Page 194 - Acquisition and Processing of Marine Seismic Data
P. 194
3.6 INLINE WAVES 185
close to the bird locations are obviously three or filters, the most important contribution to elim-
four times higher than those at channels off the inate the bird noise is supplied by the stacking
birds. Schoenberger and Mifsud (1974) manufac- process, which can remove almost all of the rem-
tured a special streamer section with 20 hydro- nant bird noise from the data.
phones located 2.4 m apart and mounted a bird
at the center of the section in order to observe
how the bird noise spreads along the streamer. 3.6 INLINE WAVES
The result indicates that the bird noise is apparent
only at nearby channels and is almost symmetri-
The noise appearing as linear events with
cal. While close hydrophones (1.2 m apart) are
trace-by-trace consistency on the raw shot
quite noisy, they observed almost no noise ampli-
records is classified as inline waves. Raw seismic
tudes at the hydrophones located farther than
data may contain several linear events from dif-
3.6 m from the bird location. Their analysis also
ferent sources, such as
provided information about the mechanism of
the bird noise: Symmetry and rapid decay are • Tail buoy noise
notthecharacteristicsoftheturbulenceeffectfrom • Mechanical cable noise (tug and strum)
the water flow around the bird fins, which sug- • Direct waves
gests that the bird noise is transferred onto the • Refracted waves
streamer and hydrophones by mechanicalmeans.
Fig. 3.15A shows an example shot record with 3.6.1 Tail Buoy Noise
distinct bird noise appearing as vertical high-
amplitude bands at the channels close to the bird The far end of the streamers is determined by
locations indicated by arrows. Amplitude resto- a tail buoy, generally with an rGPS transmitter
ration processes, especially AGC, may signifi- (Section 2.1.8). Depending on the weather condi-
cantly increase the amplitudes of the bird tions, movement of the tail buoy may generate
noise. Fig. 3.15B illustrates a shot gather filtered low-amplitude linear noise around the far off-
with an f-k filter to make the bird noise much sets of the shot gathers. Its amplitude depends
more prominent. Bird noise can easily be recog- on the ocean surface conditions, and if the sea
nized on f-k spectrum: The amplitudes come up conditions are rough, the tail buoy may act as
as horizontal bands on the 2D Fourier spectra a seismic source, such that the generated signal
because they are seen as vertical bands on time propagates from far to near offsets along the
sections (B in Fig. 3.15B). streamer. Therefore, tail buoy noise is observed
Fig. 3.16A shows a shot gather with bird noise at far offsets as events with linear but reverse-
as vertical high-amplitude stripes. The f-k spec- dip continuous amplitudes on the shot records
trum of the shot indicates that the bird noise has (Fig. 3.17). Its form does not change with the
approximately 20-Hz frequency and it is seen as tow speed for 4–8 knots range, and its frequency
a horizontal band along the wavenumber axis. band is rather low and is around 20 Hz.
The f-x spectrum, which illustrates the fre- A reasonable way to avoid tail buoy noise
quency content of each trace separately, shows during the acquisition is to tow the buoy as far
high amplitude vertical bands indicated by as possible from the last active section of the
arrows around the dominant bird noise fre- streamer. The tail buoy noise has reverse dip
quency. A suitable f-k polygon that accepts only on the shots, hence its amplitudes appear in neg-
reflection amplitudes in the f-k panel removes ative panel on the f-k spectrum, and therefore it
most of the noise arising from the birds is generally easy to remove this noise using a
(Fig. 3.16B). In addition to the band-pass or f-k suitable f-k filter.