角度域转换波地震叠前深度偏移成像方法研究

发布时间：2018-10-10 12:03

【摘要】：目前,影响叠前深度偏移的主要因素包括三个方面:第一,叠前地震数据;第二,成像速度场;第三,偏移算子的选取。在实际叠前成像中,数据是前提,算子、算法是保证,成像速度是核心,它们三个是相互影响的,一个好的成像结果也主要建立在对这三个问题很好解决的基础之上。但是直到现在,只有对第三个问题的研究比较深入,其它两个方面的研究还有很大的发展空间。同时,叠前深度偏移未来的发展方向应该为高信噪、高分辨、高保真的全波成像;因此,开展更为广泛的转换波成像方法和用于偏移速度建模的共成像点道集的研究具有理论和现实意义。叠前深度偏移(PSDM)对速度场的要求很高,速度与层位不确定性使得PSDM对速度十分敏感。因此,我们可以利用PSDM方法来构造速度模型,以成像结果的好坏来判断速度场的正确性。在偏移速度建模中,如何建立一个用于偏移速度分析的共成像道集,近年来成为地球物理领域研究的一个热点。其中,应用较为普遍的为偏移距域共成像点道集(ODCIG)和角度域共成像点道集(ADCIG),特别是ADCIG,近年来备受国内外学者的关注,且由波动方程方法得到的ADCIG被证明是惟一没有假象的道集,也被认为是目前最为合理的道集。论文首先从基于更加严格的解耦理论对全声波方程进行单程波保幅分解,得到更直观、高效的对波场的压力分量直接进行延拓的单程波波动方程。利用解耦之后的单程波算子推导了单平方根(SSR)和双平方根(DSR)保幅延拓算子,分别利用SSR算子和DSR算子研究了PP波和PSV波叠前深度偏移。然后,在CMP域,利用双平方根算子对波场进行反向延拓,得到偏移距域共成像点道集,在频率-波数域将偏移距信息映射到角度域,得到ADCIG,指出ADCIG的角度域信息主要体现在局部偏移距中,本质上,它仍旧是对地震数据中多偏移距信息的体现。同时,利用MPI(Message Passing Interface)并行算法解决了叠前偏移耗时长、计算效率低等问题,使叠前偏移用于实际地震勘探成为可能。最后,研究了ODCIG和ADCIG对速度误差的敏感性,指出ADCIG更适合作为速度分析和AVO/AVA分析道集,当速度正确时,ADCIG同相轴是拉平的道集;当偏移速度偏小时,ADCIG同相轴上翘,ADCIG与成像深度成椭圆关系;当偏移速度偏大时,ADCIG同相轴下弯,ADCIG与成像深度成双曲关系。
[Abstract]:At present, the main factors affecting prestack depth migration include three aspects: first, pre-stack seismic data; second, imaging velocity field; third, migration operator selection. In the actual prestack imaging, the data is the premise, the operator, the algorithm is the guarantee, the imaging speed is the core, the three of them are mutual influence, a good imaging result is mainly based on solving these three problems. But until now, only the research on the third problem is more in-depth, the other two aspects of the study still have a lot of room for development. At the same time, the future development direction of prestack depth migration should be full wave imaging with high signal noise, high resolution and high fidelity. It is of theoretical and practical significance to carry out more extensive imaging methods of converted wave and the study of co-imaging point gathers for migration velocity modeling. Prestack depth migration (PSDM) requires very high velocity field, and the uncertainty of velocity and horizon makes PSDM very sensitive to velocity. Therefore, we can use the PSDM method to construct the velocity model and judge the correctness of the velocity field by the imaging results. In migration velocity modeling, how to establish a common imaging gather for migration velocity analysis has become a hot topic in geophysics in recent years. Among them, the common applications are offset domain co-imaging point gathers (ODCIG) and angle domain co-imaging point gathers (ADCIG), especially ADCIG, which have attracted much attention from domestic and foreign scholars in recent years, and the ADCIG obtained by wave equation method has been proved to be the only one without false image. It is also considered to be the most reasonable collection at present. Firstly, based on the more rigorous decoupling theory, the one-pass wave equation is decomposed to obtain a more intuitive and efficient one-pass wave equation which directly extends the pressure component of the wave field. The single square root (SSR) and double square root (DSR) amplitude-preserving continuation operators are derived by using the decoupled one-way wave operators. The prestack depth migration of PP wave and PSV wave is studied by using SSR operator and DSR operator respectively. Then, in the CMP domain, the dual square root operator is used to carry out the inverse continuation of the wave field, and the common imaging point gathers in the offset domain are obtained, and the offset information is mapped to the angle domain in the frequency-wavenumber domain. ADCIG, points out that the angle domain information of ADCIG is mainly reflected in local offset, but in essence, it is still the embodiment of multi-offset information in seismic data. At the same time, the MPI (Message Passing Interface) parallel algorithm is used to solve the problems of long time consuming and low computational efficiency of prestack migration, which makes it possible for prestack migration to be used in actual seismic exploration. Finally, the sensitivity of ODCIG and ADCIG to velocity error is studied, and it is pointed out that ADCIG is more suitable for velocity analysis and AVO/AVA analysis. When the velocity is correct, the cophase axis of ADCIG is a leveling gather. When the migration velocity is small, the ADCIG cophase axis is upwarped, the ADCIG is elliptical with the imaging depth, and when the migration velocity is larger, the ADCIG cophase axis bends, and the ADCIG is hyperbolic with the imaging depth.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.44

【相似文献】