Introduction - 3-D Seismic Interpretation - page 3

The stacked trace is as it would be for a 1-D observation, with coincident source
and receiver, but with much improved signal to noise ratio. These traces can then be
displayed as a seismic section , in which each seismic trace is plotted vertically below
the appropriate surface point of the corresponding 1-D observation. The trace spacing
depends on the spacing of shots and receivers, but might be 12.5 or 25 m for a typical
survey. The seismic section is to a first approximation a cross-section through the earth,
though we need to note several limitations.
(1) The vertical axis is the time taken for seismic waves to travel to the reflector and
back again (often called the two-way time, TWT ), not depth.
(2) The actual reflection point in the subsurface is not necessarily vertically below the
trace position, if the subsurface reflectors are dipping. We can try to reposition the
reflection to the correct trace location so that the cross-section is closer to the real
subsurface structure, but this is only in part possible for a 2-D line (see section 1.2 ) .
(3) For a subsurface interface to generate a reflection, there has to be a change across it
of a quantity called acoustic impedance (which is the product of density and seismic
velocity in the layer concerned), so that not all interfaces of geological significance
are necessarily visible on seismic data. The seismic velocity is the velocity with
which seismic waves (see the glossary in Appendix 2) travel through the rock.
(4) The vertical resolution of the section, which is discussed further in chapter 4 , is
likely to be at best 5 ms. (TWT is usually expressed in milliseconds (ms): 1 ms

=

1000 s.) Despite all this, the 2-D section gives considerable insight into the
geometry of the subsurface.
Although not necessarily acquired in this way, a simple way of thinking of 3-D data
is as a series of closely spaced parallel sections. The spacing between these sections
might be the same 12.5 or 25 m as the typical trace spacing within each section. There
are two benefits to be derived from the 3-D coverage:
(a) correcting for lateral shifts of reflection points in 3-D rather than 2-D produces a
better image of the subsurface,
(b) the very dense data coverage makes it much easier and less ambiguous to follow
structural or stratigraphic features across the survey area.
We shall discuss each of these in turn.

1 /

Migration of seismic data

1.2

The process of transforming the seismic section to allow for the fact that the reflection
points are laterally shifted relative to the surface source/receiver locations is known
as seismic migration . For a 2-D section, fig. 1.1 shows how the problem arises. We
assume that the data as recorded have been transformed (as discussed above) to what
would be observed in the zero-offset case, i.e. with source and receiver coincident and
therefore no offset between them. For zero-offset, the reflected ray must retrace the