TM No. 377 



This experiment was conducted with the OMDUM I system held by a vertical 

 T-beam supported at two points on the pier. This support held the wave 

 meter rigidly in place beneath the surface waves. As in the first experiment, 

 the u meter was positioned normal to the crests of the waves. Tne meter? were 

 immersed to a depth of about 75 cm below the average trough level of the " /-es e 

 The measured waves were well-developed; i.e., they exhibited clearly defined 

 crests with a visually estimated period averaging about 1.5 seconds. The siy 

 was clear, and the wind speed was about 9 ni sec~l from the southwest, allowing 

 the waves to build up over a fetch of about 6 km. The wave height was esti- 

 mated to be about 50-75 cm, with a wave length of roughly 5-7 meters. Since 

 the water depth at the position of measurement was about 8 meters, the waves 

 measured were considered essentially "short" or "deep water" waves (as defined 

 in chapter II ). 



The voltage pulse record of the OMDUM I wave meter was processed; i.e., 

 the consecutive voltage pulse distribution was measured and converted to u 

 and w velocity components, using the procedures discussed in chapter III. 

 A Tukey spectral analysis of the data was made. The period of sampling was 

 356.4 seconds. Use of an interpolated sampling interval of 0.3 second gave 

 1188 data pairs (of u and w) for analysis. 



The variances of each velocity component Ou. and 0^, were 22.3 cm 2 sec"' 

 and 62.9 cm 2 sec" 2 , respectively. Note that there is a relatively large 

 difference in the magnitude of the variances of the orthogonal velocity com- 

 ponents u and w. The covariance function at zero lag li'w' between the 1188 

 pairs of vertical and horizontal velocity components was calculated as -7.8 

 cm 2 sec" 2 , providing an estimate of the magnitude and sign of the observed 

 Reynolds stress. The linear correlation coefficient V* between u and w was 

 -0.21. 



The auto-covariance and covariance spectra are shown in figure IV-2. 

 The auto-spectra of both velocity components ( $u, and $u, ) show peaks 

 at the frequency of the observed waves. The cc-spectrum is shown by the 

 negative curve. The area between the horizontal axis and the solid curve 

 is the (negative) covariance. 



The covariance function and the co-spectrum could have been caused by 

 some spurious effect in the meter system. As a cheek to examine the extent 

 to which the correlation function and the individual time series were stationary 

 (see chapter II), the spectral analysis was re-run using varying percentages of 

 the data. The values of the covariance function at zero lag were as follows: 



Number of Percent of Total Covariance ( ttW ) 

 Data Points Data Pieces cm 2 sec' 



1188 

 1176 



792 



5^0 



73 



