of large volumes of data. Some other type of read-out, such as punched 

 cards or punched tape, would b'e preferable. The temperature error 

 should not exceed ±0.1°C for the temperature range of -2° to 32°C. 

 The depth error should be less than two percent not including the 

 error resulting from linear interpolations between thermistors. The 

 optimum spacing of thermistors is not known but is believed to be 

 about 25 feet. 



The major expense and difficulty of handling of the Richardson 

 chain lie in the winch and mechanical chain linkage. This expense 

 largely would be eliminated if the instrument were not designed for 

 underway use. Although a requirement exists for such an instrument 

 to be used underway, a greater requirement exists for an instrument 

 which need not be engineered for underway use. Time-series studies 

 of internal waves from an anchored ship and Texas Tower applications 

 are two uses of such an instrument. 



A contract for a shipboard thermocline recorder has been let. 

 If evaluation of this equipment is favorable, it is recommended that 

 contracts be let for other types of these recorders. 



c. Airborne radiation thermometer 



This important survey tool has wide application in ice pre- 

 diction as well as in temperature prediction. Although much experi- 

 mentation still is required to realize the full potentialities of an 

 airborne radiation thermometer, it already has proven its value as 

 the most practical way of obtaining nearly synoptic, gross, sea surface 

 temperature data. The mean error of this instrument as tested by the 

 Hydrographic Office is approximately ±0.2°C, depending upon the 

 competence of the operator. Since this inherent error is primarily 

 the result of atmospheric transmission and in some cases sun glitter, 

 the relative accuracy is not much better than the absolute accuracy. 

 Better accuracy is believed possible with the use of associated instru- 

 ments for sampling the atmospheric gradients between the sensor and 

 the sea. 



It is recommended that refinement of the airborne radiation thermo- 

 meter be continued. 



F. SELECTED REFERENCES 



III-l. BEHAR, M. F., ed. Handbook of measurement and control. 

 Instrumentation and Automation, vol. 27, no. 12, Pt. 2, Decem- 

 ber 1954. 216 p. 1954. 



Ill - 1 



