108 



DEEP-WATER TRANSMISSION 



line. For the data shown in this figure, half of the ob- 

 served values of a fall within about 0.6 db per kyd of 

 the average values. This is so close to the probable 

 error of 0.5 db per kyd for a single determination of a 

 that the result may be entirely due to observational 

 errors. Certainly the reduced scatter can be attrib- 

 uted to the relatively uniform conditions prevailing 

 during these tests. The observed attenuation coef- 

 ficients for the deep hydrophone will be discussed in 

 Section 5.3. In addition, the relatively small scatter 

 shown in Figure 15, about the same as the observa- 

 tional error in the measurement of transmission loss. 



o 



I- 

 < 



'l 5 20 40 60 80 95 99 993 



PER CENT OF ALL ATTENUATIONS 

 GREATER THAN THE VALUE PLOTTED 



Figure 20. Distribution of attenuation coefficients 

 for Point Conception runs. 



suggests that in isothermal water the attenuation 

 may be relatively constant. Further information is 

 required, however, before any very definitive conclu- 

 sions can be drawn about the variability of the at- 

 tenuation coefficient when the temperature of the 

 water in the upper 100 ft of the sea is approximately 

 constant. Most of the UCDWR data have not been 

 analyzed with this purpose in mind. In reference 13, 

 where a high variability of a is found, the runs in 

 isothermal water are not treated separately. Also, 

 some of the runs included do not extend out very far; 

 when marked peaks in the anomaly curve are present, 

 the attenuation coefficient for such runs may be as 

 low as 1 db per kyd. An examination of a sample run 

 of this type, shown in Figure 21, indicates that such 

 values are not necessarily indicative of the attenua- 

 tion over longer ranges. Thus, these data do not cast 

 much fight on the variability of the attenuation 

 coefficient in isothermal water. 



It is tempting to assume that the values shown in 

 Figure 17 represent pure absorption, and that any 

 variations from these values found in approximately 

 isothermal water represent distortion of wave front 

 by temperature gradients too small to be detected 

 reliably on the bathythermograph. More accurate 

 data on transmission in mixed water and more ac- 

 curate thermal measurements would be required to 

 test such an hypothesis. 



5.2.4 Short-Range Transmission 



The goal of transmission studies is to relate the 

 sound intensity at any range to the sound output of 

 the source. Most transmission measurements at sea, 

 however, do not measure the sound level closer than 

 about 100 yd from the source. In principle it should 

 be possible to measure the absolute sound level in the 

 water, and also to measure the absolute level 1 yd 

 from the projector; in practice, the measuring equip- 

 ment has apparently not been sufficiently stable to 

 make these absolute measurements possible. Thus 

 transmission measurements give only relative sound 

 levels and may be used to give the sound level at long 

 range relative to the level at several hundred yards. 

 The methods used in computing transmission anom- 

 alies are discussed in some detail in Chapter 4. For 

 most of the data discussed here, the transmission 

 anomaly at short range, usually about 100 or 200 yd, 

 has been taken equal to zero. Thus, to find true 

 transmission anomalies at long range requires infor- 

 mation on the true value of the transmission anomaly 

 at ranges around 100 yd. 



Since refraction can be ignored at such close ranges, 

 the transmission anomaly for the sound passing di- 

 rectly from projector to hydrophone must be very 

 close to zero. If the surface were perfectly flat, sur- 

 face-reflected sound would, on the average, double 

 the sound intensity at ranges of several hundred 

 yards, provided that the intensity is averaged over 

 the interference pattern discussed in Section2.6.3;the 

 transmission anomaly A at these ranges would then 

 be —3 db. Sound intensity measurements between 

 1 yd and several hundred yards would then show a 

 gradually decreasing transmission anomaly as the 

 range increased and as surface-reflected sound ap- 

 proached the same average strength as the direct 

 sound. 



However, the sea surface is never perfectly flat, 

 and this fact may be expected to alter the simple 

 relationships to be expected for a flat surface. Al- 



