If the instrument which senses F is calibrated with reference to a blackbody cavity, it will 

 assign the value T to the surface. Equation (4) shows, however, that T will be equal to 

 T only in special cases. For e.xample: 



a. if T = T then T = T 



CO o 



b. if r = o then T = T 



o o 



c. if the sky is clear then T< T 



d. if r = r 



o c 



1, for T < T , T< T 



CO o 



2. for T >T , T>T 



CO o 



Consequently, appreciable error may result especially under clear sky conditions, e.g., 1°C. 

 Use of (4) is possible if cloud base tempei'ature and reflectivities are known. Oridinarily, 

 surface and cloud reflectivity over the 8-15 micron wave band may be taken as 0.014 for 

 normal incidence. On the other hand, if calibration of the IRT is done in the field in situ , as 

 mentioned earilier, these difficulties are eliminated. The procedure would be time consuming 

 but fruitful. As a point of interest, if two IRT's were employed simultaneously, one lab-cali- 

 brated while the other was field calibrated, it may be possible to estimate the reflectivity of 

 the water surface over the wave band considered. This should also be possible with one lab- 

 calibrated IRT in a controlled experiment where water temperature is known and a plate of 

 known optical properties and temperature is used to simulate the cloud base. 



OTHER COMMENTS 



It is hoped that the above discussion may give some guidance in considerations of the 

 nature of what an IRT measures. If reliable IRT readings are made, it is likely they may be 

 correlated with sub-surface temperatures under varying but known ambient conditions. Con- 

 sequently, if one water temperature is known the other may be estimated. From the stand- 

 point of air-sea interactions as well as of operational needs, this would appear desirable. 

 Empirical equations of this type have been obtained for this purpose (R. D. Boudreau, M.S. 

 Thesis, Dept. Oceanography and Meteorology, Texas A&M University). 



-98- 



