TRANSMISSIOiV LOSS 



369 



21. V. 2 Calibration Techniques at 

 San Diego 



The methods of calibration most commonly used 

 in target strength measurements, however, emploj' 

 calibrated transducers; at San Diego, an auxiliary 

 transducer is lowered over the side of the Jasper and 

 used with the standard echo-ranging transducer. First 

 one is used as the projector, then the other, and final 

 calibration is accomplished by referring to the con- 

 stants of the auxiliary transducer as calibrated at a 

 separate measuring station. Unfortunately, this 

 system is susceptible to errors at every step, so that 

 too much reliance cannot be placed on the accuracy 

 of the calibration. 



At San Diego, the greatest error in calibration is 

 believed to be in the measurement of the output of 

 the auxiliary transducer, which is used to calibrate 

 the echo-ranging transducer before and after each 

 run, as mentioned in Section 21.3.1. This auxiliary 

 transducer is calibrated at intervals of roughly four 

 months. Slow drifts of as much as 3 or 4 db have been 

 detected for crystal transducers between calibration 

 checks every three or four months; this drift may be 

 responsible for part of the "variation" observed dur- 

 ing target-strength runs, as described in Section 

 21.6.1. However, since it was not practicable to con- 

 trol or even measure all the factors entering into gear 

 calibration, there is no direct evidence on which to 

 base estimates of the overall calibration error of echo- 

 ranging equipment. 



21.4.3 Observed Calibration Errors 



Recent indirect evidence suggests, however, that 

 calibration errors as great as 12 db may occur. An 

 example of such large calibration errors is evident in 

 the results of San Diego echo-ranging tests on a 

 sphere.""'^ The sphere, 1 yd in diameter, was sus- 

 pended 16 ft below the surface of the ocean at ranges 

 from 24 to 166 yd; echoes from pulses from 0.5 to 

 7 msec long were received on a JK transducer. Target 

 strengths cojnputed from equation (6) in Chapter 19 

 varied from —24 to -|-3 db, approximately 12 db 

 above and below the theoretical value predicted from 

 equation (10) in Chapter 19. Although the very low 

 values are possibly the result of training errors, the 

 very high values seem rather large to be attributed to 

 errors in the estimated transmission loss, especially 

 since the values as high as 3 db were found when the 

 transmission loss was measured directly with a hydro- 

 phone placed (1) close to the projector and then 



(2) close to the target. However, the possibility that 

 the transmission loss at short ranges fluctuates by 12 

 db cannot be ruled out at the present time. This large 

 error must result either from large fluctuations in 

 short-range transmission, or from errors inherent in 

 the calibration of the gear, provided that the 

 theoretical formula in Chapter 19 for the target 

 strength of a sphere is applicable to direct measure- 

 ments. 



To provide a check on the validity of this formula, 

 an auxiliary hydrophone was placed a few yards from 

 the sphere during this series of observations and was 

 used to measure both the outgoing pulse and the re- 

 turning echo. The mean target strength of 350 echoes 

 was found by this method to be — 13.3 db, in unusu- 

 ally close agreement with the theoretical value of 

 — 12 db. A similar result was obtained at Woods 

 Hole. Thus, the 12-db discrepancy observed when 

 the JK transducer alone was used is undoubtedly the 

 result of errors in the estimated transmission loss, in 

 calibration, or in both. That large systematic errors 

 in these quantities may sometimes be present, even 

 when careful checks are provided, is suggested by the 

 anomalously high values found at San Diego for the 

 target strength of a submarine at 60 kc, and the 

 similar results obtained by Woods Hole at 12 and 

 24 kc, both reported in Section 23.6.2. 



Large errors in calibration may result from 

 (1) large-scale variability of the calibrated auxiliary 

 units employed in methods involving absolute cali- 

 bration at sea; or (2) gross deviations of the sound 

 field from the theoretical inverse square law in cali- 

 bration measurements at close ranges, because of 

 interference with reflections from the hull or from 

 other surfaces nearby. Neither of these explanations 

 seems very likely. So far no really satisfactory ex- 

 planation of the large internal inconsistencies in 

 direct target strength measurements has been ad- 

 vanced. Calibration of ship-mounted gear at sea re- 

 mains one of the most troublesome of all underwater 

 sound measurements. 



21.5 



TRANSMISSION LOSS 



It has already been pointed out that much of the 

 error in the direct measurements of target strength 

 may be due to errors in the estimated transmission 

 loss; probably a large part of the variability in ob- 

 served target strengths arises from variability in the 

 transmission loss. This quantity varies widely from 

 hour to hour and from place to place and is seldom 

 known accurately. 



