MARTIN/PERRONE : GEOGRAPHICAL VARIATION OF AMBIENT NOISE IN THE OCEAN 

 FOR THE FREQUENCY RANGE FROM 1 HERTZ TO 5 KILOHERTZ 



ridge. While this represents a small data set, the pronounced differ- 

 ences at the two sites suggest that longer term observations would 

 yield the same general conclusions. 



Figure 16 shows a composite of results discussed so far. But 

 now consider data from New Zealand (Figure 16) obtained in an area 

 known to have almost no shipping and blocked from the broad open 

 ocean areas by oceanic ridges. Note that the spread of data for 

 the range of wind speeds obtained is independent of frequency from 

 10 to 1,000 Hz and that the high wind speed curve is similar to that 

 observed in Bermuda. Also, the shape of the curves below 100 Hz is 

 inconsistent with those appearing in the literature for wind generated 

 noise. Measurements in these quiet areas to frequencies of 1 Hz or 

 lower are important in deteirmining the wind generated mechanism (s) 

 giving rise to the ambient noise in the various frequency regions. 

 Long time series are important so that results and conclusions will 

 not be criticized on the basis of sample size. The Bermuda measure- 

 ments, discussed in part earlier, were obtained over a 1-year period. 



Figure 18 presents a phenomenological occurrence of the surface 

 image effect on 1/3-octave noise in the presence of a directional 

 noise field. For AUTOBUOY Dive 1 (Figure 15) , the system slowly 

 ascended to the near surface and hovered at that shallow depth for 

 about 30 minutes. The abscissa in Figure 18 is scaled both for time 

 and for depth as determined by a depth gauge in the AUTOBUOY. The 

 ambient noise level is shown to vary slowly with decreasing depth up 

 to approximately 150 feet and then to fall off rapidly for the lower 

 frequencies in particular. If one assumes a surface interference 

 effect in each 1/3-octave band, the results agree quite well with 

 the ass\imption that the noise is uniformly directional from approxi- 

 mately to 10 degrees. While this is a first order assumption, the 



830 



