Detection and measurement of wave breaking - Breaking and Dissipation of Ocean Surface Waves - page 89

Figure 3.10 Comparison of dominant breaking probabilities (2.3) obtained by the spectrogram
(denoted as 'BYB') and bubble-detection (denoted as 'present') methods. Figure is reproduced from
Manasseh et al. ( 2006 ) © American Meteorological Society. Reprinted with permission
The acoustic signal was preamplified by a Bruel & Kjaer 2635 charge amplifier set to the
hydrophone's calibration such that 1 V output represented exactly 100 Pa sound pressure
amplitude. The signal was passed through a 400 Hz unity-gain high-pass filter and digitised
at 40 kHz. The pulsewise processing was applied in real time on 5 min of data. Since in
this laboratory testing every wave broke, the hydrophone was deliberately placed within a
few cm of the bubble-formation zone, so rather than determining the optimum trigger level
by classification-accuracy analysis as for field data, the criterion was simply to minimise
variance in the processed data while keeping the data collection time per run reasonably
brief. Typically, 500-1000 pulses were acquired.
The difference between the water elevations upstream and downstream of the wave-
breaker were used to calculate the energy loss (2.24) , a parameter assumed to represent
the true breaking severity ( Section 2.7 ). The results are shown in Figure 3.12 , where the
mean local bubble radius R 0 is shown with the 95% confidence interval calculated from
the pulsewise processing. It can be seen that there is a clear, though not necessarily linear,
increase in R 0 with the loss of energy by the plunging breaker. At higher wave ampli-
tudes than those shown, breaking increasingly occurred prior to the board and between
the upstream probe and the board, so those conditions could not be used for the present