Detection and measurement of wave breaking - Breaking and Dissipation of Ocean Surface Waves - page 85

5 were employed for
studying the wave-breaking statistics. Their spectral peak frequencies f p ranged from 0.33
to 0.50 Hz. The wind speed U 10 ranged from 8.5 to 19

Relatively young, strongly forced waves with U 10 / c p = 3 . 3-6 .

s.
Visual detection of the dominant wave-breaking occurrence is arguably one of the most
reliable methods available since it does not require any additional experimental criteria. At
Lake George, most wave records were supplemented by video-taped images of the water
surface surrounding the wave array. The video-taping allows repeated playback of wave
records to establish the breaking statistics by visual means and to verify the results. This
visual observation method was used to obtain the breaking statistics.
As mentioned above, the video records were synchronised with the surface elevation
records gathered by the array of wave probes. Both the surface elevation sampling and
video framing were performed at 25 Hz rate so that there was unambiguous correspondence
between each wave height reading and a video image. For the video-taping a computer-
controllable Panasonic model AG-7350 video recorder (VCR) was used whose time-code
generator facility allowed rapid retrieval of particular segments of the recorded wave series
corresponding to visually observed breaking events.
The hydrophone output was recorded on the VCR audio channel. The hydrophone had
two signal gains, 20 and 40 dB. Normally, for developed wind-waves the 20 dB gain
was sufficient to detect breaking waves. During data analysis, the acoustic signal was
sampled at the 8 KHz rate and digitised. These synchronised time series of the acoustic
signal contain potentially valuable additional information about visually observed domi-
nant wave breakers, particularly in relation to breaker strength (see Manasseh et al. , 2006 ,
and Section 6.2 ).
In Figure 3.5 , a spectrogram of the first minute of record 4 in Table 5.2 is plotted. The
dark crests across almost the entire 4 KHz frequency span in the spectrogram are associated
with acoustic noise from dominant breaking waves. This was confirmed through repeated
viewing of the synchronised video records. For example, the first and last breakers (near
t

. 8m /

55 s) detected in the spectrogram in Figure 3.5 are shown in the captured
video images seen in Figures 3.2 and 3.3 , respectively. It is clear that these are cases when
the crest of a breaking wave is passing through the wave array and over the bottom-mounted
hydrophone.
After the connection between the spectrogram signatures and the visually observed dom-
inant breakers was established, the acoustic spectrograms were used along with the video
records to obtain the wave-breaking statistics (see Section 5.3.1 ). As a further example,
Figure 3.8 shows another spectrogram of breaking occurrences during a 1 min-long seg-
ment of record 12 ( Table 5.2 ) when the wind and waves were much weaker and only two
isolated breakers occurred within the 1 min record.
Thus, the use of spectrograms rather than acoustic intensity time series is a preferred
method for detection of dominant breaking events in the wave field. A similar conclusion
was reached by Bass &Hey ( 1997 ), who used spectrograms to detect breakers in the natural
surf zone. The acoustic time series, however, can provide useful physical insights into the
breaking process, that is in order to estimate periods of individual breaking waves, the

= 1 s and t =