Page 449 - Acquisition and Processing of Marine Seismic Data

440 9. VELOCITY ANALYSIS

the group interval of this dataset was 6.25 m, the Thus, the normal moveout times of the
subsurface was actually sampled at every reflectionsdecreaseasthe depthofthereflec-
3.125 m, which means that there are traces in the tion increases. Velocity estimation using sem-
datawith3.125 moffsetincrement.Whatistherea- blance analysis depends on the curvature of
son for 50-m offset increment in the CDPs, and the hyperbolas, and since the curvature
whereare themissing offsetsinaspecifiedCDP? decreases with depth, the resolution in velocity
In fact, they are located within the adjacent CDPs. estimation is significantly reduced in the dee-
For instance, the trace with offset 106.25 m is in per parts of a semblance plot. The example in
CDP 47, 100-m offset is in CDP 48 etc. (Fig. 9.13). Fig. 9.14 showsthisdegradation.The upper
Based on the theory of the supergather forma- and lower limits from which a velocity value
tion, the traces of different offsets in the neighbor- can be picked from the semblance enclosures
ing CDPs are combined to form a multioffset are indicated with white dotted curves. For
CDP gather, which includes as many offsets as instance, the velocity of an arbitrary reflection
possible. Therefore, the number of traces in a from the shallow parts, e.g., the hyperbola indi-
supergather is much higher than the optimum cated by I in the supergather in Fig. 9.14,can be
fold of the line. For the example supergather in precisely picked from point A in the semblance
Fig. 9.13B, it is formed by combining five adjacent plot as V A ¼ 1472 m/s, since the enclosures in
CDP gathers of Fig. 9.13A, and according to the the region are relatively tight which prevents
offsets of this supergather (close-up number VI picking an incorrectly high or low velocity
with red borderline), four additional traces with value for this reflection. On the other hand,
different offsets are inserted among the succes- for deeper parts of the semblance, the enclo-
sive traces from the input CDP gathers. For sures become much wider along the velocity
instance, there are no traces between offsets (horizontal) axis, which prevents picking the
50 and 100 m in close-up IV in Fig. 9.13;however, precise velocity values for the reflections. For
in close-up VI of the supergather, the traces with instance, the velocity of a deep reflection, such
offsets 56.25, 62.5, 68.75, and 93.75 m are incorpo- as the one indicated by II in the supergather in
rated between the traces with 50- and 100-m off- Fig. 9.14, can be picked either from points B, C,
sets as the contributions of neighboring CDPs. or D in the semblance plot as V B ¼ 1843,
V C ¼ 1914, and V D ¼ 1984 m/s, respectively,
9.3.2 Effect of Reflection Depth since the enclosures in this area are relatively
wide. This situation indicates that any velocity
As the depth of a reflection from the seafloor
value between 1843 and 1984 m/s can be used
increases, the resolution in the velocity of that
to flatten this hyperbola during NMO correc-
reflection obtained from the semblance plots is
tion, and therefore the stack section will be fine
significantly reduced. Basically, there are two
even if we cannot determine the velocity of this
reasons for this:
reflection precisely, because the reflection
• The curvature of reflection hyperbolas will be flattened for any velocity value in this
(namely, the normal moveout) decreases as range. Anyway, we cannot know the exact
the velocity of the medium increases, and the velocity value of this deep reflection, which
velocity generally increases with depth denotes that we lose the resolution in velocity
(Fig. 9.10). determination as the depth of the reflections
• The curvature of reflection hyperbolas also increases, and the velocities can be estimated
decreases as the depth (namely, zero offset much more precisely in the shallow parts of
time) of the reflection increases (Fig. 9.5). the seismic data.