TRANSMISSION IN ISOTHKRMAL WATER 



103 



-20 



Z 



o 



z 

 < 



20 



40 



1000 2000 



RANGE IN YARDS 



Figure 15. Individual transmission anomalies in isothermal water. 



3000 



transmission anomaly shown in Figure 14. Only runs 

 with a hydrophone at 16 ft have been included. In the 

 curve for 24 kc, taken from reference 14, runs made 

 with a total temperature change of less than 0.3 de- 

 gree between the surface and 80 ft were included. 

 Since a general analysis shows that temperature dif- 

 ferences as small as 0.2 degree in the top 30 ft do not 

 affect transmission anomaUes appreciably, probably 

 much the same curve would be obtained if only those 

 runs in water isothermal to less than 0. 1 F were con- 

 sidered. In the curve for 60 kc (taken from Volume 7 

 of Division 6), all runs made with an approximately 

 isothermal surface layer were included. The slopes of 

 the two average curves in Figure 14 give attenuation 

 coefficients of 4.0 db per kyd at 24 kc and 13.5 db 

 per kyd at 60 kc. 



In Figure 15 are plotted all the individual points 

 used at 24 kc, with solid lines connecting the median 

 points and dashed lines connecting the upper and 

 lower quartiles. The average quartile deviation shown 

 in Figure 15 is 2.6 db. This is about the same as the 

 deviation of any single transmission anomaly curve 

 in isothermal water from a straight line, and is also 

 about the same as the experimental error, discussed 

 in Chapter 4, in the determination of the transmis- 

 sion loss from the average of five observations. 



Other sets of measurements give similar values of 

 the attenuation coefficient a. Many transmission runs 

 were made some time before the war by NRL. '"~" 

 In these runs, the range was commonly opened from 

 less than 1,000 to more than 10,000 yd. These data 



were originally plotted in terms of intensities rather 

 than transmission anomalies. It was found that in a 

 considerable number of runs, plots of sound intensity 

 (in decibels) against range gave essentially straight 

 lines beyond 1,000 yd. The measurements were re- 

 analyzed and the results reinterpreted in terms of 

 transmission anomalies. '^ 



For the data reported in reference 16, all the 

 straight-line graphs were obtained when the water 

 was isothermal (to +0.1 C) from the surface to more 

 than 100 ft. The average attenuation coefficients 

 found at 17.6, 23.6, and 30 kc were 1.8, 4.4, and 6.5 

 db per kyd, respectively. Half the observed values 

 lay within + 1 db per kyd of these average values. 



The NRL and UCDWR measurements described 

 above are the only ones in deep water which have 

 been analyzed in terms of the temperature structure 

 present at the time the measurements were made. A 

 considerable body of other measurements have been 

 made to determine the attenuation coefficient, but 

 these are less reliable. 



Measurements of bottom-reflected sound in shallow 

 water have been used to determine the attenuation 

 coefficient resulting from absorption and scattering 

 in the volume of the ocean. In particular, sound has 

 been sent out vertically, and the strength of the echo 

 received in different depths of water used as a meas- 

 ure of attenuation in the water. Since these measure- 

 ments depend entirely on the reflection coefficient of 

 the bottom, they can give results on attenuation 

 only if it is assumed that the reflection coefficient of 



