Clark and Yarnall 



lines is being investigated. Spectral estimates at the high fre- 

 quency end of the fluctuation spectrum (surface wave and short period 

 internal wave signatures) are impractical to compute over very long 

 test periods. For example, something on the order of 107 data points 

 would be required to compute surface wave estimates over the entire 

 140 days of LCT-1, a prohibitively expensive data load for a power 

 spectral calculation. The data have been sufficiently "clean", 

 however, to conduct visual studies of both surface wave and short 

 period internal wave signatures over extended time periods, with re- 

 sults that are judged to be a useful spectral measure of these effects. 



SURFACE WAVE SIGNATURES 



The spectral signatures of wind driven surface waves are 

 almost always present in the acoustic data from any of the hydrophones. 

 The magnitude of the effect is relatively small and, as noted above, 

 is seen as a high frequency "hash" or "smear" in most of the data 

 records presented. Both signal amplitude and phase are effected and 

 display substantially identical fluctuation spectra in the data that 

 have been examined carefully. 



A detailed visual study was made of the surface wave signa- 

 tures in signal amplitude during a four day test from January 25-29, 

 1965. The hydrophone involved is designated as "A" in Figure 2. The 

 signal was cw at 420 Hz, the case for all data discussed in this paper. 



A random sampling of the test records for the 4-day interval 

 showed the average peak to peak pressure variation to be about 5 dB . 

 The maximum variation noted was 12 dB. The minimum was 1 dB, which 

 is close to the resolution limit of the instrumentation. 



The periodicity of the fluctuation was predominantly 3-4 

 seconds over this testing period, but did increase at times to as 

 much as 12 seconds. The 3-4 second period was evident in the receiv- 

 ed signal throughout the test with the exception of one 10 hour inter- 

 val, on 26 January, during which the period was primarily 8 seconds 

 with some 6 second periods also present. Unfortunately, no surface 

 wave height data were available at the time. Changes in the pre- 

 dominating period or in the magnitude of the amplitude variations 

 tended to be gradual. However, it was not unusual to find several 

 isolated cases of 5 to 7 second periods included in an interval which 

 exhibited predominately 3-4 second fluctuations. No periods of less 

 than 3 seconds were noted during the test. 



The similarity in the spectral composition of the signal 

 amplitude and phase fluctuation was quite apparent and no attempt 

 was made to tabulate the periods of the phase fluctuation. The peak 

 to peak range of the phase fluctuation varied typically from the re- 

 solution limit of about 5° up to 30 or 40°, and increased in rare 

 case to as high as 180°. 



321 



