TM No. 377 



The & trace shows a rapid variation between C and 1^0°, followed by a 

 leveling off for 2-3 hours, and then a gradual climb to 3^0° , The 5-minute 

 samples do not indicate rapid or high frequency current direction variations, 

 However, the current speed S ? under this higher resolution (5 versus 20 

 minute sampling interval)., still contains rapid oscillations » 



Only certain incremental values of j occur This is due to the resolu™ 

 ti on of the film reader, which can be read only to within l° c The Savonius 

 rotor calibration of the Braincon meter is 0=403° per centimeter per second 

 for a ^-minute interval averaging time. Since the radarc on the film was 

 read to the nearest whole degree, the resulting speeds occur only as multi- 

 ples of 2J43 cm sec-1., This limitation on the number of velocity values is 

 due to the small angular displacement caused by the short integration time. 

 Over longer sampling intervals the current variation appears more continous 

 (see figure TV-6). For sampling over short intervals (5 minutes or less) the 

 instrument could be improved by: (l) geering the rotor rotation to give a 

 larger angular swing of the radarc per unit speed; and (2) using finer grained 

 photographic film, together with an improved optical film reader to permit a 

 reading of the radarc to fractions of a degree.- 



The speed trace shows abrupt variations of 7~10 cm sec"- 1 - occurring 

 within 5-10 minutes. This explains the differing mean values of u (in table 

 IV-3) for the wave measurements, which were taken ax successive 5-10 minute 

 intervals.; The strongest current occurs while 1? is varying most slowly, and 

 conversely. 



An auto-spectrum <$g was generated to further examine motions having an 

 ultra-tidal frequency; i.e. . in the range Q 50 mcpks < -f < l600 mcps} or 

 C320 min > T > 10 man}., A sample containing 1000 points, from 2k November 

 (1200 hours) through 2§ November (23^0 hours), was analyzed,, The results 

 are shown in figure IV~ii„ The ordinate scale is in terms of cm sec"'" per 

 mcpks. The abscissa scale is in mcpks and hours. 



The spectrum portrays a rather uninteresting pattern with most of the 

 energy concentrated well, above 200 mepkSo This undoubtedly is the influence 

 of the 6 and 12-hour peak oscillations associated with the tape shown in 

 figure 17-6* There are small peaks occurring at about 230 and 500 mcpks, and 

 two other peaks at 1^00 and 1600 mcpks , These four peaks are slightly larger 

 than the 80 percent confidence range shown on the graph* 



In general, the spectrum beyond 200 mcpks is approximately constant or 

 white., This white energy is not negligible, however-, for it constitutes 

 about 31 percent of the total variance, The phenomenon of aliasing could be 

 contributing energy from frequencies less than the Nyquist value of fn~lob77 

 mcpks (6 cycles per hour) to those frequencies from. 250 mcpks up to fn« 



In summary., these two series of measurements (I and II) show a predictable 

 pattern of current speed and direction in relation to the free surface height 

 of the tide.. However, there occurs a relatively high frequency "noise" in the 

 current speed ^ .. which contains relatively large amounts of spectral energy 



3U 



