Role of Domtnant Wave tn Spectrum of Wind-Generated Waves 
spectrum corresponding to the observed wave. It can therefore not 
be ruled out that the wave pattern is not of the equilibrium type, and 
thus the wave could be of less than maximum possible height. But it 
is not unreasonable to expect that separation, if in this case it occurs 
at all, does not cause pressure pattern which differs as much at the 
upstream and downstream faces of the wave as the laboratory case. 
Some observational evidence for this comes from the fact that Konda 
et.al. have found a very significant vertical velocity component in the 
air which was related to the period of the waves. A velocity profile 
of the vertical velocity above the wave is shown in the figure. The 
velocity is measured with a probe that was fixed on a tower at a cons- 
tant distance above an undisturbed water surface. 
The spectrum corresponding to the wave in Fig. 8 is also 
shown in Fig. 6. In this case the spectrum of the average wave drops 
off faster than the spectrum of the average laboratory wave, and its 
shape is much more like that of a Stokes wave, whose energy spectrum 
has also been drawn into Fig. 6. We conclude from this that the domi- 
nant wave of the laboratory may differ in form from the dominant wave 
of the ocean. 
How then is it possible that the similarity spectra cf the 
ocean and the laboratory are so much alike? A tentative explanation 
may be as follows, As we saw, the laboratory wave has, by accident 
or for a reason which is at present not known, a shape that yields a 
spectrum whose high frequency end obeys approximately a - 5-power 
law. The spectral analysis techniques currently practised tend to 
smear this energy of the waves over a broader spectrum, thus obs- 
curing gaps which might be in the laboratory spectrum. In contrast 
to the laboratory situation, the ocean waves are less skewed and 
tend to have lower energies associated with higher harmonics, But 
the air flow can follow the contours of the large waves and finds on- 
ly little resistance on them. Therefore, there might be generated on 
the large waves smaller waves which show the same characteristics 
as the laboratory waves. These then might ''flesh out" the high fre- 
quency end of the spectrum, the result being that the spectra in field 
and laboratory become similar in shape. 
VII. CONCLUSIONS 
I have shown in this paper that the model of Phillips and 
Longuet-Higgins of a water surface covered by waves of all possible 
frequencies above some lower cut-off value which are in the state of 
breaking, and for which the work done on the waves by the wind is 
equal to the energy dissipated by breaking, does not agree with the 
1383 
