were 1° to 3° warmer than the infrared radiometer. These lightships are in areas of great 

 turbulence and near boundaries of maximum temperature gradients, therefore the signific- 

 ance of this disagreement is not known. 



Surface water temperature checks were made on December 20, 1963 and April 17, 

 1964 with the aircraft circling the laboratory boat for simultaneous temperature data. The 

 results showed almost identical readii^s as follows: 



Date 



December 20, 1963 

 April 17, 1964 



It should be noted that the air temperature was not essentially different from the water tem- 

 peratures. The lightships, on the other hand, were in deeper water near the inshore con- 

 fluence of the Gulf Stream where the air-sea temperature differences were greater. During 

 all flights the air temperatures were lower than the water temperatures recorded in the 

 vicinity of the lightships. 



In-flight temperature checks with a bucket of water and a mercury thermometer were 

 made during all flights. The water temperature and air temperature, in the aircraft, were 

 approximately the same. These checks were made to ascertain the normal operation of the 

 instrumoit and recorder rather than for calibration. 



We have not conducted field studies of the effects of wind, spray, fog, or reflected sun| 

 light on the recorder temperature. In-flight observations over an area presumed to be rela- 

 tively constant showed no obvious temperature changes with reflected sunlight, increased ^ 

 white caps and spray from increased wind velocity, or changes in bolometer angle from 30 

 to 60° normal. Also, changes in flight altitude from 300 to 1.500 feet caused no apparent 

 change. All of these conclusions are based on observations without a known reference. 



The limits of accuracy of our instrument as we operate it have been within .5 C . of 

 the temperature of water dipped from the sea surface with a bucket and read within 30 seconds 

 with a mercury thermometer or thermistor. In our opinion this accuracy is more than ade- 

 quate for the turbulent shallow waters of the area we are surveying. For example, in April 

 1964 the sea temperature chained 10°C. in less than 5 miles at the inner boundary of the 

 Gulf Stream. The continual interaction of coastal turbulence from Capes Hatteras and Fear 

 and the Gulf Stream keep the shelf waters of our section of the coast in such a state that any 

 attempt at greater accuracy than ±.5°C. probably is meaningless. Changes with time are a 

 far greater concern. For all we know. Gulf Stream waters may invade within the 10 fathom 

 contour and recede in a few days. Conversely, cooler or warmer water from meteorological 

 conditions may dominate much of the shallow shelf of the region for relatively short periods. 



-58- 



