He 584 Å Dayglow at Saturn: a Reassessment
C. D. Parkinson, E. Griffioen, J. C. McConnell (York University), G. R. Gladstone (Southwest Research Institute), B. R. Sandel (Lunar and Planetary Laboratory)
New calculations of the Saturnian He 584 Å airglow intensity, using
radiative transfer models with partial frequency redistribution and
inhomogeneous atmospheric models, are presented. For our standard
conditions the results indicate that the eddy diffusion coefficient at
the homopause, K , lies in the range 4.3 to 1.2 cm s for Voyager 1 and 2 to 5 cm s for Voyager 2 brightness measurements. Our
standard conditions assume a helium mixing ratio, f , of 0.033,
a He 584 Å solar flux of 4 photons cm s
at 1 AU with a solar line 1/e half-width of 73 mÅ
and an atmosphere consistent with the Voyager UVS occultation results.
The results of the earlier work of Sandel et
al. (1982) and Atreya (1982) are in reasonable accord with this work.
Sandel et al. (1982) used an isothermal
atmosphere and a simpler
scattering model with a limited analysis of stellar occultation data.
We have also considered variations in the model atmosphere about the
standard one obtained from the occultation data and find that the
result is robust. This result underscores the original difference
found for values of K obtained from the He 584 Å airglow and
the occultation data (Smith et al., 1983), which yielded a much
smaller value of 5 cm s . However, we note
that there is a possibility that the He 584 Å line center
solar flux could be lower by a factor of two than our standard value.
In this case, K would have to be cm s
or greater. Hence, it would be very difficult to fit the Voyager
He 584 Å\
airglow data with reasonable values of K and the standard value of
f . Increasing f to 0.06, the upper limit of the IRIS
estimate, helps to reduce the descrepancy between measurement and model
but is far from ameliorating the problem. In any
case, the inference of high values of K remains, and also suggests
that future work modelling the Saturnian thermosphere should consider
the higher K values as more appropriate.