TRANSMISSION WITH NEGATIVE GRADIENTS NEAR SURFACE 



125 



If the temperature difference between and 30 ft 

 is about 1 F, and the surface temperature is about 

 70 F, then for sound of 24,000 cycles equation (16) 

 gives an attenuation of aljout 0.1 db per yd, or about 

 100 db per kyd. This is at least twice as great as the 

 values usually obtained. The discrepancy seems 

 somewhat too great to be explained as observational 

 error, although the fact that observed and predicted 

 attenuations are of the same order of magnitude is 

 suggestive. It is possible that the presence of thermal 

 micro.structure may explain this difference. No at- 

 tempt has been made to correlate the observed at- 

 tenuation across the shadow boundary with either 

 the frequency / or the velocity gradient dc/dy at the 

 surface. 



Scattered Sound in the Shadow Zone 



Most transmission anomaly plots for NAN pat- 

 terns are characterized by a nearly constant trans- 

 mission anomaly well beyond the computed boundary 

 of the shadow zone, amounting to between 40 and 

 60 db and extending out to the limit of measurement 

 at about 5,000 yd. 



An examination of the oscillograph record shows 

 that the signal received at ranges between 1,500 and 

 about 3,000 yd bears very little relation to the shape 

 or length of the original pulse. Figure 44 shows the 

 signals received at different ranges when a marked 

 negative gradient was present at the surface. At 

 moderately short ranges, less than 1,000 yd, the re- 

 ceived signal reproduces the outgoing pulse rather 

 faithfully. At moderate ranges into the shadow zone, 

 however, the signal has an appearance similar to that 

 of reverberation, and is much prolonged, as shown in 

 trace C made at 1,340 yd. At these intermediate 

 ranges, the use of peak amplitudes in reading each 

 signal gives a value about 7 db higher than the use of 

 average intensities. For coherent 100-msec signals, 

 the difference between peak amplitude and rms am- 

 plitude is negligible. At slightly longer ranges even the 

 few traces of the direct pulse, visible for some of the 

 signals in trace C, completely disappear. At the long- 

 est ranges the signal begins again to resemble the 

 emitted pulse, as shown in trace D. 



• The intensity of the sound received in the shadow 

 zone increases with increasing pulse length. The 

 observed difference in intensity for 100-msec and 

 10-sec pulses is shown in Figure 45. In the shadow 

 zone at intermediate range, where the received signal 

 is prolonged and incoherent, the effect is large, 

 amounting to between 4 and 8 db. At very long 



RANGE IN YARDS 

 1000 2000 



I 50S h-elOO FEET 



II IOOgh-200 FEET 

 in ZOOS h -=300 FEET 



W 300^h-=400 FEET 



Z 



o 



Figure 43. Average transmission anomalies for NAN 

 patterns with hydrophone deeper than 50 feet. 



ranges and within the direct sound field at shorter 

 ranges the signal is coherent and the difference is 

 small. This small difference results from the fact that 

 the peak amplitude of a long signal, which fluctuates 

 from minimum to maximum values many times, 

 tends to be greater than the peak ampUtude of a 

 short signal, which may be altered by fluctuation to 

 a low value. 



