论文详情
川东南地区茅口组岩溶异常体尺度小, 且多与断裂有关, 其地震反射特征不明显, 加之受限于上覆地层的强同相轴屏蔽, 因而地震资料分辨率不高, 识别难度较大, 预测精度不高。为此, 探索了基于地震波形分解的岩溶异常识别技术。与常规地震数据处理方法存在基函数选取复杂、信号混叠等问题不同, 变分模态分解技术具有分辨率高、分解信息完备等优势。基于基追踪方法, 首先采用该技术对地震数据进行全频带分解, 得到不同波形特征的模态分量, 再利用相关性系数对分量进行分级, 去除最大分量以削弱研究区龙潭组底界强信号层的能量影响, 选取能够表征异常特征的分量进行重构。处理后的地震数据可以有效削弱强反射层屏蔽的影响, 凸显下部岩溶的横向异常, 提高断层和岩溶的识别精度。基于处理后的地震数据体进行属性分析, 以完成对研究区岩溶异常带的综合识别。实际应用结果表明: 地震弱响应信息能在地震数据波形分解后得以呈现。利用实钻资料对上述应用结果进行验证, 结果表明该技术具备良好的应用效果, 对研究区后续岩溶储层预测具有一定的参考意义。
Karst anomalies of the Maokou Formation in southeastern Sichuan are associated with faults and have a small scale, and their seismic reflection characteristics are not obvious. It is also limited by the overlaying of strong event shielding, resulting in a low resolution of seismic data, difficulty in identification, and a lack of prediction accuracy. In view of the above problems, this study explored a karst anomaly identification method based on seismic waveform decomposition technology. Compared to the traditional method, which has the problems of complex basis function selection and signal aliasing, the variational mode decomposition(VMD) method has the advantages of high resolution and complete decomposition of seismic information. Based on the optimization of basis pursuit, the VMD method was used to decompose the seismic data in the entire frequency band to obtain the mode components with different waveform characteristics. The correlation coefficient method was applied to classify these components and the largest component was removed to reduce the energy influence of the strong signal at the bottom of the Longtan Formation. Simultaneously, components that could characterize abnormal features were selected for reconstruction. The seismic data processed using the VMD method can effectively weaken the shielding effect of the strong reflection layer, highlight the lateral anomalies of the karst in the lower part of the events, and further improve the identification accuracy of faults and thin layers. Additionally, seismic attributes were extracted from the reconstructed data to comprehensively identify karst anomalies in the study area. The practical application results show that seismically weak signal information can be enhanced to a certain extent after the strong shielding layer is weakened. The verification was performed in combination with drilling data, which has achieved favorable results. This indicates that the proposed method is significant as a reference for the follow-up prediction of karst reservoirs in the southeastern Sichuan area.