61 



2.000 



lnttrpr%t»fion of Ttmp^raturt Dtpth Points. 



. ObaerveJ -ie/T^ra^are clep^h 



/^ Smooth line joining pomis 



^ PossMe area of f/ucitiaiion 

 due to errors such as iniemal turves 



Actual chanqe in temperaturf / (/epih 



over wh!c/i ihe meAsareinent t/ield«J no dgh. 



clarify the problem somewhat. 



a. The primary measuring ele- 

 ment will have its own internal 

 hysteresis and thermal time 

 constant. 



b. The recorder will have its own 

 hysteresis and time constant. 



c. The time or depth scale will 

 have its own hysteresis and 

 time constant. These time 

 constants make a low pass filt- 

 er of the recorder. 



d. The phenomena we are measur- 

 ing may well be changing both 

 spatially and timewise, for ex- 

 ample, in the presence of ed- 

 dies or internal waves. 



e. In all probability our measur- 

 ing instrument is b«ing dragged 

 or is oscillating vertically and 

 horizontally through the struc- 

 ture we are trying to measure. 



f. If we are trying to measure 

 small structures we are con- 

 fronted with the finite size of 

 the measuring probe. 



Fig. 4. Typical set of temperature- 

 depth points showing range of possible 

 interpretations. 



The dynamic size of a thermal 

 element can be many times the actual 

 size and it must be known and under- 

 stood for each piece of equipment. In- 

 stead of the actual size of the probe one 

 should visualize a fictitious tempera- 

 ture element whose size is as large as 

 the actual size plus the product of its 

 thermal time constant times the speed the element is moving through the water. 

 The time constant is the time required for the thermal element to come to with- 

 in a given per cent of the final temperature. This is usually stated as the time 

 to come to within 37% of the final temperature. In a thermal system the re- 

 sponse is usually logarithmic and will have the following response: 



Time 



e 

 ze 



39 

 46 

 56 



Approach to Final Temperature 



37% 

 13% 



5% 



2% 



0.7% 



A bathythermograph will have a time constant of about 2/l0 seconds so 

 if it is falling 10 feet per second its thermal element will be effectively 4 feet 

 long if it is to respond within 13% of the temperature change and effectively 10 

 feet long if it is to respond to within 7/10% of the temperature change. 



It is very difficult to make a rugged thermal element with a rapid speed 

 of response although the state of the art is improving and some commercial 

 manufacturers are making higher speed elements. Even an infra red device 



