Detection and measurement of wave breaking - Breaking and Dissipation of Ocean Surface Waves - page 96

In Phillips et al. ( 2001 ), the radar measurements were conducted with a very high resolu-
tion and the moving sea spikes were clearly distinguishable over background backscatter.
Many interesting observations were made. Moderate wind-speed conditions were simi-
lar to those observed in the acoustic field experiments by Ding & Farmer ( 1994 ) (see
Section 3.5 ). Therefore, comparisons of the acoustic sensing and radar backscatter tech-
niques could again be made, particularly as in both experiments the spectral distribution of
the breaking occurrence was measured in terms of breaker phase speeds.
Phillips et al. ( 2001 ) found that their overall breaking statistics was close to that of Ding
&Farmer ( 1994 ), but the phase speeds of breakers were on average lower (i.e. frequencies
higher).While acousticmeasurements did detect breakingwaves propagatingwith a spectral
peak speed c p , the radar measurements did not. The fastest radar-measured breakers would
propagate with the speed of some 0 6 c p . As discussed earlier, the microwave backscattering

technique identifies breaking events at or close to breaking onset, whereas the acoustic
signature signifies wave breaking well in progress. Thus, the results reported by Phillips
et al. ( 2001 ) are consistent with the notion also mentioned a number of times previously that
the length (and therefore phase speed) of waves decreases immediately prior to breaking,
if the breaking takes place as a consequence of modulational instability. Therefore, the
overall number of incipient breakers and developed breakers should be statistically the same
averaged value, but the statistically average spectral signature of such breakers, if they are
caused by the instability of nonlinear wave groups, should not. This will be discussed in
more detail in Section 5.2 (see also Babanin et al. , 2007b , 2009a , 2010a ).
Thorough quantitative measurements of breaking fronts, such as the actual form of
the phase-speed distribution of

.

, were performed by Phillips et al. ( 2001 ) (see also
Sharkov , 2007 , for a review). This kind of measurement proved valuable for obtaining
breaking statistics and breaking probability (see Section 2.5 for definitions). Estimates of
the dissipation rates based on (3.30) , however, as discussed in Section 3.4 , seem to be of
limited value, particularly away from the spectral peak.
Apart from its scientific merit, the paper by Phillips et al. ( 2001 ) also provides an
important link with optical remote-sensing studies which concentrate on measuring

( c )

.
Advancements in computer power and storage in recent decades have made it possible to
automatically process large numbers of digital video images of the water surface. The prop-
agating whitecaps can be identified and distinguished from the background darker water
surface, and therefore the statistics of breaking fronts, their propagation speeds and other
geometric, kinematic and dynamic characteristics of the areas covered with whitecaps can
be investigated by means of analysing video sequences in much the same fashion as radar
imaging. Depending on the observational setup (e.g. aircraft or a close-by experimental
tower), video imaging can be significantly more accurate because of the much smaller
footprints of the video pixels compared to the radar cross-sections. The effects of bubbles
entrained by breaking waves on the optical scattering and backscattering coefficients were
investigated by Terrill et al. ( 2001 ).
Melville & Matusov ( 2002 ) conducted video recording of wave-breaking fronts with the
purpose of obtaining both the breaking statistics and the dissipation (3.30) . Video imaging

( c )