Inversion for subsurface anisotropy using estimates of shear-wave splitting
Inversion for subsurface anisotropy using estimates of shear-wave splitting
复制标题
DOI：
10.1111/j.1365-246x.1991.tb01418.x
复制
发表时间：
1991-12
影响因子：
2.8
通讯作者：
C. MacBeth
中科院分区：
文献类型：
--
作者：
C. MacBeth
对剪切波分裂特征参数（极化、时延和差分衰减）的反演为全波各向异性建模提供了一种获得良好的起始模型的方法，并提高了最终波形匹配的收敛性。这是必要的，因为横波在各向异性方面表现为高度非线性。在这里，我们讨论极化和时滞信息的程度，估计从分裂剪切波形，可以提供这个开始模型的细节。考虑了这些参数在不同各向异性模型下的三维模式，以研究反演方案的潜力。为提供各向异性模型的最佳分辨率，形成了震源和检波器的最佳安排准则。研究表明，在大多数情况下，不同的各向异性模型不能通过垂直地震剖面（VSPs）、反射剖面和在一个方位角进行单独极化或延时测量的井间测量来充分区分。通过结合极化和时延估计获得更好的结果，最好是从多个方位角。可以解决的模型参数类型主要取决于采集系统的几何形状，并且在较小程度上取决于背景速度结构。个别模型参数更难提取，但某些细节（原则上）对于某些类型的调查是可以解决的。VSPs最适合监测平行、垂直、排列裂纹的走向和密度，以及导致层致各向异性的细层倾角。交叉孔试验是测量裂缝倾斜度的最佳方法。裂缝含量和纵横比的分辨率主要由线奇点的位置决定，如果线在相对于裂缝走向的适当方位角上拍摄，则偏移VSP或适应的跨孔测量都足以解决这些特性。利用近偏移VSP测量极化和延时模式记录中奇异点的位置，可以评价沉积盆地中裂缝诱导各向异性与基质各向异性的比值。
SUMMARY Inversion of the parameters characterizing shear-wave splitting (polarizations, time delays, and differential attenuation) provides a way of obtaining a good starting model for full-wave anisotropic modelling, and improves the convergence to a final waveform match. This is necessary as the shear waves behave in a highly non-linear fashion with respect to anisotropy. Here we discuss the extent to which polarization and time-delay information, estimated from split shear waveforms, may supply details of this starting model. 3-D patterns of these parameters for different anisotropic models are considered, to investigate the potential of the inversion scheme. Guidelines are formed for the best arrangements of sources and geophones, to provide optimum resolution of an anisotropic model. The study indicates that in most cases different anisotropic models cannot be adequately distinguished using vertical seismic profiles (VSPs), reflection profiles, and cross-hole surveys in which separate polarization or time-delay measurements at one azimuth are made. Better results are obtained by combining polarization and time-delays estimates, preferably from more than one azimuth. The type of model parameters which can be resolved depends crucially upon the geometry of the acquisition system, and to a lesser extent on the background velocity structure. Individual model parameters are more difficult to extract, but some details appear (in principle) to be resolvable for certain types of survey. VSPs are most suitable to monitor crack strike and density of parallel, vertical, aligned cracks, and the dip of the fine layering contributing to layer-induced anisotropy. Cross-hole experiments are best for measuring the dip of the cracks. The resolution of the crack content and aspect ratio are determined principally by the position of line singularities, and either an offset VSP, or an adapted cross-hole survey may be adequate for resolving these properties if the line is shot at an appropriate azimuth with respect to the strike of the cracks. The ratio of crack-induced anisotropy to matrix anisotropy in sedimentary basins can be evaluated by measuring the positions of the point singularities in recordings of polarization and time-delay patterns using a near-offset VSP.