130 



DEEP-WATER TRANSMISSION 



9 

 z 



o 



< 



O-r' DEPTH TO THERMOCLINE IN FEET 



Figure 48. Attenuation coefficient below the therraocline. 



Figure 48 shows an increase of attenuation below 

 the thermocHne with decreasing thermocline depth. 

 This effect has already been noted in Section 5.3.4 for 

 thermocline depths greater than 40 ft ; the change of 

 R^(i with layer depth (Figure 29) was shown to be 

 consistent with the theoretical curve based on equa- 

 tion (13). It is apparent from Figure 47 that this same 

 effect persists to much shallower layers. A comparison 

 of Figures 47 and 48 indicates that layer effect in- 

 creases steadily with decreasing depth of the ther- 

 mocline. This result is also consistent with expecta- 

 tions based on equation (13). 



The increase of the attenuation coefficient in the 

 isothermal layer as the depth of the layer decreases 

 is quite marked in Figure 47. It may be noted that 

 the values shown for Dy between 20 and 30 ft are not 

 inconsistent with the attenuation coefficient of about 

 13 db per kyd found from the upper curve in Figure 

 40, drawn for D2 between 20 and 30 ft. The origin of 

 this high attenuation is hard to explain. As sound 

 travels along through the layer of nearly isothermal 

 water, with the sound rays continually reflected from 

 the surface and distorted by temperature microstruc- 

 ture, it may be expected that a certain fraction of the 

 sound would be bent out of the isothermal layer in 

 each yard of sound travel. Any such sound reaching 

 the thermocline will be bent down so sharply that it is 

 unlikely to return to the isothermal layer. It is possi- 

 ble that a quantitative theory along these lines, based 



on more accurate information on the properties of the 

 isothermal layer, may explain the observed decrease 

 of attenuation with increasing thickness of the layer. 

 In any case there is little question as to the reality 

 of the effect noted in Figures 47 and 48. 



RANGE IN YARDS 



3000 



40 



. MIKE, 30= Dj -=40 

 CHARLIE, 20 S D2<30 



Figure 49. Average transmission anomalies for 

 MIKE and CHARLIE patterns (hydrophone shallow). 



Whether the scatter evident in these figures is 

 greater than can be explained by the observational 

 scatter of all observed transmission anomalies is not 

 evident from an examination of reference 13. It has 

 already been noted, in Chapter 4, that for most of the 

 UCDWR data transmission anomalies determined by 

 averaging 5 successive received pings have a probable 

 error of about 2 db as a result of hydrophone direc- 

 tivity, training errors, and sampling errors; the errors 



