A processing method for surface correlation multiple wave model prediction

Abstract

The present invention belongs to the field of geophysical exploration technology and specifically discloses a method for predicting a surface-related multiple wave model. The method comprises the following steps, performed in sequence: S1, conventionally processing the collected seismic data to obtain pre-stack CMP gather data before migration; S2, stacking and signal enhancement processing of the pre-stack CMP gather data; S3, enhancing the signal-to-noise ratio and continuity of the gather in the mid-deep layers; S4, near-track reconstruction of the gather; S5, shallow-layer reconstruction of the gather; and S6, calculating the reconstructed gather data to obtain a predicted surface-related multiple wave model. The present invention improves the effects of low signal-to-noise ratio of land data, missing near-tracks, and poor shallow-layer data quality on the prediction of the surface-related multiple wave model, and significantly improves the coherence, signal-to-noise ratio, and continuity of the event axes of the surface-related multiple wave model. The present invention is suitable for predicting surface-related multiple wave models.

Description

Technical Field

[0001] The invention belongs to the technical field of geophysical exploration, and in particular relates to a method for processing surface-related multiple wave model predictions. Background Art

[0002] In the processing of land seismic data, surface-related multiple wave suppression is a very critical processing step. When using the model method for multiple wave suppression, how to predict the surface-related multiple wave model with high quality is crucial. If the prediction quality is poor, the suppression effect will be seriously affected.

[0003] Land surface-related multiples and ocean free-surface multiples are essentially the same type of multiples, both generated by the strong impedance interface between the air and the actual propagation medium. For these multiples, if the time difference between the primary and multiples is large, Radon transform techniques can be used to predict the multiple model, a relatively mature technique. However, for multiples with smaller time differences, more advanced prediction algorithms are required. For ocean free-surface multiples, the industry currently has a mature prediction technology, namely SRME. For land surface multiples, the most advanced prediction method is generalized SRME. This currently used generalized SRME only considers near-surface and data irregularities, but does not account for the impact of low land data signal-to-noise ratios, missing near channels, and poor shallow data quality on multiple model prediction. High signal-to-noise ratios and high-quality near channels and shallow data are crucial for predicting land surface multiple models. This is primarily due to the fact that models predicted with low signal-to-noise ratios tend to be less coherent, data lacking near channels cannot predict true multiple models, and poor shallow data quality leads to very poor multiple model quality. Therefore, due to the influence of factors such as the signal-to-noise ratio of land data, the observation system, and the near-surface, it is difficult for land data to meet the assumptions of the generalized SRME technology, resulting in the technology not being effectively utilized. Currently, similar technologies rarely have ideal effects. Summary of the Invention

[0004] The purpose of the present invention is to provide a surface-related multiple wave model prediction processing method to solve the problem affecting the quality of surface-related multiple wave prediction from the source, so as to achieve high-quality prediction of the surface-related multiple wave model.

[0005] In order to achieve the above-mentioned purpose, the technical methods adopted by the present invention are as follows:

[0006] A method for processing surface-related multiple wave model predictions includes the following steps performed in sequence:

[0007] S1. Perform conventional processing on the acquired seismic data to obtain pre-stack CMP gather data before migration;

[0008] S2. After stacking the pre-stack CMP gather data, perform signal enhancement processing to obtain post-stack data with continuity from shallow to deep layers and enhanced signal-to-noise ratio;

[0009] S3, perform mid-deep signal-to-noise ratio and continuity enhancement processing on pre-stack CMP gather data;

[0010] S4, performing energy matching on the post-stack data of step S2 and the pre-stack CMP gather data processed by step S3 for enhancing the signal-to-noise ratio and continuity of the mid-deep layers to obtain matched post-stack data, and then fusing the matched post-stack data into the pre-stack CMP gather data processed for enhancing the signal-to-noise ratio and continuity of the mid-deep layers to perform near-track reconstruction;

[0011] S5. Copying the matched post-stack data and assigning regular near-offset information to obtain a new CMP gather, performing shallow reconstruction on the new CMP gather and the CMP gather reconstructed after near-track reconstruction to obtain reconstructed data for predicting a surface-related multiple wave model;

[0012] S6. Calculate the reconstructed data to obtain a predicted surface-related multiple wave model.

[0013] As a limitation, the conventional processing in step S1 specifically includes: performing data interpolation and residual static correction processing on the acquired seismic data to obtain pre-stack CMP gather data before migration.

[0014] As a limitation: the processing for enhancing the signal-to-noise ratio and continuity in the mid-deep layer in step S3 is specifically as follows: performing dynamic correction on the pre-stack CMP gather data to obtain the pre-stack CMP gather data, then performing fine denoising processing on the pre-stack CMP gather data after dynamic correction, and then performing high-precision combination processing in the common shot point domain and the common detection point domain in turn, and finally sorting them back to the CMP domain to obtain pre-stack CMP gathers with enhanced deep signal-to-noise ratio and continuity.

[0015] As a limitation: Step S4 is specifically as follows: energy matching is performed on the post-stack data of step S2 with the pre-stack CMP gather data processed by step S3 after the mid-deep signal-to-noise ratio and continuity enhancement process to obtain matched post-stack data, and then the matched post-stack data is integrated into the pre-stack CMP gather data processed by the mid-deep signal-to-noise ratio and continuity enhancement process to form a new CMP gather A, and the CMP gather A is subjected to in-phase axis flattening processing, and near-track reconstruction is performed to form a new CMP gather B.

[0016] As a limitation: the matched post-stack data are copied and assigned regular near-offset information to obtain a new CMP gather C. The CMP gather C and the CMP gather reconstructed after near-track reconstructing are subjected to shallow mixing processing to obtain a new CMP gather D. The CMP gather D is subjected to shallow reconstruction using the reaction correction technology to obtain the reconstructed data for predicting the surface-related multiple wave model.

[0017] As a limitation: Step S5 is specifically: using the formula The reconstructed data are calculated to obtain the predicted surface related multiple wave model, where SC i,j is the common shot point gather, C i,j R is the common detection point gather.

[0018] Due to the adoption of the above solution, the present invention has the following beneficial effects compared with the prior art:

[0019] The present invention provides a processing method for surface-related multiple wave model prediction, which comprehensively utilizes pre-stack and post-stack data to process data used for surface-related multiple wave model prediction, thereby obtaining high-quality near-path and shallow layer data, as well as medium-deep layer data with a high signal-to-noise ratio. This method improves the influence of low signal-to-noise ratio of land data, missing near-path, and poor shallow layer quality of data on surface-related multiple wave model prediction, significantly improves the coherence, signal-to-noise ratio, and continuity of the event axis of the surface-related multiple wave model, and makes the predicted model closer to the actual multiple wave, providing a good data basis for further adaptive multiple wave suppression reduction, and producing good results using both model data and actual data.

[0020] The present invention is applicable to the prediction of surface-related multiple wave models. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

[0022] Figure 1 A schematic diagram of the ray path of a gun inspection point according to an embodiment of the present invention;

[0023] Figure 2 This is the effect of the short-range reconstruction of the model data on a single shot according to an embodiment of the present invention;

[0024] Figure 3 This is the effect of the superposition of the model data short-range reconstruction in the embodiment of the present invention;

[0025] Figure 4 This is the effect of the actual data of the embodiment of the present invention on a single gun;

[0026] Figure 5 This is the effect of superposition of actual data in the embodiment of the present invention;

[0027] Figure 6 This is the effect of the actual data on autocorrelation of the embodiment of the present invention;

[0028] Figure 7 The surface-related multiple wave prediction model and suppression effect diagram are obtained after conventional processing of the original gather;

[0029] Figure 8 The surface-related multiple wave prediction model and suppression effect diagram are obtained after the original data gather is processed by conventional processing, RNA and taup fitting;

[0030] Figure 9 The surface-related multiple wave prediction model and suppression effect diagram are obtained after the original data gather is processed by conventional processing, RNA and taup fitting processing, near-track reconstruction and shallow layer reconstruction;

[0031] Figure 10 The surface-related multiple wave prediction model and suppression effect diagram obtained by the embodiment of the present invention;

[0032] Figure 11 This is a diagram showing the suppression effect of the surface-related multiple wave prediction model obtained in an embodiment of the present invention on a cross section;

[0033] Figure 12 This is a diagram showing the suppression effect of the surface correlation multiple wave prediction model obtained in an embodiment of the present invention on autocorrelation. DETAILED DESCRIPTION

[0034] The present invention will be further described below with reference to the following embodiments. However, those skilled in the art should understand that the present invention is not limited to the following embodiments, and any improvements and equivalent changes made based on the specific embodiments of the present invention are within the scope of protection of the claims of the present invention.

[0035] Example A processing method for surface-related multiple wave model prediction

[0036] A method for processing surface-related multiple wave model predictions includes the following steps performed in sequence:

[0037] S1. Perform data interpolation and residual static correction on the collected seismic data to obtain the pre-stack CMP gather data X before migration. i,j,k , where i represents the line number, i=1,2,…,n, j represents the point number, j=1,2,…,m, and k represents the offset size, k=O1,O2…O q , and assume that the bin size is a×a, and the number of covers of each gather is fold i,j ;

[0038] S2, CMP gather data before stacking X i,j,k After stacking, signal enhancement is performed to obtain post-stack data Y with continuity from shallow to deep layers and enhanced signal-to-noise ratio. i,j ;

[0039] S3, CMP gather data before stacking X i,j,k The signal-to-noise ratio and continuity enhancement processing of the deep layer are carried out, specifically: i,j,kPerform dynamic correction to obtain prestack CMP gather data Then the pre-stack CMP gather data after dynamic correction is Fine denoising is performed, followed by high-precision combination processing in the common shot point domain and the common receiver point domain, and finally sorted back to the CMP domain to obtain pre-stack CMP gathers with enhanced deep signal-to-noise ratio and continuity. Among them, RNA and taup fitting processing are used for fine denoising, and vertical combination processing is used for high-precision combination processing;

[0040] S4, the post-stack data Y of step S2 i,j The pre-stack CMP gather data after the mid-deep layer signal-to-noise ratio and continuity enhancement processing in step S3 Perform energy matching to obtain the matched post-stack data Y′ i,j , and then the matched post-stack data Y′ i,j Fused into pre-stack CMP gather data after mid-deep signal-to-noise ratio and continuity enhancement processing In the process, a new CMP gather is formed CMP gathers Perform event flattening and near-track reconstruction to form new CMP gathers

[0041] S5, the matched post-stack data Y′ i,j Copy and assign regular near-offset information to obtain a new CMP gather CMP gathers and CMP gathers reconstructed from short-cut paths After shallow mixing wave processing, a new CMP gather is obtained CMP gathers Using the reaction correction process, the shallow layer reconstruction is used to obtain the reconstruction data X6 for predicting the surface related multiple wave model i,j,k ;

[0042] S6. Using formula Reconstruction data X6 i,j,k Calculation is performed to obtain the predicted surface related multiple wave model, where SC i,j is the common shot point gather, C i,j R is the common detection point gather.

[0043] The coherence, signal-to-noise ratio, and continuity of the surface-related multiple wave model predicted by this embodiment are significantly improved, greatly improving the impact of low signal-to-noise ratio of land data, missing near-tracks, and poor shallow data quality on the multiple wave model prediction. The predicted model is closer to the actual multiple waves, providing a good data foundation for further adaptive multiple wave suppression. Both model data and actual data have produced good results. Figure 2and Figure 3 It can be seen that this embodiment has shown good results in solving the problem of missing shortcuts on the model data; Figure 4-Figure 6 It can be seen that this embodiment has shown good results in solving the problem of poor quality of shallow and medium-deep gathers.

[0044] The following is a detailed demonstration of the effect using actual data. The original gather data meets the data characteristics of general land data, that is, the deep layer signal-to-noise ratio in the gather is low, the shallow layer quality is poor, and the near track is missing. Figure 7 The surface correlation multiple wave prediction model and suppression effect diagram are obtained after the original gather is processed conventionally. It can be seen from the figure that the prediction model quality is poor, and thus the suppression effect is poor; Figure 8 The surface-related multiple wave prediction model and suppression effect diagram are obtained after the original gather is processed by conventional processing, RNA and taup fitting. It can be seen from the figure that the prediction model quality has been improved to a certain extent; Figure 9 The surface-correlated multiple wave prediction model and suppression effect diagram are obtained after the original gathers are processed by conventional processing, RNA and taup fitting processing, near-track reconstruction and shallow layer reconstruction. It can be seen from the figure that the prediction model quality has been further improved; Figure 10 The surface-related multiple wave prediction model and suppression effect diagram obtained by the embodiment of the present invention are shown in the figure. As can be seen from the figure, the quality of the prediction model has been further improved, thereby achieving a better multiple wave suppression effect; Figure 11 This is a diagram showing the suppression effect of the surface-related multiple wave prediction model obtained in an embodiment of the present invention on a cross section. Figure 12 This is a graph showing the suppression effect of the surface correlation multiple wave prediction model on autocorrelation obtained in an embodiment of the present invention. Figure 11 and Figure 12 It can be seen that the surface-related multiple wave model predicted by the present invention has a better comprehensive suppression effect.

Claims

1. A method for processing surface-related multiple wave model prediction, characterized in that: The process includes the following steps: S1. Perform conventional processing on the acquired seismic data to obtain pre-stack CMP gather data before migration; S2. After stacking the pre-stack CMP gather data, perform signal enhancement processing to obtain post-stack data with continuity from shallow to deep layers and enhanced signal-to-noise ratio; S3, perform mid-deep signal-to-noise ratio and continuity enhancement processing on pre-stack CMP gather data; S4, performing energy matching on the post-stack data of step S2 and the pre-stack CMP gather data processed for mid-deep layer signal-to-noise ratio and continuity enhancement in step S3 to obtain matched post-stack data, then fusing the matched post-stack data with the pre-stack CMP gather data processed for mid-deep layer signal-to-noise ratio and continuity enhancement to form a new CMP gather A, performing event flattening processing on CMP gather A, and performing near-track reconstruction to form a new CMP gather B; S5. Copy the matched post-stack data and assign regular near-offset information to obtain a new CMP gather C. Perform shallow mixing processing on CMP gather C and the CMP gather reconstructed after near-track reconstruction to obtain a new CMP gather D. Use the reaction correction technique to shallowly reconstruct CMP gather D to obtain reconstructed data for predicting the surface-related multiple wave model. S6. Using formula The reconstructed data are calculated to obtain the predicted surface-related multiple wave model, where For the common artillery point road collection, It is a common detection point gather.

2. The method for processing surface-related multiple wave model prediction according to claim 1, characterized in that: The conventional processing in step S1 specifically includes: performing data interpolation and residual static correction processing on the acquired seismic data to obtain pre-stack CMP gather data before migration.

3. The method for processing surface-related multiple wave model prediction according to claim 1, characterized in that: The processing for enhancing the signal-to-noise ratio and continuity of the mid-deep layer in step S3 is specifically as follows: performing dynamic correction on the pre-stack CMP gather data to obtain the pre-stack CMP gather data, then performing fine denoising on the pre-stack CMP gather data after dynamic correction, then performing high-precision combination processing in the common shot point domain and the common detection point domain in turn, and finally sorting back to the CMP domain to obtain the pre-stack CMP gather data with enhanced deep signal-to-noise ratio and continuity.

Images