Page 284 - Acquisition and Processing of Marine Seismic Data

5.6 GAIN RECOVERY 275

reflectors almost always have much lower the distribution of the Q value in three dimen-
amplitudes and frequency content when com- sions cannot be obtained from the seismic data
pared to the shallow reflections. Assuming a in sufficient resolution.
constant velocity medium, a lower dominant Because of these effects, amplitudes of the late
frequency implies a higher signal period and arrivals in raw seismic data are not visible on the
hence a longer wavelength, which ultimately raw data although they do exist. These invisibly
causes much lower resolution in the deeper small amplitudes are boosted by gain recovery,
parts of a seismic section. which is actually a time variant scaling of the
Although we can compensate for the spheri- seismic traces. There are various methods to
cal divergence effect by applying a suitable gain recover the low amplitudes of the seismic data.
function to the seismic traces, the effect of The most common techniques are known as true
absorption cannot be completely eliminated by amplitude recovery (TAR), which is applied as
P
existing processing techniques, since we cannot either spherical divergence correction or t cor-
model the subsurface distribution of Q values in rection, and automatic gain control (AGC). In
three dimensions along the entire survey area. principle, a proper gain function is defined for
Today, there are several inverse Q filtering a specific time gate along the time axis, and
approaches (e.g., Wang, 2002; Nunes et al., the amplitudes within this gate are multiplied
2011) in the literature to compensate for the by the gain function (Fig. 5.34)as
absorption effects, but most of them consider a
constant Q value for the whole survey area since stðÞ ¼ gtðÞ atðÞ (5.7)

FIG. 5.34 (A) A filtered seismic trace, (B) gain function, and (C) trace after gain recovery.