Page 292 - Acquisition and Processing of Marine Seismic Data

5.6 GAIN RECOVERY 283

FIG. 5.44 Example stack sections after 500 ms AGC indicating low amplitude “quiet” zones located (A) just above the sea-
bed, and (B) just below a strong subsurface reflection. Width of this quiet zone always equals half of the AGC operator length
(Δt/2 ms) for an application to the central samples of the AGC window, here 250 ms for both sections.

The most important advantage of AGC is that an almost completely transparent feeder chan-
it is very fast to apply and does not require com- nel filled with mud breccia reaching from sea-
plicated processing parameters to be selected bed to considerable depths. Since the mud
accurately. However, it also has a very substan- breccia has no reflective internal composition
tial disadvantage: it completely removes relative for seismic waves, the feeder channel appears
amplitude information embedded in the seismic as a vertical chimney with no internal ampli-
data. On the other hand, trace-by-trace ampli- tudes with respect to the surrounding stratigra-
tude variations sometimes indicate subsurface phy, which makes the mud volcanoes easily
stratigraphic transitions within the seismic data. distinguishable on the seismic data. Fig. 5.45B
For instance, gas accumulations in a reservoir or shows the section after the t P correction, in
oil-gas interfaces produce high-amplitude local which the mud volcano is still quite distinct
reflections on the seismic data. In order to delin- and the feeder channel is still transparent
eate such high-amplitude reflections, trace-by- with almost no internal amplitudes. In addition,
trace amplitude consistency must be preserved the amplitudes of the relatively deeper reflec-
during the gain recovery. Fig. 5.45A shows a tions are slightly increased while the shallow
stack section with no gain containing a mud vol- amplitudes are diminished as expected. In
cano at the central part. A common characteristic Fig. 5.45C, however, the minor amplitudes, pos-
of underwater mud volcanoes is that they have sibly from high-frequency random noise, in the