Other problems of consequence have been associated mostly with power sources, 

 chart recorder, and the instrument itself. Errors caused by atmospheric radiation of in- 

 frared are considered to be of little consequence because of precautions used in the survey 

 procedure. Water vapor in the air path undoubtedly contributes some error to the readings, 

 but in flying at low altitude (500 ft.) this is minimized. 



INSTRUMENT PROBLEMS 



Our experience has been primarily with the Barnes IT-1, operated from a basic power 

 source of 21-24 VDC (from two 12-volt storage batteries) and 27-28 VDC (from the aircraft J 

 auxiliary power unit) through an ATR No. 28 URSF 115 VAC inverter. A NESCO JY 110-2 ' 

 single -channel recorder was used. The three units were mounted in a portable rack for con- 

 venience. 



Troubles with the inexpensive ATR inverter arose from variations in frequency output 

 associated with variation in voltage input; e.g., 21 VDC yielded 59.9 c.p.s. and 28 VDC 61.1 

 c.p. A difference of about 1 c.p.s. in the power input to the IT-1 produced a difference of 

 about 3°F. in the temjDerature readout. This effect can be corrected by instrument calibration 

 (gain adjustment) if the inverter input voltage is known or through use of a knowTi temperature 

 reference target. A better solution is to use a frequency controlled power source. 



The most severe problems with the instrument itself appeared to be related to the 

 battery-supplied bias voltage circuit. In the IT-1 the bias voltage is supplied by 4 standard 

 radio B batteries. With use, the decreasing voltage output of the batteries, initially 178 V, 

 has the effect of causing a decrease in readout values. This has been corrected by gain ad- 

 justment. However, this calibration procedure is rather awkward, particularly since the 

 batteries appear to recharge between periods of use and then decline rapidly on reuse. 



In our work this problem has given rise to a possible error of about 2^F. However, 

 the record can be corrected for this effect if frequent in-flight calibration checks are made. 

 With the IT-1 we now make these checks routinely; every 12-15 minutes when possible. The 

 reference target is a container of water maintained within the range of sea surface tempera- 

 ture prevailing during that time (ice chips are carried aboard). The necessity for these pre- 

 cautions should be obviated by changes made in subsequent models which have electronic- 

 bias voltage supply. 



WEATHER PROBLEMS 



In the IT-1 thermometer, radiation from the atmosphere per se is nearly eliminated 

 by the optical system which received radiation only in the 8 to 13 micron range. Thus dry 

 air should offer no interference to measurement of sea surface temperature with the IT-1. 

 However, radiation or attenuation from water vapor in the atmosphere can interfere. But, 

 we have not detected significant effects from water vapor in our surveys even at high relative 

 humidities. Condensed vapor has been a problem, however, in the form of medium to heavy 

 fog, rain, or snow and we have avoided flyir^ in these conditions or have rejected such data 



-136- 



