Astron. Astrophys. 356, 347-356 (2000)

6. Discussion

In the following, we compare our results with previous infrared observations of Saturn (Gezari et al. 1989, Tokunaga et al. 1978). These observations have been obtained at several wavelengths, different or not from ours, and give additional constraints on the analysis of thermal structure and hydrocarbon abundances in the atmosphere of Saturn. The following discussion only focus on the NPE. Gezari et al. (1989) have made 3 images of Saturn at 7.8, 11.6 and 12.4 m. On each of their images they saw the NPE. Concerning the images at 11.6 and 12.4 m, their observations are consistent with our two assumptions since the wavelengths used are almost the same as ours, i.e. sensitive to the ethane emission. As regards the 7.8 m image, it is sensitive to the methane emission band. So the NPE seen on this image implies either an enhancement in the methane abundance or a higher stratospheric temperature (the contribution function at 7.8 m peaks in the lower stratosphere). Due to the photochemical origin of ethane and acetylene, their abundances depend directly on that of methane, and a greater methane abundance would produce a greater abundance of ethane and acetylene, which would explain the enhanced flux at 11.6 and 12.4 m. A problem arises with the increased methane abundance. Indeed, the methane mole fraction is constant in the deep atmosphere of the giant planets and it would be difficult to explain an enhanced methane mole fraction. It is thus likely that the observed structure at 7.8 m is caused by an enhanced temperature and not by a larger methane abundance.

Tokunaga et al. (1978) studied Saturn's infrared emission using north-south scans at 17.8, 19.7 and 22.7 m obtained in 1975 and 1977. Their observations were made when Saturn presented a geometry opposite to the one we have: the tilt of the south pole of Saturn towards the sun was about 15^o, whereas during our observations, Saturn presents the same tilt angle, with the north pole being towards the sun. Furthermore we observed Saturn in 1992, 15 and 17 years after the observations of Tokunaga et al., which corresponds to half a period of Saturn's revolution around the Sun. Therefore we have symetrical observation conditions which would imply that we should observe the same features as Tokunaga et al. (1978), but now for the north pole. Bézard & Gautier (1985), who computed the seasonal cycle of Saturn's stratosphere, showed that, at 5 mbar, the temperature of the north pole in 1992 should be nearly the same as that of the south pole in 1975-1977. The thermal inversion in Saturn's atmosphere is located around 60 mbar, below the 5 mbar level, but it is likely that the thermal variation calculated by Bézard & Gautier extends to levels around the thermal inversion. Tokunaga et al. (1978) observations revealed the presence of a strong enhanced emission at the south pole at a wavelength range (around 20m) which is not sensitive to hydrocarbon emissions. Moreover, simultaneous equatorial scans showed that the limb brightening at the equator is smaller (Caldwell et al. 1978). This implies, as reported by Tokunaga et al. (1978), that the temperature of the atmospheric levels around the thermal inversion is hotter at the south pole. From the above considerations, it is likely that the levels around the tropopause at the north pole should show a higher temperature at the epoch of our observations. We did not observe Saturn in the 20m window, but the images at 10.91 m also probe the thermal continuum. The interesting point is that at 10.91 m, we do not see any polar brightening. The maximum of the contribution function at 17.8 m is located just around 100 mbar and near 500 mbar at 10.91 m. In order to see if a hotter tropopause could produce a brightening at 17.8 m and could be invisible at 10.91 m, we calculate the emission at these two wavelengths with two different thermal profiles, the only difference being the hotter tropopause (+3K). For emission angle corresponding to high latitudes, we find an enhancement of about 20% at 17.8 m, but a similar 10.91 m flux in the two cases. Therefore, the absence of a polar emission at 10.91 m does not imply the abscence of a polar emission at 17.8 m. With all this information, we can postulate that the seasonal evolution of insolation produces thermal effects, the insolated pole being hotter, and these effects concern only the upper troposphere (above 200-300 mbar) and the stratosphere. This agrees quite well with the results of Conrath and Pirraglia (1983) who derived the latitudinal variation of the temperature at three pressure levels, 730, 290 and 150 mbar. The thermal asymmetry between the northern and southern hemispheres, which is an effect of the seasonal varying insolation, is clear only at 150 mbar but not at 730 mbar, in agreement with our results. Moreover, from calculations with radiative-dynamical models, Conrath et al. (1990) and Barnet et al. (1992) showed that, at north summer solstice, levels below 300 mbars have a constant temperature over all latitudes (Fig. 9b of Conrath et al. (1990) and Fig. 6 of Barnet et al. (1992)), the upper levels being sensitive to the seasonal variation. And the calculated structure of the atmosphere seems to be in good agreement with the observed one. Indeed, we have already noticed that these calculations lead to a good agreement with the observations of Gezari et al. (1989) which were obtained near north summer solstice. Our observations were conducted when the solar longitude of Saturn was about 145^o, during northern summer. Due to the phase lag between the insolation and the thermal response of the atmosphere (Conrath & Pirraglia,1983, Conrath et al.1990), the northern hemisphere is still warming at the time of our observations, and these two radiative-dynamical models are in good agreement with the global structure of the atmosphere derived from our observations. Nevertheless, the precise calculation of the structure of the atmosphere at the precise time of the observations would be necessary to determine exactely the order of magnitude of the latitudinal temperature variations, in order to compare with our results (models should find a variation of the order of 10K between 15^o N and 60^o N). Furthermore, the NEB, which we associate with the temperature enhancement between 10 and 20^o N observed by Voyager/IRIS (Conrath & Pirraglia, 1983) seems to be difficult to reproduce with these models, and further improvements of such models are necessary to provide a better understanding this particular feature and the thermal structure of Saturn's atmosphere.

Finally, we conclude that the hypothesis of a hydrocarbon abundance enhancement is not necessary. Nevertheless, it is likely that such variations occur but probably not with the order of magnitude found here. The enhanced abundances found here are more likely the upper limits of ethane and acetylene abundances. The processes which may be responsible for this enhancement still remain to be identified. For instance, one important parameter for the photochemistry is the temperature profile since chemical reactions are temperature-dependent. Therefore, the hydrocarbon abundances should vary with latitude as the temperature profile varies. Furthermore, processes involving energetic particles could produce significant effects which could lead to a variability of hydrocarbon abundances (as suggested by Kostiuk et al. 1987, in the case of Jupiter).

© European Southern Observatory (ESO) 2000

Online publication: March 28, 2000
helpdesk.link@springer.de