- 1302 — 
Discussion 
The results achieved u:ing this global cozrzection are 
encouraging since the error is minimized from tho 0.5 9C offer- 
ed by the Barnes PRT-5 to 0.39C for an altitude of 400 meters. 
This error may be reduced taking eccount of the non blackbody 
characteristic of sez water (i. e. teking a non unity valued 
emissivity coefficient; and introducing the value of the correct 
ing term for reflection. This may result in an agreonent of 
C.1 9C between ground truth measurements and remotely sensed 
temperatures. In doing so, the limit of sensitivity of the 
Barnes PRT-5 uould be reache2. Tris is a strong limit and the 
question is raised uether it is useful or not to perform such 
a (sometime ) tedious calculation to obtzin this accuracy in- 
stéad of the 0,5 9C offered by the radiomster after a more aasy 
calculation. The trouble of measuring the interface accounting 
fector is for some purposes not worth of the achieved accuracy. 
For other purposes like microstructural studies this exactitude 
is felt to be worthuile. 
An other factor to be considered is the variability of the 
radiometer's calibration. Some scientist have found that the 
blackbody conversion from the radiometric data is not maintained 
with time(23). Hence that the applicability of this proce:iure 
depends mainly on the aim of {hes experiment for otherwise the 
required calculations involved th it may bes not worth of the 
exactitude. It is felt that monitoring purposes the best 
accuracy atteinable From the radio 
S h € 
D 
n 
er is of D.5 SL end th 
only in special ca where t mination of tempe 
ases i rat 
of primary importance the glot ection should bz per 
  
  
  
  
The authors wish to thank Prof. ^A. Ballester for inspiring 
this work and for his many fruitful discussions; Prof. J. Casas 
for his assistance; R. Blanco for his interesting discussions 
and comments; The aircraft was set at our disposition by the 
Ministery of Air under the coordinztion of INTA-COJIE to uhom 
we are in debt.