each seismic data body together constitute the spectral characteristics of the seismic data. Because the spectral characteristics of each seismic data body are different, there are differences between the overall spectral characteristics and local spectral characteristics of the seismic data.

[0003] At present, when deconvolution is performed on seismic data, deconvolution is only performed once, and the prediction step of deconvolution is selected from a limited number of prediction step values ​​by trial and error. When the selected prediction step corresponds exactly to the overall spectrum characteristics of the original seismic data, the data after deconvolution can retain the overall spectrum characteristics of the original seismic data; when the selected prediction step corresponds to the local spectrum characteristics of the original seismic data, the data after deconvolution can retain the local spectrum characteristics of the original seismic data, but the data after deconvolution cannot retain both the overall spectrum characteristics and the local spectrum characteristics of the original seismic data, and the effect of deconvolution is not good. Moreover, when the deconvolution can only retain the local spectrum characteristics of the original seismic data, the spectrum characteristics of the processed seismic data will change significantly, and various instabilities will occur in the processed seismic data, especially the high-frequency components of the seismic data will be severely attenuated during the propagation of seismic waves, and the high-frequency components of the seismic data will be severely unstable. Summary of the invention

[0004] The purpose of the present invention is to provide a deconvolution method for seismic data, so as to solve the problem that when deconvolution is performed on seismic data using the current deconvolution method, the deconvolution data cannot retain both the overall spectrum characteristics and the local spectrum characteristics of the original seismic data.

[0005] To achieve the above object, the present invention provides a deconvolution method for seismic data, the method comprising the following steps:

[0006] (1) Obtaining seismic data;

[0007] (2) performing spectrum analysis on each seismic data volume in the seismic data to determine the main frequency of each seismic data volume, performing statistical analysis on the main frequency of each seismic data volume, and selecting the main frequency with the highest distribution probability as the main frequency of the seismic data;

[0008] (3) performing surface consistency prediction deconvolution on the seismic data to retain the main frequency characteristics of the seismic data;

[0009] (4) Performing single-channel small-step prediction deconvolution or single-channel pulse deconvolution on the seismic data after surface consistency prediction deconvolution processing to retain the local spectrum characteristics of the seismic data and obtain the final seismic data after deconvolution processing.

[0010] The beneficial effects of the present invention are as follows: the present invention includes two deconvolution processes when performing deconvolution processing on seismic data: the first deconvolution process is to perform surface consistency prediction deconvolution on the seismic data in order to retain the main frequency characteristics of the original seismic data. Since the main frequency of the seismic data in the present invention is the main frequency of the seismic data body with the highest distribution probability in the seismic data, the main frequency characteristics of the seismic data in the present invention can characterize the main frequency characteristics of most seismic data bodies in the seismic data. Since the main frequency characteristics are the most important part of the frequency spectrum characteristics, the main frequency characteristics of the seismic data in the present invention can characterize the overall frequency spectrum characteristics of the seismic data. That is to say, the first deconvolution process can The overall frequency spectrum characteristics of the original seismic data can be retained; the second deconvolution processing is to perform single-channel small-step prediction deconvolution or single-channel pulse deconvolution on the seismic data after the surface consistency prediction deconvolution processing, which is to perform deconvolution processing on the seismic data after the surface consistency prediction deconvolution processing from the perspective of single-channel data, and can highlight the fine frequency spectrum characteristics contained in the single-channel data, so that the local frequency spectrum characteristics of the original seismic data can be retained through the second deconvolution processing; therefore, the deconvolution processing method of the present invention can retain both the overall frequency spectrum characteristics of the original seismic data and the local frequency spectrum characteristics of the original seismic data, and can take both into account, and the deconvolution processing effect is good.

[0011] Furthermore, in the above method, the step (3) includes the following steps: determining a prediction step for performing surface consistency prediction deconvolution on the seismic data according to the main frequency characteristics of the seismic data, wherein the determination is based on: performing surface consistency deconvolution on the seismic data according to the determined prediction step can minimize the difference between the main frequency characteristics before and after the seismic data processing; performing surface consistency prediction deconvolution on the seismic data according to the determined prediction step.

[0012] The beneficial effect of doing so is: determining the prediction step length according to the main frequency characteristics of the seismic data. Since the main frequency characteristics of the seismic data in the present invention can characterize the main frequency characteristics of most seismic data bodies in the seismic data, when the prediction step length is used to perform predictive deconvolution processing on the seismic data, it is equivalent to using the spectral characteristics of most seismic data bodies in the seismic data as a basis to bring the spectral characteristics of other few seismic data bodies closer to the spectral characteristics of most seismic data bodies, so that the overall spectral characteristics of the original seismic data can be retained in the processed data, and the difference between the spectral characteristics of the seismic data after processing and the spectral characteristics of the seismic data before processing can be minimized. While highlighting the overall characteristics of the seismic data, the consistency of the seismic wavelet can be improved, outliers (especially high-frequency outliers) can be effectively suppressed, and reverberation and short-period multiple waves can be effectively suppressed, so as to obtain a smooth seismic data signal.

[0013] Furthermore, in the above method, the method also includes the step of preprocessing the acquired seismic data, wherein the preprocessing includes prestack noise suppression processing and amplitude compensation processing.

[0014] The beneficial effect of doing so is that preprocessing the acquired seismic data can make the amplitude energy of the preprocessed seismic data obtain a relatively stable state, which meets the requirements of deconvolution processing as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of a deconvolution method for seismic data in an embodiment of the method of the present invention;

[0016] Figure 2 is a schematic diagram of preprocessed seismic data in an embodiment of the method of the present invention;

[0017] Figure 3 is a schematic diagram of spectrum analysis of a seismic data volume in an embodiment of the method of the present invention;

[0018] Figure 4 is a schematic diagram of the main frequency distribution of each seismic data volume in the embodiment of the method of the present invention;

[0019] Figure 5-a is a schematic diagram of seismic single shot data before deconvolution processing in an embodiment of the method of the present invention;

[0020] Figure 5-b Yes Figure 5-a Schematic diagram of single shot data processed using the existing surface consistent pulse deconvolution method;

[0021] Figure 5-c Yes Figure 5-a Schematic diagram of single shot data processed using the existing surface consistency prediction deconvolution method;

[0022] Figure 5-d Yes Figure 5-a Schematic diagram of single shot data processed by the method of this embodiment;

[0023] Figure 6-a yes Figure 5-a Schematic diagram of spectrum analysis of single-shot data;

[0024] Figure 6-b yes Figure 5-b Schematic diagram of spectrum analysis of single-shot data;

[0025] Figure 6-c yes Figure 5-c Schematic diagram of spectrum analysis of single-shot data;

[0026] Figure 6-d yes Figure 5-d Schematic diagram of spectrum analysis of single-shot data;

[0027] Figure 7 yes Figure 5-a , Figure 5-b , Figure 5-c and Figure 5-d Schematic diagram of spectrum analysis comparison of each single shot data;

[0028] Figure 8-a It is a schematic diagram of the autocorrelation of single shot data after being processed by the existing surface consistent pulse deconvolution method;

[0029] Figure 8-b It is a schematic diagram of the autocorrelation of single shot data after being processed by the existing surface consistency prediction deconvolution method;

[0030] Figure 8-c Schematic diagram of the autocorrelation of single shot data after being processed by the method of this embodiment. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0032] The present invention provides a deconvolution method for obtaining the dominant frequency band (i.e., main frequency) of seismic data based on spectrum analysis and probability statistics, first highlighting the overall spectrum characteristics of the seismic data and then adjusting the local spectrum characteristics, so as to solve the problem that the existing deconvolution method cannot make the deconvolution processed data retain the overall spectrum characteristics and local spectrum characteristics of the original seismic data at the same time, resulting in a large change in the spectrum characteristics of the processed seismic data, poor consistency of the seismic wavelet, unclear formation characteristics, obvious high-frequency noise and many outliers.

[0033] Method Example:

[0034] Combine the following Figure 1, taking certain three-dimensional seismic data as an example, the deconvolution method of seismic data of this embodiment (hereinafter referred to as the method of this embodiment) is introduced in detail.

[0035] The method of this embodiment includes the following steps:

[0036] (1) Obtaining seismic data;

[0037] (2) performing pre-stack noise suppression and amplitude compensation processing on the acquired seismic data to obtain pre-processed seismic data, and using the pre-processed seismic data as the seismic data to be processed;

[0038] In this embodiment, the preprocessed seismic data is as follows: Figure 2 As shown, the purpose of pre-stack noise suppression and amplitude compensation is to make the amplitude energy of the pre-processed seismic data obtain a relatively stable state, which meets the requirements of deconvolution processing as much as possible.

[0039] As another implementation, when the seismic data obtained in step (1) is of high quality, the step of preprocessing the seismic data can be omitted. In this case, the seismic data obtained in step (1) is the seismic data to be processed.

[0040] (3) performing frequency spectrum analysis on each seismic data volume in the seismic data to be processed to obtain the main frequency of each seismic data volume in the seismic data to be processed, and performing statistical analysis on the main frequency of each seismic data volume to select the main frequency with the highest distribution probability as the main frequency of the seismic data to be processed;

[0041] In this embodiment, the spectrum analysis result of a seismic data volume is as follows: Figure 3 As shown, Figure 3 The horizontal axis is the main frequency value, and the vertical axis is the detection point number. Figure 3 The curve in the figure shows the cumulative percentage of each frequency as the frequency increases. For example, when the frequency is 28 Hz, the corresponding curve value is the percentage of data below 28 Hz. The main frequency distribution of each seismic data volume is shown in the figure below. Figure 4 As shown by Figure 4 It can be seen that the main frequency distribution of each seismic data body conforms to the normal distribution law. The main frequency with the highest distribution probability (hereinafter referred to as the main frequency of most seismic data bodies) is taken as the main frequency F of the seismic data to be processed. Figure 4 The frequency band with the highest distribution probability is 24-26 Hz. This frequency band is used as the main frequency band of the seismic data to be processed, and any frequency value between 24 and 26 Hz is selected as the main frequency F.

[0042] (4) performing surface consistency prediction deconvolution processing on the seismic data to be processed to retain the main frequency characteristics of the seismic data to be processed;

[0043] The specific implementation method of this step is: using the main frequency F of the seismic data to be processed and the empirical formula between the main frequency of the seismic data and the prediction step length to determine the prediction step length α of the surface consistency prediction deconvolution (hereinafter referred to as prediction deconvolution). In this embodiment, the empirical formula between the main frequency of the seismic data and the prediction step length is: α=1200 / F; applying the determined prediction step length α to the existing surface consistency prediction deconvolution method to perform surface consistency prediction deconvolution processing on the seismic data to be processed, the main frequency characteristics of the seismic data to be processed can be retained.

[0044] Among them, the empirical formula α=1200 / F is an empirical formula determined by performing predictive deconvolution tests on seismic data of different main frequencies, and the determination basis is as follows:

[0045] The period T of the seismic data can be obtained through the main frequency F of the seismic data, and the period T = 1000 / F; because the period T is a time value, the prediction step α of the prediction deconvolution is also a time value, and it can be assumed that α = nT.

[0046] Through the prediction deconvolution test of seismic data with different main frequencies, it is found that when the empirical formula between the main frequency of seismic data and the prediction step length is about α = 1.2T, the spectrum characteristics of the seismic data after the prediction deconvolution process are basically consistent with the spectrum characteristics of the seismic data before the prediction deconvolution process (that is, the difference is the smallest), and the high and low frequency components of the seismic data after the prediction deconvolution process are effectively broadened. Therefore, the empirical formula between the main frequency of seismic data and the prediction step length is: α = 1.2T, that is, α = 1200 / F.

[0047] By executing step (4) to perform the first deconvolution processing on the seismic data to be processed, the following effects can be achieved:

[0048] First, since the main frequency of the seismic data to be processed in this embodiment is the main frequency of the seismic data body with the highest distribution probability in the seismic data to be processed, the main frequency characteristics of the seismic data to be processed in this embodiment can characterize the main frequency characteristics of most seismic data bodies in the seismic data to be processed. Since the main frequency characteristics are the most important part of the spectrum characteristics, the main frequency characteristics of the seismic data to be processed in this embodiment can characterize the overall spectrum characteristics of the seismic data to be processed. Therefore, the prediction step size is determined according to the main frequency characteristics of the seismic data. When the seismic data is predicted and deconvolved using the prediction step size, it is equivalent to bringing the spectrum characteristics of other few seismic data bodies closer to the spectrum characteristics of most seismic data bodies based on the spectrum characteristics of most seismic data bodies in the seismic data, so that the overall spectrum characteristics of the original seismic data can be retained in the processed data, and the spectrum characteristics of the seismic data after processing can be minimized from the spectrum characteristics of the seismic data before processing. While highlighting the overall characteristics of the seismic data, the consistency of the seismic wavelet can be improved, outliers (especially high-frequency outliers) can be effectively suppressed, and reverberation and short-period multiple waves can be effectively suppressed, so as to obtain a smooth seismic data signal.

[0049] After the processing in step (4), the high-frequency and low-frequency components of the seismic data are broadened to a certain extent. Although the spectrum broadening range of the seismic data is limited, especially the spectrum broadening range of the high-frequency components is limited, the mid- and low-frequency components of the seismic data are effectively enhanced, and the overall characteristics of the seismic data are prominent.

[0050] As another embodiment, the empirical formula between the main frequency of seismic data and the prediction step length can also be: α = N / F ± b, where α is the prediction step length, N is the fitting coefficient, N is determined according to the best relationship curve between the main frequency of seismic data and the prediction step length (hereinafter referred to as the best relationship curve), the value range of N is 1000 to 1200, F is the main frequency of seismic data, b is the prediction step length adjustment amount, and the value range of b is 0 to b max , b max is the maximum adjustment of the prediction step length, and the maximum adjustment of the prediction step length b maxIt is determined according to the main frequency of the seismic data to be processed and the relationship curve between the main frequency of the seismic data and the maximum adjustment amount of the prediction step (referred to as the maximum adjustment amount curve); wherein the optimal relationship curve and the maximum adjustment amount curve are determined by performing deconvolution tests on multiple single-shot seismic data with different main frequencies, specifically as follows: deconvolution processing is performed on each single-shot seismic data using multiple different prediction step sizes, and the optimal prediction step size and critical prediction step size for deconvolution processing of each single-shot seismic data are determined; wherein the optimal prediction step size is determined based on the principle of minimizing the difference in spectral characteristics before and after deconvolution processing of single-shot seismic data; the basis for determining the critical prediction step size is: when deconvolution processing is performed on single-shot seismic data, when the prediction step size used is less than or equal to the critical prediction step size, the spectral characteristics after deconvolution processing of the single-shot seismic data are compared with the spectral characteristics before deconvolution processing, the main frequency changes and the high-frequency components occur Periodic oscillation, and the periodic oscillation becomes more and more obvious as the prediction step size decreases; the main frequency of each single-shot seismic data and the optimal prediction step size and critical prediction step size for deconvolution processing of each single-shot seismic data are recorded to obtain multiple groups of seismic data main frequencies and optimal prediction step sizes, multiple groups of seismic data main frequencies and critical prediction step sizes; a plane coordinate system is established with the main frequency of seismic data and the prediction step size as coordinate axes, and the obtained multiple groups of seismic data main frequencies and optimal prediction step sizes are plotted in the plane coordinate system to obtain the optimal relationship curve between the main frequency of seismic data and the prediction step size; the obtained multiple groups of seismic data main frequencies and critical prediction step sizes are plotted in the plane coordinate system to obtain the critical relationship curve between the main frequency of seismic data and the prediction step size; the optimal relationship curve and the critical relationship curve between the main frequency of seismic data and the prediction step size are subtracted to obtain the relationship curve between the main frequency of seismic data and the maximum adjustment amount of the prediction step size.

[0051] As another implementation, the prediction step α can also be determined based on experience through trial and error, as long as the surface consistent deconvolution processing is performed on the seismic data to be processed according to the determined prediction step α, the difference between the main frequency characteristics of the seismic data before and after processing can be minimized.

[0052] (5) Performing single-channel pulse deconvolution processing on the seismic data after surface consistency prediction deconvolution processing to retain the local spectrum characteristics of the seismic data to be processed and obtain the final deconvolution processed seismic data.

[0053] This step performs deconvolution processing from the perspective of single-channel data. After the processing in step (4), the seismic data signal is stable and the overall characteristics are prominent. In particular, the high-frequency noise of the seismic data is well suppressed, which can ensure the stability of high-frequency information during the subsequent single-channel pulse deconvolution, and lay a good foundation for further improving the high-frequency components of the seismic data. Since the consistency of the seismic data after the processing in step (4) is good, the single-channel pulse deconvolution can effectively compress the seismic wavelet, highlight the fine spectral features contained in the single-channel data, adjust the local features of the seismic data, highlight the detailed structural features, that is, highlight the local features of the seismic data, effectively improve the high-frequency components of the seismic data, effectively prevent the enhancement of high-frequency noise, and obtain stable high-frequency information.

[0054] As another implementation, in step (5), the single-channel pulse deconvolution may be replaced by a single-channel small-step prediction deconvolution, wherein the step size of the single-channel small-step prediction deconvolution may be set according to personal experience, for example, to between 2 and 12 ms.

[0055] After a series of processing is performed on the seismic data to be processed using this implementation method, both the overall spectral characteristics of the original seismic data and the local spectral characteristics of the original seismic data can be retained, and both can be taken into account. In addition, the processed seismic data obtained has a wide frequency bandwidth, obvious medium and low frequency information characteristics, suppressed reverberation and short-period multiple waves, good consistency of seismic signals, and stable high-frequency information, which effectively improves the vertical resolution of the seismic data.

[0056] The effectiveness of this implementation method is described in detail below in conjunction with Figures 5 to 8:

[0057] Figure 5-a is the seismic single shot data before deconvolution processing, Figure 6-a yes Figure 5-a The spectrum analysis results of the single shot data are as follows: Figure 5-a It can be seen that the reverberation of the seismic single-shot data before deconvolution is quite serious, and it is difficult to distinguish effective seismic information, which is something geologists do not want to see.

[0058] Figure 5-b It is a single shot data processed by the existing surface consistent pulse deconvolution method. It can be seen that by compressing the wavelet, the reverberation is effectively suppressed. However, while suppressing the reverberation, the effective information is also suppressed, and the high-frequency noise is strengthened, which reduces the signal-to-noise ratio of the seismic data. Figure 5-b The spectrum analysis results of the single shot data (i.e. Figure 6-b ) It can be seen that the high-frequency components of the data (the part greater than 30Hz) are strengthened, the low-frequency information (the part less than 30Hz) is relatively weakened, and the 20-30Hz part of the data is the main skeleton of the seismic data imaging.

[0059] Figure 5-c The single shot data is processed by the existing surface consistency prediction deconvolution method. It can be seen that the reverberation is suppressed to a certain extent, but the suppression effect is not ideal. At the same time, there is a lack of high-frequency information. Figure 5-c The spectrum analysis results of the single shot data (i.e. Figure 6-c ) It can be seen that both high-frequency and low-frequency information are suppressed, and only the data in the middle 20-40Hz range is highlighted. Such data has a narrow frequency band and single information, making it difficult to meet processing requirements.

[0060] Figure 5-d This is the single shot data processed by the method of this embodiment. It can be seen that the reverberation is well suppressed and the effective information is well maintained. Figure 5-d The spectrum analysis results of the single shot data (i.e. Figure 6-d ) It can be seen that the spectral characteristics of the seismic data have been well maintained, the high-frequency and low-frequency effective information have been effectively broadened, and the effective processing bandwidth is above 64 Hz and above 5 octaves.

[0061] Figure 7 This is a spectrum analysis comparison diagram of single shot data before deconvolution processing (i.e., original data), single shot data after existing surface consistency pulse deconvolution processing (i.e., existing pulse deconvolution data), single shot data after existing surface consistency prediction deconvolution processing (i.e., existing prediction deconvolution data), and single shot data processed by the method of this embodiment (i.e., deconvolution data of the present invention). It can be seen that: the data after existing surface consistency pulse deconvolution processing loses low-frequency information; the data after existing surface consistency prediction deconvolution processing has a narrow frequency band and lacks high and low frequency information; the data processed by the method of this embodiment basically maintains the spectrum characteristics of the original data, and at the same time, the high and low frequency information are effectively broadened.

[0062] Figure 8-a , Figure 8-b , Figure 8-c They are respectively the autocorrelation diagrams of single-shot data processed by the existing surface consistency pulse deconvolution method, the existing surface consistency prediction deconvolution method, and the method of this embodiment. It can be seen from the autocorrelation diagrams of single-shot data processed by these three deconvolution methods that the existing surface consistency deconvolution and the existing surface consistency prediction deconvolution have unsatisfactory sidelobe suppression effects; the method of this embodiment effectively suppresses the sidelobes, which further reflects that the effective suppression of interlayer multiple waves is relatively ideal.

[0063] In summary, the method of this embodiment adopts the concept of probability statistics to give full play to the advantage of a large amount of seismic data. First, by performing spectrum analysis on each seismic data body in the seismic data to be processed and performing statistical analysis on the main frequency of each seismic data body, the main frequency of the seismic data to be processed is determined, that is, the main frequency of most seismic data bodies is used as the main frequency of the seismic data to be processed; secondly, the prediction step size is determined in combination with the main frequency characteristics of the seismic data to be processed; then, the seismic data is subjected to surface consistency prediction deconvolution according to the determined prediction step size, so that the spectrum characteristics of a few seismic data bodies in the seismic data to be processed are brought closer to the spectrum characteristics of most seismic data bodies, thereby highlighting the overall spectrum characteristics of the seismic data, improving the consistency of the seismic wavelet, and making the processed seismic data signal stable; finally, the local characteristics of the seismic data are further adjusted by single-channel pulse deconvolution or single-channel small-step prediction deconvolution, highlighting the local characteristics of the seismic data, and effectively improving the high-frequency components of the seismic data; so that the overall characteristics of the deconvolution-processed seismic data finally obtained are prominent, the local characteristics are obvious, the frequency band is wide, the amplitude preservation is good, and the fidelity is high.

[0064] The beneficial effects of the implementation method are as follows: (1) before deconvolution of the seismic data to be processed, the frequency spectrum characteristics of the seismic data to be processed are studied, and the main frequency characteristics of most seismic data bodies in the seismic data to be processed are taken as the main frequency characteristics of the seismic data, so that the main frequency characteristics of the seismic data can represent the overall frequency spectrum characteristics of the seismic data, and the prediction step length for predictive deconvolution of the seismic data is determined according to the main frequency characteristics of the seismic data. The prediction step length is used to perform predictive deconvolution on the seismic data, which can effectively enhance the medium and low frequency components of the seismic data while widening the bandwidth of the seismic data, highlight the overall frequency spectrum characteristics of the seismic data, effectively suppress outliers (especially high-frequency outliers), effectively suppress reverberation and short-period multiple waves, improve the consistency and stability of the seismic data, and prepare data for subsequent single-channel pulse deconvolution or single-channel small-step predictive deconvolution; (2) on the basis of the seismic data after predictive deconvolution, the local characteristics of the seismic data are further adjusted by single-channel pulse deconvolution, highlighting the local characteristics of the seismic data, effectively widening the high-frequency information, and effectively preventing the generation of high-frequency noise while improving the high-frequency components, thereby improving the resolution of seismic data.

Claims

1. A deconvolution method for seismic data, characterized in that: The method comprises the following steps: (1) Obtain seismic data; (2) performing spectrum analysis on each seismic data volume in the seismic data to determine the main frequency of each seismic data volume, performing statistical analysis on the main frequency of each seismic data volume, and selecting the main frequency with the highest distribution probability as the main frequency of the seismic data; (3) The method comprises the following steps: determining a prediction step length for performing surface consistency prediction deconvolution on the seismic data according to the main frequency characteristics of the seismic data, wherein the determination is based on: performing surface consistency deconvolution on the seismic data according to the determined prediction step length can minimize the difference between the main frequency characteristics before and after the seismic data is processed; performing surface consistency prediction deconvolution on the seismic data according to the determined prediction step length to retain the main frequency characteristics of the seismic data; (4) Performing single-channel small-step prediction deconvolution or single-channel pulse deconvolution on the seismic data after surface consistency prediction deconvolution processing to retain the local frequency spectrum characteristics of the seismic data and obtain the final seismic data after deconvolution processing.

2. The deconvolution method of seismic data according to claim 1, characterized in that: The method also includes the step of preprocessing the acquired seismic data, wherein the preprocessing includes prestack noise suppression processing and amplitude compensation processing.

Images