Astron. Astrophys. 321, L13-L16 (1997)

3. Interpretation

3.1. The 4.5-5.5 µm region

As for Jupiter, the 5-µm window of Saturn's spectrum probes the troposphere, in the 2-5 bar pressure range. The 5-µm Saturn window has been studied from the KAO by Fink and Larson (1978) and Larson et al. (1980), at a slightly higher resolution (R = 2500), but lower sensitivity than the ISO data. The 4.5-5.0 µm region was also studied from the ground at high spectral resolution (above 20000) by Noll et al. (1986, 1988, 1989), Noll and Larson (1990) and Bézard et al. (1989), leading to the detection of CO, GeH₄ and AsH₃ in Saturn's troposphere.

As pointed out by Bézard et al. (1989), a major difference between the Jupiter and Saturn spectra in the 5-µm window is the presence, in the case of Saturn, of a significant contribution coming from reflected sunlight. This is a consequence of the lower temperature of the tropospheric region of Saturn probed at 5 µm, which gives a much lower thermal emission at this wavelength as compared to Jupiter. Another difference between the Jupiter and Saturn spectra lies in the lower abundance, in Saturn's upper troposphere, of NH₃ and H₂ O, and the much larger abundance of PH₃.

Figure 1 shows a comparison between the Saturn spectrum and a synthetic calculation. A cloud layer at 0.55 bar, with a transmission of 0.9, reflects the sunlight. Another deeper cloud, around 1.55 bar, with a transmission of 0.2, acts as a grey absorbing layer for the thermal emission coming from lower tropospheric levels. Calculations show that the synthetic spectrum is not strongly sensitive to the altitudes of these clouds. In the calculations of the thermal component, contributions due to AsH₃, GeH₄, NH₃, PH₃, CO, CH₃ D and H₂ O were included with the following mixing ratios: AsH₃ /H₂ = 2.0 10^-9, GeH₄ /H₂ = 2.0 10^-9, NH₃ /H₂ = 1.0 10^-4, PH₃ /H₂ = 4.4 10^-6, CO/H₂ =1.8 10^-9, CH₃ D/H₂ = 3.2 10^-7 and H₂ O/H₂ = 2.0 10^-7. CO is only very marginally detected. The GeH₄ and CH₃ D abundances bear a large uncertainty. The abundances of PH₃, GeH₄ and AsH₃ in the upper troposphere, affecting the solar reflected component, have to be lowered to fit the spectrum, as first discussed by Noll et al. (1989), Noll and Larson (1990) and Bézard et al. (1989), and a vertical profile decreasing with height was introduced for these molecules. The ISO-SWS spectrum shows the first evidence for tropospheric H₂ O on Saturn. An acceptable overall fit of the H₂ O lines is obtained with a constant mixing ratio of 2 10^-7 below the 3-bar level, and a cutoff at higher altitudes. Saturated profiles give too large absorption features for H₂ O, which implies tropospheric water undersaturation; a similar result was recently found in the case of a Jovian hot spot by the Galileo probe (Niemann et al., 1996), confirming previous ground-based measurements (Bjoraker et al, 1986). The NH₃ vertical abundance is also strongly depleted above the 1.2-bar level in the model, to account for saturation. The synthetic and observed spectra agree within 2.5% between 4.5 and 5 µm. Our determinations of the AsH₃ and PH₃ mixing ratios are in good agreement with the results of Bézard et al. (1989) and Noll and Larson (1990), while our value of the GeH₄ mixing ratio is 1.6 to 2 times higher than their values; our NH₃ value is consistent with the upper limit (3 10^-4) set by Noll and Larson (1990). The ISO spectrum confirms the oversolar abundance of phosphorus in the deep atmosphere of Saturn. In contrast, the abundance of NH₃ at 5 µm corresponds to a N/H ratio closer to the solar value. It should be mentioned that all these measurements are still preliminary.

Fig. 1. Observed ISO-SWS spectrum (upper curve) and synthetic spectrum of Saturn (lower curve) in the 5-µm region. Spectral absorptions are due to NH3, PH3, AsH3, GeH4, CH3 D and H2 O. In the lower curve, the narrow line corresponds to a calculation without H2 O.

3.2. The 7-15 µm region

The atmospheric region probed by the 7-15 µm spectrum is the upper troposphere and the lower stratosphere of Saturn. The corresponding pressure levels range from about 0.5 bar at 9 µm to about 0.4 mb in the Q-branch of the CH₄ band at 7.7 µm. Figures 2 and 3 show a comparison between the ISO data and a nominal synthetic spectrum in the 7 -11 µm and 11 - 15 µm range respectively. Between 7 and 11 µm, the observed spectrum exhibits stratospheric emission features due to CH₄ ( band), and tropospheric absorption features from CH₃ D (), PH₃ ( and ) and NH₃ (). The 11 - 15 µm range shows emissions by C₂ H₆ () and C₂ H₂ ().

Fig. 2. Comparison between the Saturn ISO-SWS data (above) and a synthetic model (below) in the 7-11 µm range. The spectrum shows stratospheric emission features due to CH4, and tropospheric absorption features due to CH3 D, PH3 and NH3.

Fig. 3. Observed ISO-SWS spectrum of Saturn (upper curve) and synthetic spectrum (lower curve) between 11 and 15 µm. Molecular emissions are due to C2 H6 and C2 H2.

We present here a first attempt to model the ISO spectrum. Our fit is quite satisfactory, except in the 7.0-7.2-µm region where the observed spectrum shows more flux than the model. We first determined a temperature profile which allowed us to reproduce the H₂ -He continuum between the molecular features, and the 7.7-µm methane band assuming a mixing ratio of 4.4 10^-3 in the lower stratosphere and troposphere (Courtin et al., 1984). In the stratosphere, this profile is colder (by 7 K at 0.5 mbar) than the Voyager radio-occultation profile recorded at N in 1981, whereas the tropospheric part is warmer (by 5 K at 400 mbar and 3 K at 150 mbar). Spatial or temporal variations might be responsible for the differences observed between the Voyager profile, measured in a single point, and the present disk-averaged profile, retrieved 15 years later.

We used a CH₃ D mixing ratio of 3.2 10^-7, as derived from previous 5-µm investigations (Bézard et al., 1989; Noll and Larson, 1990) and in agreement with our 5-µm model. A PH₃ profile with a mole fraction of 4.5 10^-6 below 600 mbar (as derived from the 5-µm spectrum), decreasing to 2.5 10^-6 at 300 mbar, and sharply depleted above, allowed us to reproduce the various absorption features observed between 8.8 and 11.1 µm. No emission was observed in the core of the strong Q-branches from the (10.1 µm) and (8.95 µm) bands, indicating that the stratospheric abundance of PH₃ is negligible in contradiction with the analysis of Voyager spectra by Courtin et al. (1984). The cutoff of the PH₃ vertical distribution in Saturn's atmosphere around the 300-mbar level is quite consistent with the analysis of PH₃ rotational multiplets detected in the ISO-LWS spectrum of Saturn (Davis et al., 1996). The C₂ H₆ mixing ratio in our model is held constant at 4 10^-6 above the 10-mbar level, in agreement with Courtin et al. (1984). We infer a C₂ H₂ mixing ratio of 3.5 10^-6 at 0.1 mbar decreasing to 2.5 10^-7 at 1 mbar. This abundance is slightly higher than the Voyager determination (Courtin et al., 1984). The decrease with depth, expected from photochemical models, is needed to reproduce the relative strengths of the fundamental and the associated hot band . A few weak absorption features from NH₃ are detected and can be reproduced with a mixing ratio profile having a 50% humidity.

3.3. Detection of CO₂, CH₃ C₂ H, and C₄ H₂

Figure 4 shows the Saturn spectrum observed between 14 and 16 µm. As mentioned above (Section 2), a procedure was applied to these data to remove some of the instrumental fringes that affect observations in Band 3A. This treatment also modified the intensities of the lines from the C₂ H₂ P-branch, which are then fitted with a mixing ratio profile 30% lower than before. This illustrates that the present uncertainty on the C₂ H₂ mixing ratio derived from ISO cannot be better than 30%.

Fig. 4. Observed ISO-SWS spectrum of Saturn (upper curve) and synthetic spectrum (lower curve) between 14 and 16 µm. The 14.98 µm emission feature is attributed to CO2. Emission from C4 H2 (15.92 µm) and CH3 -C2 H (around 15.8 µm) are also detected. Below 15 µm, lines from the P-branch of the band of C2 H2 dominate the spectrum.

Besides the components of the C₂ H₂ P-branch, the spectrum clearly shows an emission peak at 14.98 µm, a position where no C₂ H₂ line is expected. We attribute this emission to the strong Q-branch of the CO₂ band. A possible fit of this feature is obtained with a CO₂ mixing ratio of about 3 10^-10 above the 10-mbar pressure level, assuming that this ratio is constant with height. The corresponding column density is 9 10¹⁴ molecule cm^-2. Smaller amounts are still consistent with the data if CO₂ is located higher in the stratosphere. We note that the CO₂ emission feature is not sensitive to the CO₂ abundance below the 10-mbar level. Carbon dioxide may indirectly originate from the infall of oxygen bearing material. The recent detection of H₂ O emission lines in the ISO-SWS spectra of Saturn, Uranus, and Neptune (Feuchtgruber et al., 1996) gives evidence for such an exogenic source of material, which may originate from Saturn's rings, moons, or interplanetary particles (Connerney and Waite, 1984). Reaction of CO with the OH radical, formed through photolysis of H₂ O molecules, appears as a possible mechanism for the production of CO₂. Detailed modelling of the Saturn CO₂ emission feature and investigation of its production through photochemical processes will be presented in a forthcoming publication.

The spectrum also shows an emission peak from the band of diacetylene (C₄ H₂) at 15.92 µm, and a broad emission feature centered at 15.8 µm due to the band of propyne (or methylacetylene, CH₃ C₂ H). These two species are produced by methane photodissociation. The observed features can be reproduced with constant mixing ratios of 9 10^-11 (C₄ H₂) and 6 10^-10 (CH₃ C₂ H) above the 10-mbar level. The corresponding column densities are about 3 10¹⁴ and 2 10¹⁵ molecule cm^-2 respectively. No constraints on the actual vertical distribution of these species in the stratosphere are available from the data. These two gases are produced by photochemistry of methane occuring in the upper atmosphere of Saturn. Comparison of the ISO observations with predictions from photochemical models is deferred to a subsequent publication.

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998
helpdesk.link@springer.de