OCEANIC TURBULENCE AND AMPLITUDE FLUCTUATIONS IN ACOUSTIC SIGNALS 



Laurence C. Breaker William D. O'Neil, Jr. 



U. S. Naval Oceanographic Office The Bissett-Berman Corporation 

 Washington, D. C. 20390 Santa Monica, California 90404 



INTRODUCTION 



This paper deals with an experiment in the fluctuations in 

 transmission loss for underwater acoustic signals. There are con- 

 siderable gaps in our empirical knowledge of these fluctuations and 

 this in itself might be considered sufficient reason to conduct such 

 an experiment. This experiment, however, proceeded out of a 

 particular theory of the origin and nature of transmission-loss 

 fluctuations. A certain knowledge of this theory (which will be 

 glossed, not altogether accurately, as the "turbulence induced" 

 theory of fluctuation in transmission loss or amplitude) is necessary 

 if one is to understand why the experiment was planned, executed, and 

 analyzed in the manner about to be described. Moreover, this theory 

 (to the extent to which it may actually be valid) offers a means to 

 predict some operationally significant sonar-performance parameters. 



Thus this paper will open with a survey of the theoretical 

 background to the experiment. Next will come sections dealing with 

 the experiment itself, the planning which preceded it, and the data 

 reduction which followed. As data reduction and analysis is not, at 

 this writing, complete, report of results and conclusions must 

 await a later occasion. Persons having especial interest in these 

 matters, however, are invited to correspond with the authors. 



THEORETICAL BACKGROUND 



Suppose an omnidirectional source is placed some distance 

 below the surface of the ocean and a number of hydrophones are 

 arrayed about it, each at a different depth but all at a common slant 

 range. Now a very short pulse is transmitted from the source. If 

 in a typical case one observes the amounts of energy received by the 

 various hydrophones over direct paths (i.e., excluding any boundary 

 reflections) it will be found that they differ significantly one 



366 



