Geoscience Reference
In-Depth Information
Fig. 2.31
Seismic migration. During seismic data recording the reflection from point A will be
recorded by a coincident source and receiver at location B. It will be plotted on a stacked section as
a point A
∗
vertically below B with distance given by length AB. Migration moves the point back to
its subsurface reflection point and plots it vertically below point C, thus giving a section that looks
more like a cross-section through the earth.
2.5.5
Stacking and migration
The next steps in the processing sequence are to prepare the data for seismic migration.
Migration is one of the key steps in seismic data processing - it is the step which
attempts to move the recorded data so that events lie in their correct spatial location
rather than their recorded location
(figs. 2.31
and
2.32)
. The mathematics of how to
perform this process is well defined by the wave equation and has been extensively
researched over the past three decades. There are a number of texts, such as Stolt &
about seismic migration that describe the mathematics in detail. Once more there are
a large number of options ranging from migrating all the pre-stack data to stacking
data in a CMP followed by post-stack migration. There is also the issue of whether to
use time or depth migration and also the type of algorithm (Kirchhoff, implicit finite
difference, explicit finite difference, FK, phase shift, etc.). In recent years the choice
has become even wider with the ability of some algorithms to incorporate the effects
of velocity anisotropy. To a large extent the details of the algorithm are unimportant to
the interpreter. What matters are the accuracy and cost. These are determined by issues
such as the largest dip that can be properly migrated, the frequency content of the final
migration and the time required to perform the operation. The choice of whether to
migrate data before or after stacking is largely dependent on the velocity regime and
the subsurface dips present in the data. Large dips may mean that shallower (slower)
events arrive at the same time as deeper events, giving rise to the two events needing
to stack at the same time but with different velocities
(fig. 2.33)
. Such an occurrence is
known as a stacking conflict and the solution is to migrate the data before stack. The
