Obs. Paris and IUF/Paris 6, DESPA, 92195 Meudon Cédex Principal, France
Astronomy Department, Cornell University, Ithaca NY 14853-6801 USA
170-25, California Institute of Technology, Pasadena CA 91125 USA
MS 245-3, NASA Ames Research Center, Moffett Field, CA 94035-1000 USA
Astronomy Department, Wellesley College, Wellesley MA 02181-8286 USA
Lunar and Planetary Laboratory, University of Arizona, Tucson AZ 85721 USA
Department of Earth and Space Sciences, State University of New York, Stony Brook NY 11794-2100 USA
Astronomy Department, Cornell University, Ithaca NY 14853-6801 USA
Department of Astronomy, University of Michigan, Ann Arbor MI 48109-1090 USA
MS 245-3, NASA Ames Research Center, Moffett Field, CA 94035-1000 USA
Keywords: Solar system, Saturn, rings
Every half orbital period of Saturn, i.e., about every 15 years, the Sun crosses Saturn's ring plane. Around that equinox period, and depending on the special geometry of the event, the Earth crosses this plane either once or three times. For several months, the rings are thus illuminated and observed at very grazing angles, allowing specific scientific goals to be pursued.
More precisely, the ring brightness diminishes considerably during several months, allowing the retrieval and/or discovery of small moons otherwise lost in the glare of the ring system. Also, the faint, but non-zero flux from the rings observed edge-on, provides an estimate of the ``equivalent thickness'' of the rings. This may include the contributions of a global warp of the ring plane, local bending waves, a tenuous atmosphere of dust, or usually undetectable, but vertically perturbed, narrow rings. Furthermore, the edge-on configuration increases the apparent optical depth of faint rings. This allows one to scan Saturn's equatorial plane outside the bright ring system, searching for material usually too tenuous to be imaged. Finally, this is the occasion to observe many mutual events between satellites (eclipses and occultations), a powerful tool to improve satellite ephemerides.
There were three Earth ring plane crossings (RPX's) in 1995/96, namely on May 22, 1995, August 10, 1995 and February 11, 1996, while the Sun RPX was centered on November 19, 1995. The present paper is devoted to the August and November observations by the Hubble Space Telescope (HST), during Cycle 5. Due to the recent character of these experiments, only preliminary results will be presented here. A more detailed account of our observations and results are published elsewhere (see Nicholson et al. 1996).
Note also that the May observations, made during Cycle 4, are discussed elsewhere (Bosh et al. 1996a,b). Finally, the February 11, 1996, crossing is angularly too close to the Sun to allow any HST observations.
A total of 15 orbits were allocated to our team for the August and November RPX's, using the Wide Field and Planetary Camera 2. Six orbits were centered on the crossing of August 10, 1995, with two more the day before to image the faint E and G rings. Five orbits were dedicated to the Sun RPX around November 21, 1995, with again two more on November 27 and 28 to study the E and G rings. The main rings were observed at 890 nm (methane band) in order to reduced the scattered light from the planet. Broadband filters at 300, 450, 555 and 675 nm were used to observe the diffuse, more remote, G and E rings.
During the time span of our run (from August 10, 1995, 14 h to August 11, 1 h UT), the Earth went from the southern (unlit) side of the rings to the northern (illuminated) side. During this period, the elevation of the Earth above the ring plane went from to +0.005, with an expected ring plane crossing at 20:54 UT (see the discussion by Nicholson et al. 1996). Meanwhile, the rings were illuminated by the Sun with an incidence angle of +1.5
One of the most surprising results of the August observations is that the main rings, when observed edge-on, exhibit an almost featureless profile as a function of distance to the planet center. In addition, instead of dropping at the outer edge of ring A (136,770 km), the signal keeps on increasing slightly, up to the radius of ring F (140,180 km). In particular, no significant brightening is observed at the locations of the C ring and Cassini Division during the 12 hours or so bracketing the RPX. This is in contrast to the ground-based observations taken 12 hours or so before the RPX (at 9:00 UT on 10 August), in which a clear brightening is detected at these two locations.
The flux from the rings between 80,000 and 120,000 km is particularly smooth, and its average value can be used to estimate the equivalent thickness z of the rings in this range of radius. As explained in the Introduction, several effects can increase z above the local physical thickness. In essence, z may be viewed as the vertical extent of a homogeneous ``box-like'' ring, composed of the same particles as the actual rings, and reflecting the same amount of light as measured from the Earth, when observed edge-on. Note that the quantity z is thus model dependent, since assumptions must be made as to the particle albedo and phase function (see Sicardy et al. 1982, Dones et al. 1993).
Figure 1 shows the vertically-integrated value of (translated into km) vs. time on August 10, 1995, derived separately for the eastern and western ansae. This value will be used later to derive the equivalent thickness z of the rings, assuming some optical properties of the particles, as explained above.
Figure: Vertically-integrated brightness of the rings (in km), averaged between 80,000 and 120,000 km, vs. time, around the expected ring plane crossing (indicated by the vertical dotted line). Note the conspicuous discrepancy between the RPX times for the east and west ansae, as well as the clear difference in slopes after the RPX.
Note the clear asymmetry of the ring brightness right after the expected RPX, the western ansa being 30% brighter than the eastern ansa 1 hour or so after the RPX time. Also, the RPX times for the two ansae differ by about 45 minutes, the ring plane crossing on the western side appearing to occur earlier.
The residual equivalent thickness at RPX, about 1 km, is consistent with the value derived, with lower accuracy, by Sicardy et al., 1982 from the March 1980 RPX, z= 1.1 km. This figure was interpreted at that time as being due to bending waves excited by the vertical 5:3 resonance with Mimas, observed by Voyager in the outer A ring (at the radius 131,800 km). The total amplitude of these waves, about 1 km (Shu et al. 1983, Gresh et al. 1986), agreed quite well with the value of z derived above.
The HST spatial resolution (better than 700 km at Saturn), definitely shows, however, that no special feature is associated with the radius of the 5:3 bending waves. In particular, no drop of signal is noted outside that radius, where only the physical thickness of ring A, 50 m or less, should be observed.
Altogether, these results clearly show that something is hiding the main ring from us during the few hours bracketing the Earth RPX. Because the thin line of the rings is then detected up to the F ring radius, the latter must be the main contributor to the edge-on ring brightness. Note, however, that the F ring equivalent thickness is barely large enough to overcome what is expected from the bending wave contribution, since both should have a z of about 1 km.
Since these recent observations are still being analyzed, only brief comments will be given here. The November run was intended to watch the last rays of the Sun grazing the ring plane, with the very last light on the northern side of the rings being expected on November 21, 1995, around 12:00 UT. This configuration should enhance the visibility of any global warp of the ring plane, or localized bending waves. An advantage of the November Sun RPX is that the Earth had a planetocentric elevation of 2.7, about the maximum possible for such an event. This yields a good 2-D resolution of the ring ansae, contrary to the August crossing, from which we essentially got a unidimensional view of the rings.
Although the C ring and the Cassini Division are conspicuous in the November images (because of the solar transmitted light), the F ring is again the prominent feature of the ring system (Nicholson et al. 1996). Nevertheless, faint details are visible at the location of the Encke gap and at the outer edge of the A ring. These local brightenings could be due to the real physical thickness of the rings at those locations, and are discussed elsewhere (Ibid.). These authors also discuss azimuthal asymmetries in the B and A rings, which probably result from the planet's illumination for the former, and from ``quadrant asymmetries'' for the latter, as was observed by Voyager (Franklin et al. 1987).
The appearance of the F ring near the west ansa indicates that this ring has a small, but non-zero () inclination relative to the A ring. More precisely, the F ring consistently disappears out of view on the west side, a little bit beyond the ansa. The most straightforward explanation for that phenomenon is that a shadow is cast by the A ring on the F ring, a configuration requiring a relative inclination between the two rings.
In any case, it appears that some care should be taken in analyzing the August data, since the F ring appears to be the major contributor to the edge-on ring brightness. Also, the probable inclination of the F ring is likely to complicate the interpretation of the crossing times derived in August. In particular, it is not yet clear whether the discrepancy observed between the two ansae at that time is solely due to the F ring, or also to a global warp of the ring plane. Finally, no conclusion is drawn for the moment as to the position of Saturn's pole, derived from the ring plane orientation. Consequently, we have not yet deduced any value for the precession rate of the planet, from the combination of the 1980/81 and 1995 observations.
Both the E and G rings are readily visible in the HST images. While the E ring was first glimpsed in 1966, and re-observed in 1980/81 from the ground and from Voyager, the G ring was only marginally observed by Voyager in 1980/81 (see Showalter & Cuzzi 1993), and the 1995 campaign yielded the first opportunity since then to detect this ring from the Earth.
Several features associated with the E ring have been confirmed by the HST, and extended to the UV part of the spectrum. In particular, the blue color of this ring is now well established, with better error bars than in 1980 (Nicholson et al. 1996). Note also that numerous ground-based observations now provide new photometric data on that ring in the IR (in particular at 2.2 m). The resulting spectrum, from 0.3 to 2.2 m, will be used to constrain the particle size distribution in the E ring. It seems likely, in the meantime, that the unimodal distribution around the size 1 m, as derived by Showalter et al. (1991), will be confirmed.
The HST observations also confirm the maximum of brightness of the E ring at the orbit of Enceladus ( 240,000 km), and shows a hint of a dip at the location of Mimas' orbit ( 180,000 km). Finally, the vertical scale height of the E ring is resolved everywhere, with a local minimum at Enceladus' orbit (full thickness at half maximum 8,000 km), and a flaring at larger radii (more than 15,000 km beyond the orbit of Dione, at 380,000 km).
In contrast with the E ring, the G ring essentially exhibits a grey or slightly red spectrum between 0.3 and 0.9 m (Nicholson et al. 1996). Further analysis is still required to detect any internal structure in this ring. In particular, the November observations should tell us whether or not azimuthal asymmetries and/or embedded bodies are present in the G ring.
As stated in the Introduction, one of the goals of the HST observations was to retrieve and possibly discover new moons around Saturn. While Janus and Epimetheus were first observed from the ground during the 1966 crossing, only more remote bodies (the trojan librators Helene, Telesto and Calypso) were discovered from the Earth in 1980. In the meantime, the Voyager spacecraft in 1980/81 revealed two more objects on each side of F ring (Pandora and Prometheus), a small object just outside the A ring edge (Atlas), and a moonlet embedded in the Encke gap (Pan).
The HST August data show the presence of several unresolved objects (Nicholson et al. 1995,1996). One of them, already observed during the May crossing (Bosh & Rivkin 1995) has a mean motion coincident with that of Prometheus, except that the object lags the expected longitudinal position of Prometheus by 19. The positions observed for the other small satellites seen in the HST images (Janus, Epimetheus, Pandora, Helene, Telesto and Calypso) agree with their predicted ephemerides. In any case, the accuracy of the 1980/81 Voyager observations of Prometheus should not have caused a longitude discrepancy larger than 3 in 1995. The link of the August Prometheus observations to the May and November observations is discussed more thoroughly by Nicholson et al. (1996). Although no definitive answer for this lag is given yet, some possible explanations are discussed by these authors.
The August HST images show three objects, denoted 1995 S5, 1995 S6 and 1995 S7 (Nicholson et al. 1995,1996). Their orbital radius is derived from their mean motion (see Figure 2), using the known value of Saturn's gravity coefficients , and .
Figure: Observations of Pandora, Prometheus, 1995 S5, 1995 S6 and 1995 S7 by HST on August 10, 1995. The various symbols show the observed separations of each body from Saturn's center vs. time, while the solid curves are fits assuming circular motion around the planet. The horizontal dotted lines show the central times of each HST observational ``windows'', and the horizontal dashed line indicates the expected ring plane crossing time. Finally, the grey region in the center corresponds to the position of Saturn's disk.
These radii appear to be consistent, within the error bars, with that of the F ring (140,180 km). However, none of these objects are retrieved in an obvious way in the May and November data (ibid.). This suggests that they are transient clumps or arcs associated with the F ring, and not fully accreted moonlets. This interpretation is confirmed in the case of 1995 S5, which becomes much fainter at its greatest elongation, as expected for an optically thick arc from purely geometric considerations.
The data presented here are still under study, and the derived results are consequently still preliminary. However, we can quote the following firm points:
Bosh, A.S. & Rivkin, A.S. 1995, IAU Circ. 6192
Bosh, A.S. et al. 1996a, Icarus, submitted
Bosh, A.S. et al. 1996b, Science, submitted
Dones, L., Cuzzi, J.N., & Showalter, M.R. 1993, Icarus, 105, 184
Franklin, F.A., Cook II, A.F., Barrey, R.T.F., Roff, C.A., Hunt, G.E., & De Rueda, H.B. 1987, Icarus, 69, 280
Gresh, D.L., Rosen, P.A., Tyler, G.L., & Lissauer, J.J. 1986, Icarus, 68, 481
Nicholson, P.D. et al. 1995, IAU Circ. 6243
Nicholson, P.D. et al. 1996, Science, submitted
Showalter, M.R., Cuzzi, J.N., & Larson, S.M. 1991, Icarus, 94, 451
Showalter, M.R. & Cuzzi, J.N. 1993, Icarus, 103, 124
Shu, F.H., Cuzzi, J.N., & Lissauer, J.J. 1983, Icarus, 53, 185
Sicardy, B., Lecacheux, J., Laques, P., Despiau, R., & Augé, A. 1982, Astron. Astrophys., 108, 296