- journal_title：Geophysics
- Contributor：Leon Thomsen
- Publisher：Society of Exploration Geophysicists
- Date：1999-
- Format：text/html
- Language：en
- Identifier：10.1190/1.1444577
- journal_abbrev：Geophysics
- issn：0016-8033
- volume：64
- issue：3
- firstpage：678
- section：Articles
Converted-wave processing is more critically dependent on physical assumptions concerning rock velocities than is pure-mode processing, because not only move-out but also the offset of the imaged point itself depend upon the physical parameters of the medium. Hence, unrealistic assumptions of homogeneity and isotropy are more critical than for pure-mode propagation, where the image-point offset is determined geometrically rather than physically. In layered anisotropic media, an effective velocity ratio gamma eff identical with gamma 2 2 /gamma 0 (where gamma 0 identical with V p /V s is the ratio of average vertical velocities and gamma 2 is the corresponding ratio of short-spread moveout velocities) governs most of the behavior of the conversion-point offset. These ratios can be constructed from P-wave and converted-wave data if an approximate correlation is established between corresponding reflection events. Acquisition designs based naively on gamma 0 instead of gamma eff can result in suboptimal data collection. Computer programs that implement algorithms for isotropic homogeneous media can be forced to treat layered anisotropic media, sometimes with good precision, with the simple provision of gamma eff as input for a velocity ratio function. However, simple closed-form expressions permit hyperbolic and posthyperbolic moveout removal and computation of conversion-point offset without these restrictive assumptions. In these equations, vertical traveltime is preferred (over depth) as an independent variable, since the determination of the depth is imprecise in the presence of polar anisotropy and may be postponed until later in the flow. If the subsurface has lateral variability and/or azimuthal anisotropy, then the converted-wave data are not invariant under the exchange of source and receiver positions; hence, a split-spread gather may have asymmetric moveout. Particularly in 3-D surveys, ignoring this diodic feature of the converted-wave velocity field may lead to imaging errors.