|
|
|
| Wave-equation tomography for anisotropic parameters | |
|
Next: Numerical test of the
Up: Li and Biondi: Anisotropic
Previous: Parameterization
Anisotropic WETom is a non-linear inversion process that aims to find
the anisotropic model that minimizes the residual field
in the image space. The residual image is derived
from the background image
, which is computed with current
background model. In general, the residual image is defined as
(Biondi, 2008)
|
(7) |
where
is a focusing operator acting on the background image.
In the least-squares sense, the tomographic objective function can be
written as follows:
|
(8) |
To perform the WETom for anisotropic parameters, we first need to
extend the tomographic operator from the isotropic medium
(Shen, 2004; Sava, 2004; Guerra et al., 2009) to the anisotropic medium. We
define the image-space wave-equation tomographic operator T for anisotropic
parameters as follows:
where m is the anisotropy model, which in this case includes vertical
slowness
and anellipticity parameter
;
is the background anisotropy model, consisting of the background
slowness
and background anellipticity
; I is the image. This WETom operator T is a
linear operator that relates the model perturbation
to
the image perturbation
as follows:
|
(10) |
In this paper, we evaluate the anisotropic tomographic operator in the
shot-profile domain.
Both source and receiver wavefields are downward continued in the shot-profile
domain using the one-way wave equations (Claerbout, 1971):
|
|
|
(11) |
and
|
|
|
(12) |
where
is the source wavefield at the image
point
with the source located at
;
is the receiver wavefield at the
image point
with the source located at
;
is
the source signature, and
defines the
point source function at
, which serves as the boundary
condition of Equation 11;
is the
recorded shot gather at
, which serves
as the boundary condition of Equation 12. Operator
is
the dispersion relationship for anisotropic wave propogation:
|
|
|
(13) |
where
is the angular frequency,
is the slowness
at
,
is the anellipticity at
;
is the spatial wavenumber vector. Dispersion
relationship 13 can be approximated with a rational
function by Taylor series and Padé expansion analysis (Shan, 2009):
|
(14) |
where, to the second order,
. Using
binomial expansion, Equation 14 can be further expanded
to polynomials:
|
(15) |
The background image is computed by applying the cross-correlation
imaging condition:
|
|
|
(16) |
where the overline stands for the complex conjugate, and
is the subsurface half-offset.
Under the Born approximation, a perturbation in the model parameters
causes a first-order perturbation in the wavefields. Consequently, the
resulting image perturbation reads:
where
and
are the background source and
receiver wavefields computed with the background model
,
and
are the perturbed
source wavefield and perturbed receiver wavefield, respectively, which
result from the model perturbation
.
To evaluate the adjoint tomographic operator
, which
backprojects the image perturbation into the model space, we first
compute the wavefield perturbation from the image perturbation using
the adjoint imaging condition:
The perturbed source and receiver wavefields satisfy the following
one-way wave equations, linearized with respect to slowness and
:
|
|
|
(19) |
and
|
|
|
(20) |
where
is the row vector
, and
is the transpose of
.
When solving the optimization problem, we obtain the image
perturbation by migrating the data with the current background model
and performing a focusing operation (Equation 7). Then the
perturbed image is convolved with the background wavefields to get the
perturbed wavefields (Equation 18). The scattered
wavefields are computed by applying the adjoint of the one-way
wave-equations 19 and 20. Finally, the
model space gradient is obtained by cross-correlating the upward
propagated scattered wavefields with the modified background
wavefields (terms in the parentheses on the right-hand sides of
Equations 19 and 20).
|
|
|
| Wave-equation tomography for anisotropic parameters | |
|
Next: Numerical test of the
Up: Li and Biondi: Anisotropic
Previous: Parameterization
2010-05-19