Page 531 echo sounding 5613 



Mean salinity curves are also required. As compared to temperature, however, 

 the variation normally encountered in salinity has comparatively slight effect on 

 velocity and, except in extreme cases, one mean salinity curve can be used throughout 

 the entire season, even when a number of mean temperature curves are required. 



After the areas and time periods have been decided on, mean representative curves 

 are drawn. This may be done graphically or mathematically. In the former method, 

 the curves to be averaged are plotted on Form 5-1528-5 and by inspection the mean 

 curve is drawn. This method will be found preferable in most cases, and it has the 

 advantage of disclosing immediately any individual curve that varies excessively from 

 the average, and which should be rejected or at least not be included in this group. 

 Various colors should be used to differentiate between the different curves, black being 

 used for the mean curves. 



In the mathematical method, selected representative depths are chosen and 

 the temperatures and salinities at these depths scaled from the separate serial tempera- 

 ture curves (see 5611 (1)) . The arithmetic mean at each depth is then found and plotted 

 on Form 5-1528-5, and a smooth curve is drawn through the points; or the arithmetic 

 means, if the depths have been chosen with this in mind, may be entered directly in 

 columns B and C in the form on which the velocity correction computations are made 

 (see table 20). This mathematical averaging of the temperatures and sahnities elim- 

 inates the necessity for replotting all the serial temperature curves and the difficulty 

 that may be experienced when a number of curves plot very close to one another. 



For depths greater than 120 fathoms, each temperature curve shall be plotted in 

 two parts, the shoaler part at the scale provided on Form 5-1528-5, and the deeper 

 part at a scale one-tenth of that used for the shoal section. These should be drawn 

 and the separate parts identified in accordance with instructions in 6342. 



5613. Numerical Determination of Velocity Corrections 



Velocity corrections may be derived numerically or graphically, steps (1) and (2) 

 below being the same in either case. For the purpose of illustration an example of 

 velocity correction computations by the numerical method is given in table 20, with 

 columns identified by the letters {A) to (/) for reference purposes only. This is the 

 standard form which shall be used for all corrections derived by the numerical method. 

 The graphic method of deriving velocity corrections is described in 5615. 



It must not be assumed that all depth layers, for the purpose of determining 

 velocity corrections, need to be chosen as shown in this example, but from past expe- 

 rience 5-fathom layers in the upper hundred fathoms, 20-fathom layers in the second 

 hundred fathoms, and 200-fathom layers in greater depths have usually been found 

 satisfactory. Where the change in temperature is regular with respect to depth and 

 not too great, thicker layers may give sufficient accuracy. On the contrary, for precise 

 results in shoal water 5-foot intervals may even be required at times. 



