TRANSMISSION IN ISOTHERMAL WATER 



105 



10^ 10° 



FREQUENCY IN CYCLES 



10' 



10" 



Figure 17. Dependence of attenuation coefficient on frequency. The points on the curve are from the following refer- 

 ences: n NRL, 13 and 16; V UCDWR, 14; O UCDWR, 18; . CNRC, 19; • Fresh Water, 20; H Fresh Water, 22; 

 A Fresh Water, 23; A WHOI, Chapter 9 of this book. 



coefficients of 18, 26, and 32 db per kyd, respectively, 

 for an assumed amplitude reflection coefficient of the 

 bottom of 0.5 (energy loss of 6 db per reflection), cor- 

 responding to the SAND-AND-MUD bottoms over 

 which the measurements were made. A change of the 

 reflection coefficient by 0.2 in either direction changes 

 the attenuation coefficient by about 2.5 db per kyd 

 in the same direction; this variation may be taken 

 as a rough estimate of the probable error of the re- 

 sults. Owing to this high probable error, the values of 

 2.0 and 7.0 db per kyd, found at 24 and 40 kc re- 

 spectively, are of relatively low weight and may be 

 disregarded. 



At frequencies between 500 and 2,500 kc, extensive 

 measurements of attenuation have been made by the 

 Canadian National Research Council." A projector 

 was mounted on a dock in Vancouver Harbor in 13 to 

 25 ft of water. The receiver was also mounted on the 

 same dock at distances varying up to 100 ft. As a 

 result of the high directivity of the projector, surface 

 and bottom-reflected sound were largely eliminated. 

 The slope of the transmission anomaly was measured 



to give an attenuation coefficient at each frequency. 

 Relative probable errors of these coefficients, esti- 

 mated from the reproducibility of the results, 

 averaged about 7 per cent. No temperature measure- 

 ments were made. Over such short ranges any gradi- 

 ents would have had a negligible effect. 



No measurements at frequencies above 3 mc are 

 available for sound in the ocean. However, such 

 measurements have been made in the laboratory.™"^' 

 Those of reference 20 extend down to 2.8 mc, where 

 the values found are of the same order of magnitude 

 as those determined in the ocean. Other determina- 

 tions of absorption in fresh water in the frequency 

 range between 200 and 4,000 kc are about ten times 

 as high as those found in the sea.^^"^^ These fresh- 

 water measurements are not in good agreement with 

 each other and may be affected by systematic errors. 

 Since the sea-water values taken from reference 16 

 were made over a much greater sound path, these 

 should be much more reliable, and in any case, consti- 

 tute better evidence for the attenuation of sound in 

 the sea; the fresh-water measurements in references 



