Page 199 - Acquisition and Processing of Marine Seismic Data

190 3. NOISE IN MARINE SEISMICS

offset) and negative (from far to near offset) indicates that the shot involves moderate ampli-
dips, respectively. The amplitude of the cable tude cable noise in negative dip and relatively
noise is dependent on the weather conditions, low amplitude in positive dip. An application
tow speed, balance of the streamer in the water, of an f-k filter with a suitable f-k polygon almost
efficiency of the stretch sections at both ends of completely removes cable noise from the data
the streamer, and concordance between lead-in (Fig. 3.19B).
cable and active sections (Larner et al., 1983).
Application of too much gain or usage of unnec- 3.6.3 Direct Waves
essarily low AGC operator lengths may increase
the amplitudes of the cable noise. Direct wavespropagatehorizontallyalongthe
The amplitude of the noise is proportional to sea surface from the source to hydrophones and
the third order of the vessel speed, and if the ves- are recorded by whole available channels, and
sel speed is doubled, S/N decreases with a ratio they are always the first arrivals at each channel.
of 1/11. On the contrary, if we decrease the tow Because they travel only in the water column,
speed on the order of 18%, then the S/N will be their propagation velocity always equals to the
doubled (Schoenberger and Mifsud, 1974), P wave velocity in the sea water (approximately
which indicates that high tow speeds are not 1500 m/s). Their amplitude diminishes with
practical for seismic acquisition. Noise test increasing offset: It is maximum for the very first
results obtained at 5.3 and 8.6 knot speeds using recording channel while minimum amplitude is
a streamer with no birds deployed show that recorded at the farthest one.
noise amplitudes do not increase or decrease Fig. 3.20 shows direct waves on two filtered
with the distance to the tow point, but they shot gathers recorded at water depths of 90 and
exhibit rather regular distribution along the 180 m, respectively. As the water depth increases,
streamer, which suggests that the noise is pro- the time between seabed reflection and the direct
duced in a continuous manner within the wave increases at all offsets. In fact, direct wave is
streamer. In addition, the frequency content of an asymptote to the seabed reflection merging at
the noise does not change significantly with infinite offset. Minimum offset distance in the
the tow speed and only absolute amplitude source-streamer layout can be verified using the
increases (Schoenberger and Mifsud, 1974). arrival time of the direct wave at the first channel
The amplitude of the cable noise depends of the streamer. By using the inclination of direct
strongly on the vessel speed and it is signifi- waves on the shot or CDP gathers, it is possible
cantly reduced by collecting the data with an to obtain P wave velocity along the surficial
optimum tow speed. One can use f-k filters to waters,whichisacrucialparameterforsomemul-
remove the cable noise during the processing tiple elimination methods.
stage, because they are linear on the raw shot It is not possible to avoid recording direct
records, especially those with negative dip on waves during the acquisition. However, they
the shots. During the f-k filtering for the positive always appear before the seabed reflection
dip cable noise, however, extra attention should hyperbola, which is the first meaningful signal
be paid since their dip may coincide with those carrying primary reflection information from
of the primary reflection amplitudes, particu- the seabed, and since direct wave is an asymp-
larly at far offsets where the reflection hyperbo- tote to the seabed reflection, they theoretically
las becomes relatively linear. Fig. 3.19 shows an never interfere. Therefore, the direct waves are
example removal of the mechanical cable noise generally not problematic for reflection seismic
by an f-k filter. The f-k panel in Fig. 3.19A studies and they can be eliminated using a