Next: Datuming from a planar
Up: Mo: Datuming
Previous: Introduction
Poststack and prestack datuming, and various migration schemes,
all take surface
recorded data as containing subsurface primary reflections only.
This means that the surface can be regarded
as an absorbing boundary when recording the seismic wavefields.
In poststack finite-difference reverse-time migration (McMechan, 1983), these
surface recorded wavefields are taken as boundary conditions for running the
two-way wave equation backward in time.
The wavefield at time t=0 is the
subsurface reflectivity image, according to the imaging principle.
The recorded wavefields
can also be taken as the boundary condition for running
the two-way wave equation
forward or backward in time in a thin strip containing the two datums.
The desired wavefields are the history wavefields at
the output datum, with the four sides designed as absorbing boundaries.
The wave equation runs
forward in time (beginning from the minimum time) if one is extrapolating
from a lower datum to a higher datum, and backward in time
(from the maximum time) if one is
extrapolating from a higher datum to a lower datum, because the
actual earth wavefields expand up from the subsurface.
The theory and results of datuming by the Kirchhoff-method can be found
in Berryhill (1979, 1984) and Shtivelman and Canning (1988).
Here I do the datuming by solving the two-way acoustic wave equation
by the finite-difference method.
Even though any recursive wavefield extrapolation
method can be applied to accomplish datuming,
I have chosen this particular
method because it is more accurate and
can honor arbitrary velocity models.
The following sections describe upward and downward datuming with increasing
degrees of complexity.
The results show that a traveltime trajectory distorted
by irregular recording topography can be corrected
by datuming to a planar datum, either upward or downward.
Next: Datuming from a planar
Up: Mo: Datuming
Previous: Introduction
Stanford Exploration Project
11/17/1997