TM No. 377 



is thus tempting to speculate that these data show evidence of interaction "between 

 a tidal current and the sea bottom, producing both a mean shear ^/32 and a 

 gradient of turbulent energy. However, the existence of such a gradient is not 

 certain; since u was not measured simultaneously at the two depths, but over a 

 20-minute time lag. These measurements do demonstrate the use of the ducted 

 meters to study the bottom turbulence. 



The problem of assuming stationarity in the time series records was dis- 

 cussed in chapter III. Ideally, in a purely stationary process: 



j^ $ff) = o Mtp sL I £#)<# ^ |2" 2 S Q 



(v-9) 



where the frequency range from to «p^ covers the region of interest. One way to 

 estimate the degree of stationarity in a process is to make a relatively long 

 record, divide it into smaller subdivisions, and examine their spectra. The 

 degree of similarity of the auto-spectra and a comparison of the variances would 

 be indicative of the stationarity of the whole record. During BBELS-11, a 

 single record was made at 0.5 meter for 11 minutes (see table IV- 3 j 057A and 057BJ. 

 From this record, which contained 3? 300 data points at 0.2-second intervals, two 

 sub-samples were abstracted: the first had 1001 points, and the second 1070 points, 

 The auto- spectra $to are shown in figure V-l6. The degrees of freedom DF for 

 both samples was about h-2. Referring to figure III- 3, the ratios of 80 percent 

 confidence limits are above 0.7^- and below 1.29 with respect to the average value. 

 This amplitude is shown in figure V-l6. Examination of spectra and vistial obser- 

 vation of sea and wind conditions indicated that the 5-10 minute sampling period 

 of wave motions should be relatively unaffected by gross time variations in the 

 wave structure. Sudden wind shifts did occur, but these were unusual and were 

 well-monitored as they happened. 



Short period fluctuations in the horizontal current were also found to occur 

 at BBELS (see chapter IV). This phenomenon, although not visibly affecting the 

 wave structure, can affect the spectra; especially the u(t) spectrum. Because of 

 this, it is reasonable to assume that in the 0MDUM (u,w) record, w(t) has a higher 

 degree of stationarity than u(t). 



Auto-Spectra of Off-Angle Wave Measurements - Calibration of 0MDUM III for 

 off-angle response (see chapter II ) was done with a simulated steady flow in the 

 towing tank. This simulated motion was not very wave-like. Because of this, an 

 attempt was made to roughly imitate the off-angle test in actual wave conditions 

 at the BBELS. 



109 



