Figure 1-A is the simplest type; levels a and b are easily located; 

 and the sharp turn (almost 90 ) of the trace at the interface indicates 

 active mixing and probable horizontal flow in different directions in the 

 mixed layer and the thermocline. The various levels are less certain in 

 Figures 1-C and 1-J, the traces in Figures 1-G and 1-1 curve mildly, and 

 Figures 1-C and 1-1 indicate a decaying mixed layer. Level c in Figures 

 1-B and 1-L show weak thermoclines. A weak thermocline is a small thermo- 

 cline within the mixed layer. 



The depth of the thermocline bottom is not always easily determined; 

 there may be considerable difference in evaluation. Sometimes a decisive 

 change in temperature gradient occurs as seen in Figures 1-A and 1-C. 

 Normally, the gradient is very weak and decreases slowly as seen in Fig- 

 ures 1-B, 1-D, 1-H, 1-K, and 1-L. In these particular cases, the thermo- 

 cline bottom is established as a level where the temperature gradient is 

 less than 1 F per 50 feet. The exact depth is irrelevant, because varia- 

 tion of temperature below the thermocline is slight. Evaluation of At 

 (temperature difference between the sea surface and the bottom of the 

 thermocline) will vary little, even if evaluation of the depth of the 

 thermocline bottom varies considerably. The temperature at a given depth 

 below which the thermocline bottom migrates only in winter would be the 

 best reference level for determination of At. For existing BT data, the 

 ij-00-foot depth is most convenient, because traces often end shortly after 

 reaching this depth. 



When there is a positive gradient in the thermocline (Figure 1-F), 

 stability of the thermocline is uncertain; the water mass comprising the 

 mixed layer will often become unstable and sink; and a deep mixed layer 

 will be produced by the instability mixing process. Sometimes water 

 masses with a positive gradient are stable, because considerably higher 

 salinity water in the thermocline balances the cold, less saline water 

 in the mixed layer. 



Figures 1-1 and 1-J show surface effects of warming and cooling, 

 respectively. Figure 1-3 indicates the origination of or the vanishing 

 of a mixed layer. 



Reliability of the Bathythermograph 



Bathythermographs are not precise instruments, and they often lose 

 calibration. Comparison of a bathythermogram and plotted temperatures 

 of a Nansen cast taken simultaneously yields satisfactory agreement most 

 of the time. The greatest discrepancies of temperature and depth usually 

 occur in the thermocline. 



BT temperature corrections based on the average difference between 

 bucket or injection temperatures and the surface temperature of the BT 

 are made during processing. The mean error probably does not exceed 

 1 F. Actual temperature values are irrelevant in this prediction method; 

 temperature differences between various depths in a BT are of more inter- 

 est. Comparisons with simultaneous Nansen casts provide evidence that 

 temperature differences are more reliable than actual temperatures. 



