Geology Reference
In-Depth Information
6.2.1.2 Surface consistent corrections
AVO analysis is performed on a gather of traces that
have a common reflection point but are acquired
from different shot
1200
1300
receiver pairs. Clearly, the effect
of shot-to-shot variation in signal strength and vari-
ations in receiver sensitivity needs to be removed. In
the marine case, these variations are likely to be fairly
small. Air guns emit highly reproducible signals and
receivers are manufactured to close tolerances, and
both source and receiver are surrounded by seawater
the properties of which do not usually vary laterally.
In the land case, the problem is much more serious.
Sources and receivers are still manufactured to give
uniform performance, but coupling to the subsurface
is strongly affected by the nature of the near-surface
layering. Sometimes it is possible to achieve fairly
uniform conditions, for example by burying geo-
phones or by sweeping loose material away so that a
Vibroseis source rests on solid rock, but often this is
not practical and corrections need to be applied in
processing. In general, seismic amplitudes are more
reliable for marine than for land data.
The near-surface effects are dealt with by surface-
consistent corrections (Taner and Koehler,
1981
). The
idea is that near-surface effects associated with a par-
ticular surface position remain constant regardless of
the raypath involved. Thus, source strength will affect
all of the traces recorded from that source, and geo-
phone coupling will be the same for all traces
recorded at a particular receiver station from different
source positions. Unwanted surface-consistent factors
can be split into three categories: source response,
receiver response, and an offset-related response
which takes account of directivity in source or
receiver arrays. Owing to the possibility of modifying
first order offset amplitude changes related to geology
it is seldom advised to apply corrections for energy
directivity. In most instances corrections are applied
to source and receivers only. Since there are traces for
many source
-
1400
1500
1600
1700
Figure 6.3
Example of amplitude shadow caused by shallow gas
(after Rudiana
et al.
,
2008
).
propagation effects from shallower rock layers. For
example,
Fig. 6.3
shows a section from a gas field
offshore Indonesia (Rudiana et al.,
2008
), where a
shallow gas accumulation strongly attenuates seismic
amplitudes below it, giving rise to an amplitude
'
. High impedance layers such as basalts and
thick carbonate layers can have similar effects. The
extent of amplitude shadows will vary with offset,
depending on how the source and receiver paths are
affected. Important factors are source
shadow
'
receiver dis-
tance as well as the depth and lateral extent of the
attenuating layer. The effects on measured AVO at a
target horizon can be complicated; near traces will
almost certainly be affected but it is possible that at
long offsets the raypaths may fully or partially under-
shoot the feature. This means that in the affected area
it will not be possible to relate AVO on a deep reflector
to properties of the interface concerned without carry-
ing out some form of data normalisation.
For simple amplitude mapping, a possible
approach is to equalise amplitudes over a long gate
by means of automatic gain control (AGC). This aims
to scale the seismic data so that the average RMS
amplitude in a window is the same across all traces,
and assumes that the average reflectivity is invariant.
This is dangerous if applied over a short window,
potentially destroying lateral amplitude variations of
interest. If applied over a long window (e.g. 1.5
-
receiver pairs, the source and receiver
responses can be calculated in a least-squares sense.
The process is analogous to calculating shot and
receiver statics and, in a similar way, is quite effective
at dealing with short-wavelength variations.
-
2s
TWT), however, it can remove much of the
shadowing effect of the overburden without dam-
aging lateral amplitude variations in a target reflector.
This solution can be applied to partial offset stacks,
but it will obviously affect the intercept vs gradient
relationships determined from the stacks. Long gate
AGC is dubious for angle stacks because the offsets
(and the degree of undershooting) contributing to any
one trace will vary with TWT.
-
6.2.2 Long-wavelength overburden effects
A key issue for the interpreter to appreciate prior to
amplitude interpretation is variation in amplitudes at
the level of
113
interest
that may be the result of
