TM No. 377 



aimed approximately at the wind waves (as in serial 120), one would expect little 

 or no change in the total spectral curve of $ w as the meter is rotated. On the 

 other hand, as 9 is varied from 0° to 90° (see figure V-6a), the swell peak of 



^u. (A*) should increase and the wind wave peak (B*) should decrease with respect 

 to each other. This is shown, at least qualitatively, in the curves (figure V-17). 

 Peak A remains about the same from 120 through 123, whereas A' starts off relatively 

 small, drops off slightly, then increases from 121 through 123. Peak B, although 

 fluctuating, remains roughly constant; whereas peak B* shows strong attenuation, 

 especially from 121 to 123. 



The variance OZt 7 " displays approximately constant values (within 6-8f ) over 

 the four measurements. On the other hand, the horizontal variance Ou, 1, displays 

 a general decrease (associated with the decrease in the high frequency wind wave 

 peak B'), which is further evidence of the damping out of the wind wave pertur- 

 bations as the u meter axis rotates to larger values of 0. 



Quantitatively, little can be stated with regard to the directionality 

 response of the u component meter. In general, it appears to behave as one would 

 expect with respect to the cosine law approximation described in chapter II. 

 Further tests should be made in more fully developed seas and under more steady 

 wind conditions to better assess the directionality response of the OMDUM type 

 wave meter. 



Comparison of the Auto-Spectra of Wave Motions with Those 

 of Free Surface Oscillations - The Coastal Engineering Research Center (CERC) wave 

 staff system on the tower afforded an opportunity to compare some free surface 

 wave records tn (t) with the particle motion records produced by the wave meters. 

 Comparison of the auto-spectra is not easy; since the CERC records are analyzed on 

 an analog computer (discussed in chapter IV ) , which renders to the spectra certain 

 biasing idiosyncrasies not found in the digital Tukey spectra. 



The auto- spectra plots of the CERC free surface observations are given in 

 appendix B. Two BBELS-13 wave observations (serial 077 and 080) at 1 meter 

 depth were made almost simultaneously with CERC observations WS-U and WS-5. 

 Figure V-l8 shows the auto-spectra $y. and $ w for both BBELS observations. 

 Below the wave meter curves are superimposed traces of the two wave staff spectra 

 (WS-U and WS-5). The frequency scales (abscissas) are identical for both sets of 

 curves. The ordinate on the left (cm2 sec-1) refers to the 077 and 080 curves; 

 the ordinate on the right (cm^ sec) refers to the <^w spectra. This is a 

 difficult comparison to make, since the low frequency spectral resolution is 

 obviously much higher for the free surface spectra than for the wave motion 

 spectra. The peaks of the <£rf spectra are much narrower and fall unquestionably 

 at 200 mcps (5 sec). The ^^ spectra reach a maximum between 200 mcps (5 sec) 

 and 250 mcps (h sec). The <^^ spectra show peaks (at the higher frequency) at 

 about 250 mcps. 



Because of the incompatability of the CERC analog versus the Tukey digital 

 spectrum analysis, a comparison of c^^ with ^^ was not very productive. 



Ill 



