Page 477 - Acquisition and Processing of Marine Seismic Data

468 10. NORMAL MOVEOUT CORRECTION AND STACKING

FIG. 10.8 Various types of anisotropy models. (A) Vertical transverse isotropy (VTI), (B) horizontal transverse isotropy
(HTI), and (C) orthorhombic symmetry (ORT). VTI and HTI are called transverse isotropy, since the material is essentially
isotropic in its horizontal or vertical plane.

regime. The medium is stiffer along the medium, they create hyperboloids with ellipti-
fracture directions, resulting in a higher cal cross-sections for an HTI medium, indicating
propagation velocity along the cracks high- and low-velocity directions, which results
(Fig. 10.8B). The medium is isotropic along in an inaccurate velocity analysis performed,
thin vertical planes perpendicular to the assuming an isotropic subsurface. However, it
symmetry axis. is generally not a straightforward process to
• Orthorhombic symmetry (ORT): This can be obtain the anisotropic NMO velocities from
considered to be the combination of the two the surface seismic data, which require several
preceding types with three mutually elastic parameters to define the anisotropic
orthogonal planes of mirror symmetry. It is velocity. Grechka and Tsvankin (1998) showed
caused by parallel vertical cracks within a that variations in NMO velocity depending on
medium of thin horizontal layering the azimuth are characterized by an ellipse in
(Fig. 10.8C). the horizontal plane, and the axis orientations
of this ellipse are controlled by the rock proper-
The existence of anisotropy may distort the ties and the direction of the reflector normal.
normal moveout velocity for a small-spread They have also indicated that at least three azi-
approximation in a horizontally stratified sub- muthal measurements are essential to obtain
surface and results in nonhyperbolic moveouts the NMO velocity in all azimuthal directions.
(Tsvankin, 1995), where two-term NMO approx- Therefore, wide azimuth 3D seismic data is nec-
imations in Eqs. (10.1) and (10.7) become inaccu- essary for a complete measurement of azimuthal
rate with increasing offset, and therefore NMO anisotropy to construct the azimuthal NMO
velocities accounting for the anisotropic effects velocity field. These factors make the anisotropy
of the subsurface may be considered even for studies much more complex compared to the
the case of a single homogenous azimuthally conventional processing steps, such as velocity
anisotropic medium. Wallace et al. (2007) analysis for isotropic media, especially because
showed that degradation of the reflections in of the limited azimuths in data recording
the supergathers after isotropic NMO correction (Alkhalifah and Tsvankin, 1995).
is apparent in the far offset traces resorted by Several different approaches have been sug-
their azimuths, which is associated with velocity gested so far to mathematically describe the type
variations of the azimuth of propagation. and degree of the subsurface anisotropy.
Although the reflections produce hyperbo- Thomsen (1986) defined three parameters, ε, γ,
loids with circular cross-sections for an isotropic and δ, to describe the amount of anisotropy of