378 



DIRECT MEASUREMENT TECHNIQUES 



recorded oscillographically at San Diego.' The echo 

 amplitude was measured at small intervals along the 

 length of the echo, squared to give the echo intensity, 

 and averaged over the echo length as closely as the 

 echo length could be estimated; then the enclosed 

 area was calculated. Division of this area by the 

 signal length gave a new intensity, the intensity 

 which presumably would have resulted if the echo 

 length had equaled the signal length. This sample 

 analysis, although based on data not sufficient to 

 warrant definitive conclusions, showed an insignifi- 

 cant difference between peak echo intensities and 

 mean echo intensities corrected for signal length. 



In general, however, the peak echo intensity differs 

 from the uncorrected mean echo intensity, and this 

 difference is a function of the signal length. It was 

 pointed out in Sections 19.3 and 20.7 that for long 

 pulses, the echo will reproduce the signal envelope 

 while for short pulses fluctuations in intensity will re- 

 sult in an irregular structure, where sharp peaks 

 stand out against a weak background. In the latter 

 case, the peak echo amplitude may be considerably 

 different from the mean echo amplitude, and may 

 vary with signal length quite differently (see Sec- 

 tion 23.5.1). 



The variability of echoes is responsible for a large 

 part of the uncertainty in the echo level and trans- 

 mission loss values which are used to compute target 



strengths. Since echoes are often so irregular that 

 visual estimates of peak intensities are, at best, in- 

 telligent guesses, UCDWR observers estimate that 

 systematic errors of as much as 2 or 3 db may result 

 from the difference in personal judgments of different 

 observers. 



Because in practice fluctuations and variations be- 

 have as very large accidental errors, only a statistical 

 analysis of many echoes may be considered reliable. 

 Hundreds of individual echoes must be carefully aver- 

 aged, corrected, and analyzed to give target strength 

 results of any significance. At San Diego, a camera 

 has been installed aboard the Jasper to record, at the 

 same time as each signal is transmitted, the roll and 

 pitch of the vessel, the true bearing of the ship, the 

 relative bearing of the transducer, and the time and 

 pulse number. Such a record should be useful in 

 analyzing and evaluating each echo, but so far has 

 not been applied to a large number of measurements. 

 So far target strength runs have been analyzed from 

 a reasonably large number of individual observations; 

 first, successive groups of five echoes each have been 

 averaged, then an overall average computed con- 

 sidering changes in transmission loss with range and 

 changes in target aspect. Cumulative distributions 

 and computations of probable errors and quartile 

 deviations have been useful in interpreting the re- 

 sults and assessing their reliability. 



