106 



DEEP-WATER TRANSMISSION 



o 



o 



I- 



1600 



1800 

 FEB 14 



2000 



2200 



0000 



0200 



0400 0600 



FEB 15 



0800 



1200 



1400 



TIME OF DAY 



Figure 18. Variability of attenuation coefficient during one day. 



24, 25, and 26 may therefore be ignored in the present 

 discussion. 



At frequencies below 14 kc no data on deep-water 

 attenuation in the surface layers of the sea are avail- 

 able. The long-range measurements of explosive 

 sound propagation, discussed in Chapter 9, give an 

 upper limit on the attenuation coefficient deep in the 

 ocean. Although the data are uncertain, the results 

 quoted in Chapter 9 indicate that the attenuation at 

 2,000 c is probably less than 5 X 10~^ db per kyd. 



These different determinations of the attenuation 

 coefficient are combined in the plot of a against fre- 

 quency shown in Figure 17. The dashed line gives a 

 curve of best fit drawn through the plotted points. 

 The solid line in this figure gives the value of a to be 

 expected from viscous damping, taken from Section 

 2.5. Evidently, at frequencies above 1 mc the at- 

 tenuation coefficient is three to four times the classi- 

 cal value. In a UCDWR memorandura,^'' this dis- 

 crepancy is attributed to an additional viscous force 



proportional to the rate of compression of the water. 

 Such a force has usually been neglected in hydro- 

 dynamics, since ordinarily water flows like an in- 

 compressible fluid. Although no tests of such an 

 hypothesis liave been suggested, it is entirely possible 

 that this "compressional viscosity" may be respon- 

 sible for the observed values of a at frequencies above 

 1 mc. 



No explanation has yet been advanced for the ob- 

 served values of the attenuation at somewhat lower 

 frequencies. Since all the measured values between 

 10 and 100 kc were obtained in temperate latitudes in 

 water well above the freezing temperature, it is possi- 

 ble that these values are not applicable for all oceano- 

 graphic conditions. It is still not wholly certain that 

 the attenuation observed for supersonic frequencies 

 in isothermal water is entirely the result of absorp- 

 tion rather than scattering; however, the weakness of 

 scattered sound observed for backward scattering 

 (reverberation) and for forward scattering (incoher- 



