Explanatory note on Neptune's continuum model by Emmanuel Lellouch (13/12/96)
_____________________________________________________________________________


This is a modified version of a continuum model for Neptune based on 
temperatures derived from the Voyager radio-occultation experiment taking
into account that this experiment sounded one of the coldest points on 
the planet. It considers all the infrared data (25-50 microns) from 
Voyager/IRIS plus some ground based measurements.  

In short, the observational constraints to meet are: 

(tb represents a disk-average brightness temperature)

Tb(350 mu) = 61.2+/-2.5 K (ground based, Griffin and Orton 1993 Icarus 105, 537).
Tb (50 mu) = 63 K  (Voyager/IRIS)
Tb(25 mu)  = 58 K   (Voyager/IRIS)
Tb(20 mu)  = 57 K  (ground based, Orton et al 1987 Icarus 70, 1)
Tb(16.7 mu)= 62 K  (ground based, Orton et al 1987 Icarus 70, 1)

In order to reproduce the observations in the 20-25 micron range the 
radio-occultation temperature profile was required to be warmer by about 
4.5 K. The excess flux so generated near 50 microns was reduced by 
introducing a cloud at 0.33 bar with transmission 0.6. Physically, it should
represent a CH4 cloud (although one would expect a CH4 cloud to lie somewhat
deeper in the atmosphere). Then , the model gives a reasonable match of the
data longward of 20 micron, except that the Tb at 350 micron is about 4 K too
warm (in terms of flux, this is fortunately only a 10 % error). The most
serious problem is the behaviour shortward of 20 micron.  At 16.67 micron 
(600 cm-1) the model underestimates the data of Orton et al.  by 4 K, i.e. a 
factor 2.5 in flux. 

The model is not specifically tuned to match the SWS data of Neptune, but 
at 30-45 micron, it gives a good fit, within 10 % or so.


Estimated accuracy of the model:

At 20 to 50 micron, the model matches all the measurements quite well. Thus, 
the accuracy is probably limited by the measurements accuracy, which I 
estimate to be 1.5 K. In flux, this gives : 12 % at 50 mu, 25 % at 25 mu, 
30 % at 20 mu. Longward of 50 micron, the uncertainty in Tb is probably more 
than 1.5 K, but this should remain within 10 % on flux everywhere. 
Shortward of 20 mu, the uncertainty increases dramatically.

In summary:

Wavelength        Estimated accuracy of the model
  350 mu              10 %
  50 mu               12 %
  25 mu               25 %
  20 mu               30 %
  16.7 mu            factor 2.5 !