|
|
|
| High-order kernels for Riemannian Wavefield Extrapolation | |
|
Next: Conclusions
Up: Sava and Fomel: Riemannian
Previous: Examples
Accurate wave-equation migration using Riemannian wavefield
extrapolation requires a choice of coordinate system that exploits its
higher extrapolation accuracy. An effective choice of coordinate
system would be one that minimizes the difference between the
extrapolation direction and the direction of wave propagation. If this
condition is fulfilled, we can achieve high-angle accuracy using
low-order extrapolation kernels. Otherwise, we need to extrapolate
seismic wavefields with high-order kernels, like the ones described in
this paper.
Shot-record migration requires selection of
coordinate systems for the source and receiver wavefields. Optimal
selection of coordinate systems in this situation is not a trivial
task, since the source and receiver wavefields are optimally described
by different coordinate systems which also vary with
location. However, if we employ high-order extrapolation kernels,
different seismic experiments may share the same approximately optimal
coordinate system. An easy way to illustrate this idea is represented
by imaging in (tilted) Cartesian coordinate systems, which are just
special cases of Riemannian coordinates
(Sava and Fomel, 2006). A complete treatment of this
topic remains subject for future research.
|
|
|
| High-order kernels for Riemannian Wavefield Extrapolation | |
|
Next: Conclusions
Up: Sava and Fomel: Riemannian
Previous: Examples
2008-12-02