HEWITT ET AL.: DEVELOPMENT AND USE OF SONAR MAPPING 



The correlation coefficient (r) = -0.969. Table 

 1 summarizes horizontal school area contri- 

 butions by size class for observed frequencies 

 corrected for edge bias and for frequencies de- 

 rived from the exponential model. In both cases 

 more than 90% of the area was contributed by 

 schools larger than 20 m. The importance of hori- 

 zontal school area is that it is probably propor- 

 tional to the tonnage offish in schools and, in this 

 sense, decreases the significance of any bias in 

 the counts of small schools. 



Table l. — Cumulative percent of total horizontal school 

 area contributed by size class for observed frequencies (cor- 

 rected for edge bias) and for frequencies derived from an 

 exponential model. 



Diurnal and Seasonal Effects 



Time specific frequency distributions were 

 drawn for data collected on cruises in April-May 

 and in November 1973 for the purpose of discern- 

 ing variations in sizes and detection of schools 

 during various times of the day. While variations 

 were noticed, their pattern was neither pro- 

 nounced nor consistent from cruise to cruise. This 

 is not to say that daily changes in schooling be- 

 havior do not exist, but that our data base is in- 

 sufficient, at present, to delineate them. In the 

 evening, discrete, well-formed schools of anchovy 

 have been observed to disperse into a thin scat- 

 tered layer but no program of study on this prob- 

 lem has been undertaken. 



The data base is insufficient to detail seasonal 

 changes in school size distributions, although, 

 from communication with Mais and several 

 commercial fishermen, we have reason to expect 

 somewhat larger schools in the fall and smaller, 



scattered schools in the spring. Mid-spring is con- 

 sidered to be the main spawning season of the 

 northern anchovy. 



Target Strength 



Acoustic target strength is proportional to the 

 ability of an object or group of objects to reflect 

 sound waves. Acoustic reflections from schools of 

 fish are not presently well enough understood for 

 rigorous characterization of the biomass of a fish 

 school by the use of sonar. Nevertheless, we have 

 measured apparent fish school target strengths 

 with the objective of providing data which may 

 lead to the quantification of fish schools in terms 

 of total biomass. 



Peak echo amplitudes were collected and cor- 

 rected for propagation and absorption losses by 

 employing the active sonar equation: 



EL = SL - 2TL + TS 



where EL 

 SL 



echo level in decibels (dB) 

 source level in decibels, reference 

 1 yubar at 1 m 

 TL = transmission loss in decibels 

 TS = target strength in decibels. 



Solving for target strength and using signal vol- 

 tage level as a measure of echo level: 



TS = 20\ogV - k + 40 \ogR + 2 ozR 



where V = peak echo signal amplitude in 



volts 

 k = calibration coefficient which is 

 the algebraic sum of source 

 level, receiver sensitivity, and 

 system gain expressed in 

 decibels 

 40 log R + 



2a:i? = range dependent transmission 

 loss (assuming spherical losses 

 as in a homogeneous fluid) 

 where R = midrange of target 

 (as an approximation of the (lo- 

 cation of peak echo amplitude), 

 and cc = absorption coefficient 

 expressed in decibels per meter. 



Figure 4 illustrates five samples of peak target 

 strengths computed from data taken by the 



285 



