Page 452 - Acquisition and Processing of Marine Seismic Data

9.3 VELOCITY ANALYSIS IN PRACTICE 443

FIG. 9.16 A comparison of the semblance plots of two marine supergathers with (A) longer (1500 m length), and (B) shorter
0
0
(600 m length) streamer cables. Considering two reflections indicated by A and B in (A) and their counterparts A and B in (B),
0
a precise reflection velocity for A and B can be derived whereas only a range of velocity values can be estimated for A and B 0
using short offset data from the areas indicated by white horizontal bars on the semblance plot in (B).
noise amplitudes. Fig. 9.18 shows the effect of
9.3.4 Effect of Fold Number
fold number on the quality of semblance plots
Increasing the fold number of a specified CDP on a real marine supergather. Although the
gather simply results in extra traces involved major enclosures on the semblance plots are pre-
in the semblance calculation, and therefore served, several artificial enclosures appear as
increases the quality of semblance plots. the fold number decreases. This situation may
Fig. 9.17 shows two semblance plots calculated become an issue when picking the velocities at
from two synthetic CDP gathers (parameters the deeper parts of the semblance plots of low-
of the synthetic data are given in Fig. 9.10) with fold seismic data, where the amplitudes of the
different fold numbers. A 20% random noise is primary reflections are relatively weak.
also included in the synthetic data to demon-
strate the effect of fold on the semblance plots 9.3.5 Effect of Multiple Reflections
in the case of random noise. Fig. 9.17A and
B show the semblance plots calculated from a One of the most important agents that com-
120-fold CDP with a 12.5-m trace interval and plicate the semblance plots, even making the
a 30-fold CDP with a 50-m trace interval, respec- precise velocity picking quite difficult in rela-
tively. The semblance plot of the higher fold data tively deeper parts of the data, is the existence
in Fig. 9.17A is of higher quality, with no degra- of multiple reflections. In general, reflections
dation due to the random noise. The semblance from the seabed as well as shallow subsurface
plot quality, however, is poor for the lower fold reflectors repeat themselves, which sometimes
data in Fig. 9.17B which has artificial enclosures, makes it impossible to discriminate primary
especially in the shallow parts, arising from the reflections underneath the multiple reflections.