302 



DEEP-WATER REVERBERATION 



reason for the rapid decrease is not understood, but 

 that there are a number of factors which may play a 

 part. Very hkely, all the physical factors which have 

 been discussed previously are included to some de- 

 gree. In addition, there may well be other causes 

 which have not been considered. It should be noted 

 that at ranges less than 500 yd the measured rever- 

 beration for wind speeds less than 8 mph decreases 

 only as the inverse first power of the range in Figure 

 23. This rate of decay is even slower than the pre- 

 dicted inverse square decay of volume reverberation. 

 No definite explanation has been offered for this 

 feature of Figure 23, but it might be caused by a 

 gradual increase in the value of the volume-scattering 

 coefficient as the deep layer is approached. This effect 

 would, of course, be noticeable only in sea states so 

 low that volume reverberation can be measured at 

 short ranges. 



14.2.2 Dependence on Ping Length 



The theoretical formulas (22), (39), and (52) for 

 volume, surface, and bottom reverberation in Chap- 

 ter 12 all have the reverberation intensity propor- 

 tional to the ping length. The theoretical assumptions 

 required to obtain this result have been discussed in 

 Chapter 12. In this subsection we shall discuss 

 whether or not this strict proportionality may be 

 expected in practice. The only data bearing on this 

 question are reported in reference 1, Section IV, and 

 are summarized later. Unfortunately reference 1 does 

 not state whether the reverberation studied was 

 volume, surface, or bottom reverberation, but rever- 

 beration received from ranges as low as 100 yd and 

 as great as 5,000 yd was included in the analysis. 

 However, ranges less than five times the ping length 

 were not included. 



Figure 29 shows, qualitatively, that the reverbera- 

 tion intensity increases with increase in the signal 

 length. In that illustration, a record A of reverbera- 

 tion following a 70-msec ping is compared with a 

 record B of reverberation following a 10-msec ping. 

 The attenuator settings were the same for both cases; 

 however, because of the higher level of the 70-msec 

 reverberation, each attenuator step was removed a 

 little later for the 70-msec ping. For this reason, the 

 records are directly comparable only in the intervals 

 1-1, 2-2, and 3-end in which the amplification is the 

 same for both records. In these intervals, the 70-msec 

 reverberation is clearly much higher than the 10-msec 

 reverberation. 



To test quantitatively the predicted relation 

 between reverberation intensity and signal length, 

 sets of data were taken on two successive days 

 of the reverberation following pings of lengths, 

 very nearly 10, 20, 40, and 70 msec. Ten pings 

 were measured on each day for each signal length. 

 For each ping length, the average reverberation 

 amplitude was measured at a set of logarithmically 

 equispaced positions, by the band method. The 

 squares of these average amplitudes were assumed 

 proportional to the average reverberation inten- 

 sities, in accordance with the usual procedure de- 

 scribed in Chapter 13. 



The agreement between theory and experiment is 

 shown in Figure 30. In that illustration, ten times the 

 logarithm of the ratio of any two ping lengths is taken 

 as the abscissa, and the decibel difference between 

 the corresponding measured reverberation levels is 

 taken as the ordinate. "^ If reverberation intensity 

 were in fact exactly proportional to the ping length, 

 all the observed points should lie on the 45-degree 

 straight line drawn in the figure. In Figure 30 the 

 points for which the longer ping length of a pair was 

 20, 40, and 70 msec are designated differently so that 

 any systematic departure depending on ping length 

 can be discerned. On the whole, the agreement in 

 Figure 30 between theory and experiment is satis- 

 factory. For some reason, the agreement is better for 

 the ratios involving the shorter signals (10, 20, 40 

 msec) than for the ratios including the longest signal 

 (70 msec). 



The deviations from the straight line in Figure 30 

 are not too great to be ascribable to experimental 

 error. Thus these data give no reason for doubting 

 the prediction of equations (22), (39), and (52) of 

 Chapter 12, that, under the conditions specified in 

 that chapter, reverberation intensity should be pro- 

 portional to ping length. However, in view of the im- 

 portance of knowing the dependence on ping length 

 for comparison of reverberation measurements made 

 with different ping lengths, and for the determination 

 of scattering coefficients, further investigation of this 

 dependence is desirable. The measurements should 

 be repeated for a wider range of ping lengths and for 

 all types of reverberation. 



<> The reason that the abscissas of some of the pairs of 

 points in Figure 30 are not the same is that the signal lengths 

 as measured from the film records were not exactly the same 

 on both days. However, the equipment was set on each day 

 for nominal ping lengths of 10, 20, 40, and 70 msec. 



