Detection and measurement of wave breaking - Breaking and Dissipation of Ocean Surface Waves - page 82

logarithmic scale, with darker patches corresponding to higher values (i.e. louder recorded
sound levels associated with dominant-wave breakers).
If the temporal resolution of the acoustic signal in Figures 3.4 a-b is increased, as in
Figures 3.4 c-d, other features become apparent. For the breaker occurring in the first sec-
ond ( Figure 3.4 c), there is an apparent alteration of the frequency of the sound: the acoustic
carrier downshifts in frequency. This is not the case, however, for the breaker at t

=

55 s
( Figure 3.4 d), in spite of the fact that the acoustic signature of this breaker has quite a
distinct amplitude enhancement above the ambient noise. Nevertheless, both breakers are
clearly seen as breaking crests in the spectrogram in Figure 3.5 .
Felizardo&Melville ( 1995 ) applied thepassiveacoustics techniquesof Lowen &Melville
( 1991a ) and Melville et al. ( 1992 ) in the field, where breaking waves of all scales and
various dissipation rates can be present at the same time. They argued that the dependence
of ambient noise on wind is indirect, which argument signified an essential move away
from the decades-long tradition of associating wave breaking with the wind, towards wave
hydrodynamics which drives the physics of wave breaking as we understand it now. They
indeed found correlations between the ambient noise level and wave parameters related
to the incidence of wave breaking, and also between the total dissipation, estimated in
a number of different ways, and the acoustic noise. No attempts were made to obtain a
spectral distribution of the total dissipation. It should be mentioned that, in addition to the
issue of the much larger levels of ambient noise in the field described above, extending the
Lowen & Melville ( 1991a ) and Melville et al. ( 1992 ) approach into the multi-scale wave
environment may not be as straightforward as considering the duration of the hydrophone
signal above a threshold even if it was clearly determined. The acoustic energy radiated by
a breaking event and even the threshold itself (e.g. Babanin et al. , 2007b ) can be altered
because of multiple breakings nearby, or due to simultaneous breakings of different scales
at the measurement spot.
A number of other passive acoustic techniques have been further developed to work on
breaking detection and statistics. Bass & Hey ( 1997 ) and Babanin et al. ( 2001 ) both used
the spectrograms of hydrophone-recorded noise to detect breaking events. As illustrated
in Figures 3.2 - 3.5 , the identification of distinct crests in the spectrograms, spanning a
frequency range from 500 Hz to 4 kHz, appears to be a more reliable means of breaking
detection in the complex spectral environment than the integrated ambient noise exceeding
a threshold. The spectrogram method, however, can only be applied for the detection of
dominant breakers.
Babanin et al. ( 2001 ) obtained wave-breaking data at the experimental site at Lake
George near Canberra in south-eastern Australia during 1997-2000. Since the Lake George
field experiment will be frequently referred to through-out the rest of the topic, it is relevant
to provide a brief description of pertinent details here. For further details on the experi-
ment, its layout, instrumentation and measurements, we refer the reader to Young et al.
( 2005 ).
A contour map of Lake George, shown in Figure 3.6 , indicates a simple bathymetry,
with the bed sloping very gently toward the eastern shore of the lake. Since its bed form is