TM No. 377 



also that these higher frequency motions are phase shifted similar to those of 

 serial 023 and 025 (figures V-2 and V-3). However, the w trace clearly contains 

 a strong lower frequency component having an estimated period of from 5-7 seconds. 



The last sample record (figure V-5) was made using LIMDUM I, measuring w at 

 depths of 0„0 and 2.0 meters. During this record from serial 086 (BBELS-lU), 

 winds were 8.0 m sec" 1 from the SSW and no swell was observed. 



Comparison of the 0.0-meter record (top) with the 2„0-meter record (bottom) 

 indicates a high degree of positive correlation (or in-phase motion) between the 

 observed motion at the two levels. The attenuation of the motion with depth is 

 also evident in the amplitude differences of the two traces. There^is also 

 evidence of what appears to be a beat effect (from 5 through 25 sec) occurring on 

 both records. 



The velocity data portrayed in figures V-2 through V-5 would seem to indicate 

 that the time variation of Eulerian wave velocities measured at a point is not 

 particularly sinusoidal. The velocity functions appear rather to have somewhat 

 flattened crests with abrupt slopes and jagged narrow troughs. It is interesting 

 to note that this is roughly the inverse description generally given of the 

 profiles of waves (e.g., Kinsman i960). This odd shape associated with the 

 velocity functions might result from some property of the wave meter. This does 

 not seem to be the case, since the OMDUM II system faithfully reproduced an 

 impressed sinusoidal oscillation simultaneously registered by an accelerometer 

 (see figures II- 3 and II-U). However, some further examination of possible 

 distortion of the wave motion patterns by the wave meter seems called for at this 

 point. 



It will be recalled (see chapter II ) that the OMDUM III system has a threshold 

 velocity of about 7-8 cm sec" 1 . This means that no output is registered when the 

 speed of flow through the cylinder is less than this value; thus, a gap is 

 periodically produced in the data as the wave motions oscillate. On the other 

 hand, a high concentration of data points is produced in regions of high absolute 

 velocity (this is shown in figure III-8). However, with interpolation, a uniform 

 distribution of data points (as a function of time) is obtainede The velocity 

 data were sampled at high enough frequencies to provide proper resolution, so that 

 the plot of the interpolated velocity data should provide a reliable picture of 

 the velocity (see figure III-8). This is particularly true since the velocity 

 functions appear to change rapidly as they pass through zero, whereas they linger 

 at the higher absolute values (vis., a boxcar signal as the extreme case,;. Hence, 

 over the whole record relatively few sub-threshold data points have to be inter- 

 polated. 



From examination of the actual velocity traces, it becomes obvious that, m 

 order to properly compare the various records and individual u and w motions, one 

 must progress further to the analysis of the variances and spectra. 



Variances of the Observed Wave Motions - The results of the ^ series of 

 measurements made in Narragansett Bay (discussed in chapter IV ) indicated a strong 



97 



