P43B-01 13:40h
The Surface Composition of Saturn's Moon Phoebe As seen by the Cassini Visual and Infrared Mapping Spectrometer
The Cassini-Huygens spacecraft encountered Phoebe on June 11, 2004. The Visual and Infrared Mapping Spectrometer (VIMS) obtained spatially resolved hyperspectral images of Phoebe at 352 wavelengths (0.4-5 microns) from 8:47 UT June 11, 2004 at at an initial range of 245,833 km and phase angle of 84.9 degrees, to 10:22 UT June 12, 2004 at a final range of 338,401 km and a phase angle of 92.2 degrees. The closest image was obtained on June 11 at 19:32 UT at a range of 2,178 km and solar phase angle of 24.6 degrees. The spatial mapping of the VIMS, with an instantaneous field of view of 0.25 by 0.5 milliradian resulted in spatial coverage at full spectral resolution as small as 1 km/pixel. The spatially resolved spectra of Phoebe indicate a low surface albedo, from < 1 to ~6% reflectance with a variety of absorption features due to materials which occur with variable abundances and/or grain sizes in different locations on the body. These include: water ice (previously identified by Owen et al, 1999), bound water, and trapped CO2. A broad 1-micron feature is interpreted to be due to Fe2+ bearing minerals. Water ice is observed with absorptions at 3.1, 2, 1.5, 1.25, and 1.04 microns. Variable absorption strengths indicate a variety of ice abundances ranging from almost non detect to > 50% and grain sizes less than a few hundred microns. Absorptions located in the 3.3 and 1.7 micron region indicate the presence of organic molecules, and a prominent absorption at 2.42 microns is best explained by a cyanide compound. Spectral structure at 4.8-5 microns is also consistent with cyanide compounds. Absorptions at 4.5, 3.3 and 1.7 microns indicate a probable nitrile compound. These compounds argue for either an outer solar system origin for Phoebe, or that its surface has been coated with outer solar system materials.
P43B-02 13:55h
Ultraviolet Observations of Icy Saturnian Satellites from Cassini
We present results from the Ultraviolet Imaging Spectrograph (UVIS) measurements of Phoebe during the Cassini flyby on June 11, 2004. Cassini UVIS mapped the illuminated side of Phoebe at resolutions on the order of 5-10 km/pixel. Phoebe is known to be dark and spectrally flat at all visible wavelengths, similar to C-type asteroids. Unresolved Earth-based observations have detected the presence of water ice. We find that the normal albedo of Phoebe is very low ( < 0.05) at far-UV wavelengths (110-190 nm), similar to carbonaceous chondrites. Phoebe's FUV spectral albedo, however, is not flat like such meteorites, suggesting that another species strongly affects the FUV spectrum. We investigate candidate species, and study albedo and spectral variations across the surface, to map distributions of species. We also present Phoebe's UV lightcurve, measured during both the inbound and outbound portions of the flyby. In addition to the reflectance studies we investigate the presence of any emissions from gaseous species. Phoebe's spectrum does not display any strong emission features, suggesting that gaseous species are not present and/or that the density of solar wind electrons in Phoebe's environment is too low to produce detectable excitation emission features. We also report on the disk-integrated observations of Iapetus throughout October, 2004, at ranges between 1 and 2 million km. The FUV spectra of Iapetus will be compared with the Phoebe data to investigate a possible relationship between these two satellites. On October 28, 2004 UVIS will obtain disk-resolved measurements of Tethys from a range of 255,000 km. Tethys represents an icy body that orbits well within Saturn's magnetosphere and the E-ring; we will compare and contrast FUV spectra of Tethys with those of the outer satellites to look for effects of different exogenic processes.
P43B-03 14:10h
Cassini Composite Infrared Spectrometer (CIRS) Observations of Phoebe's Thermal Emission
The Cassini Composite Infrared Spectrometer (CIRS) has three focal planes that together cover the thermal spectrum from 10 to 1420 cm$^{-1}$ (1 mm to 7 $\mu$m). Focal plane 1 (FP1) covers 10 to 600 cm$^{-1}$ with a circular field of view of 3.9 milliradians diameter. Focal plane 3 (FP3) covers 600 to 1100 cm$^{-1}$, and focal plane 4 (FP4) covers 1100 - 1420 cm$^{-1}$, both with a linear array of 10 pixels, each 0.3 milliradians square. Spectral resolution is selectable from 0.5 to 15.5 cm$^{-1}$ (apodized). During the Cassini flyby of Phoebe on June 11$^{\rm th}$ 2004, CIRS obtained numerous observations of thermal emission from its surface with all three detectors, though FP4 only detected emission from the very warmest regions. FP3 achieved a spatial resolution as small as 12 km for full-disk observations, and 600 m for local observations near closest approach. Spectral resolving power for most observations was about 50. The signal to noise ratio (S/N) of the FP3 observations allows measurement of brightness temperatures as low as $\sim$ 75 K but is not sufficient to detect any deviations from blackbody behavior in the Phoebe data. There is strong topographic control of temperature, particularly around the prominent large crater seen in Cassini visible-wavelength images, where temperatures were observed over nearly half a Phoebe rotation. Low-latitude temperatures on Phoebe vary between 82 K before dawn to 112 K near the subsolar point. The diurnal variation can be matched with a thermal inertia near 3 x 10$^4$ erg cm$^{-2}$ s$^{-1/2}$ K$^{-1}$, about half the thermal inertia derived from diurnal temperature variations on the Galilean satellites but similar to that of Rhea, Dione, and Tethys. This low thermal inertia implies that the upper centimeter of Phoebe's surface is covered in very porous material. FP1 had lower spatial resolution than FP3 (near 9 km at closest approach), but higher S/N, allowing extraction of spectral information. At 50 cm$^{-1}$ (200 $\mu$m) brightness temperatures varied from 68K in predawn locations, to 101K near the subsolar point. However, strong spectral gradients were apparent in all spectra, with brightness temperatures at 400 cm$^{-1}$ (25 $\mu$m) in the above instances varying from 76K to 111K. No strong emissivity variations are present. Consequently, these spectral characteristics are due to a combination of unresolved thermal inertia and slope (shadowing) variations. These effects will be discussed.
P43B-04 14:25h
A Gas-poor Planetesimal Feeding Model for the Formation of Giant Planet Satellite Systems: Prediction for the Composition of Iapetus
Given our presently inadequate understanding of the turbulent state of the solar nebula and planetary nebulae, there are two sensible approaches to satellite formation that avoid over-reliance on specific choices for essentially free parameters. The first one postulates turbulence decay. If so, Keplerian disks must eventually pass through quiescent phases, so that the survival of satellites (and planets) ultimately hinges on gap-opening. In this scenario, the criterion for gap-opening itself sets the value for the gas surface density of the satellite disk (Mosqueira and Estrada 2003b). The second approach assumes that steady turbulence is sufficiently strong to cause the evolution of the gas disk on a shorter timescale than that for satellite formation. This approach uses the turbulence of the subnebula to remove gas from the disk but not to fine-tune the conditions of the subnebular environment. In this case, the gas surface density is left unspecified, though the presence of some gas may help to explain the observations. Satellite formation is then understood in terms of planetesimal dynamics that are largely uncoupled from the gas (somewhat analogous to the case of the terrestrial planets). We will discuss a gas-poor model with the following features: First, collisions between planetesimals in the vicinity of the giant planet leads to the formation of a protosatellite swarm of prograde and retrograde objects extending as far as $\sim R_H/2$ (Ruskol 1975, Safronov {\it et al.} 1986). Second, this circumplanetary swarm has a small net specific angular momentum which results in the formation of close-in, prograde satellites. Third, close to the planet, hypervelocity impacts can ultimately lead to a variety of outcomes (i.e., Jovian-like versus Saturnian-like satellite systems). Fourth, satellitesimal collisional removal from the outer disk is balanced by planetesimal collisional capture. Excluding satellite embryos, at any given time this disk mass is less than the mass of the regular satellites. Fifth, a satellite formation timescale of $10^5-10^6$ years (consistent with a partially differentiated Callisto) controlled by the feeding of planetesimals onto the circumplanetary disk (Mosqueira et al. 2000). It might be possible to concoct a turbulent mechanism operating following a giant impact between Titan and a Triton-sized differentiated interloper (Mosqueira and Estrada, this conference) that leads to the spread of a volatile-rich disk. However, such a mechanism is very unlikely to work inasmuch as it would require an unrealistic angular momentum budget, particularly if one considers gas drag inward migration of Iapetus (gas drag would be needed to account for the lack of objects between Titan and Iapetus). Instead, the angular momentum of material fed from heliocentric orbit (gas or solids) strongly implies that Iapetus (like Callisto, $\rho = 1.85$ g cm$^{-3}$) should be of roughly solar composition. This statement constitutes a prediction of this model and requires that the present value of the density of Iapetus ($1.14\pm 0.1$ g cm$^{-3}$, Jacobson, pers. comm.) be in error. That is, {\it within the context of a planetesimal feeding model}, Phoebe's density suggests that one should expect $\rho > 1.6$ g cm$^{-3}$ for Iapetus.
P43B-05 14:40h
Cassini RADAR Observations of Phoebe
The Cassini RADAR instrument, operating in its scatterometry mode, obtained continuous-wave (cw) echo power spectra from Phoebe during the inbound and outbound legs of the flyby, 4 h before and 2.5 h after closest approach. Phoebe's distance and subradar coordinates were approximately (93,000 km, 247 deg W, 26 deg S) inbound and (56,000 km, 323 deg W, 26 deg N) outbound. The durations of the cw sequences were 6 and 5 minutes. Larger intervals in the RADAR windows were devoted to observations with a chirp waveform able to provide range as well as Doppler resolution, and to passive radiometry; those data are not yet reduced. For Phoebe (and Dione, Mimas, Iapetus, Enceladus, Rhea, Hyperion, and Tethys), scatterometry aims to use estimates of radar albedo and angular scattering law to constrain the near-surface bulk density and/or the relative cleanliness of the icy regolith. The RADAR instrument's wavelength is 2.2 cm, vs. 3.5 cm or 13 cm for most groundbased radar astronomy, but Arecibo and Goldstone observations of the icy Galilean satellites and of asteroids give us no reason to expect significant wavelength dependence in this regime. Comparison of RADAR measurements of Titan and Iapetus with groundbased results will let us evaluate this expectation and will be key to calibrating both our measurements and their interpretation. Our inbound and outbound Phoebe echoes indicate Lambertian scattering, which requires structural complexity at scales no smaller than a centimeter. However, despite Phoebe's prominent large-scale topography, our spectra are nearly featureless, suggesting that the radar roughness is sub-topographic. We probably are seeing a combination of single scattering and multiple scattering from surface and subsurface structure. At this writing, our calibration indicates that Phoebe's average radar albedo is much closer to that of Iapetus than to those of the icy Galilean satellites.
P43B-06 14:55h
Phoebe and the Icy Saturnian Satellites: Implications for Satellite Origins
Phoebe's retrograde, eccentric and inclined orbit marks it as an object captured from heliocentric orbit. Accordingly, its composition may be indicative of its origin in the solar nebula. Analogous arguments have been made extensively in connection with the origin of Pluto-Charon (see, e.g., McKinnon et al. 1997) as well as Triton (McKinnon and Mueller 1989). Indeed, the demarcation between nebula and subnebula objects has led a number of workers (see, e.g., Johnson et al. 1987; Lunine et al. 1993; Podolak et al. 1993) to argue that the regular satellites of the giant planets did not derive the bulk of their material directly from heliocentric orbit. The recent Cassini flyby of Phoebe has yielded a mass for this object of $GM = 0.5527 \pm 0.001$ km$^3$/s$^2$ Jacobson et al. 2004. Its density of $1.6$ g/cm$^3$ indicates a rock to ice ratio of at least $50 %$ (Porco et al. 2004; Science, to be submitted). Phoebe's high rock/ice ratio when compared to the icy Saturnian satellites reinforces the argument that Phoebe is an object that formed in heliocentric orbit and became captured. Yet, given that it may be misleading to lump together satellites with quite different formation histories, we refine the comparison on the basis of models for regular satellite formation. Because it derives condensables directly from heliocentric orbit and fails to consider planetesimals, the model of Canup and Ward (2002) does not provide a context for understanding such compositional differences. We will therefore discuss two models of satellite formation we are developing, which differ mainly in their treatment of turbulence (decaying vs steady). In both models the inner (located inside Titan's orbit), icy Saturnian satellites represent a second generation of objects. Mosqueira and Estrada (2003a,b) has these satellites forming $10^4-10^5$ years after Titan as the disk became optically thin and water rich due to preferential gas drag loss of silicates as Saturn cooled. On the other hand, the gas-poor planetesimal-capture model of Estrada and Mosqueira (2003, 2004, this conference) has them forming from the impact ejecta (which presumably avoided re-accretion by gas drag inward migration) between Titan and a Triton-sized differentiated interloper, leading to Titan's eccentricity and likely causing it to differentiate (assuming it was not differentiated to begin with). In either case, these satellites may not be representative of the bulk composition of regular satellites. Furthermore, Titan's higher density and possible size-selective devolatilization (Stevenson et al. 1986) may also cloud the link between origin and composition. However, we argue that Iapetus would not have been affected by these processes, and so it may furnish a more direct test of whether the regular satellites of Saturn could have derived the bulk of their material directly from heliocentric orbit. At the time of submission, Iapetus' mass $GM = 118 \pm 11$ km$^3$/s$^2$ (though a systematic source of error hasn't been ruled out; Jacobson, pers. comm.) and mean radius of $718 \pm 8$ km (Davies and Katayama 1984) yields a density of $1.14 \pm 0.1$ g/cm$^3$, which implies processing of planetesimal composition prior to regular satellite formation and favors the satellite formation model of Mosqueira and Estrada (2003a,b), but future Cassini Iapetus flybys may be needed to settle this issue. It is possible that by the time of this conference an improved constraint on this number will be available (Jacobson, pers. comm.).
P43B-07 15:10h
Cold Compaction of Porous Ice and the Density of Phoebe
A series of experiments was carried out to measure the hydrostatic compaction of granular ice I as a function of pressure at low temperatures. Most runs were conducted on the same initial size distribution of ice granules (0.18 - 0.25 mm) and most were conducted at the same temperature of 77 K. For one run the size range of the granules was wider (0.25 - 2.0 mm), and for one run the temperature was 120 K. Granulated ice was sealed in an indium metal jacket, the sample placed into a cyrogenic pressure vessel, and hydrostatic pressure applied. Initial and final porosity were calculated from mass and volume measurements. Porosity during the experiments, when the sample was inside the pressure vessel undergoing compaction, was calculated from the length of the sample (measured in situ) assuming that porosity change and length change were proportional. The starting and ending shapes of the capsule were always right cylindrical, suggesting that the compaction was uniform throughout the samples. In the experiments, the rate of change of volume with pressure was highest at low pressure and decreased monotonically with increasing pressure. Surprisingly, the samples continued to compact with increasing pressure even at the highest pressures achieved (150 MPa). This observation is consistent with the existence of substantial residual porosity (10-15%) measured in the samples and confirmed by SEM observation after testing. The sample tested at 120 K had a compaction curve indistinguishable from the those of the other three samples of 0.18-0.25-mm ice, all tested at 77 K. Because creep is generally a very temperature-sensitive phenomenon, we infer that creep was not an important process in these tests. The sample with the wider range of granule size started with lower porosity (as expected), and ended with lower porosity. We conclude that over the interior pressures found in smaller midsize icy satellites and Kuiper Belt objects (KBOs), significant porosity can be sustained over solar system history in the absence of significant heating and sintering. Phoebe's Cassini-derived density of 1.6 g cm$^{-3}$ is consistent with a solid, non-porous density of 2.0 (from Pluto and Triton, the largest KBOs), mechanical domination by ice I, and an average porosity of 20%. Interior pressures for Phoebe reach $\sim$5 MPa; the porosity of the experimental sample with the wider granule size range was $\sim$25-30% in this modest pressure range. Thermal evolution models for porous ice-rock bodies of Phoebe's size ({\it Proc. ACM 2002}, {\bf ESA-SP-500}, 29-38) indicate peak interior temperatures $ < $150 K after 50 Myr, so in the absence of creep-driven densification, we hypothesize that Phoebe's "modest" porosity is due to its being built from a wide size spectrum of ice-rock fragments.
P43B-08 15:25h
Orbital Evolution of Impact Ejecta from some of the Biggest Craters on Saturn's Icy Satellites
We use the Swift numerical integration package written by Levison and Duncan (1994, Icarus 108, 18-36) to compute the orbital evolution of impact ejecta in the Saturn system. We consider four giant craters on three satellites: Herschel on Mimas, Odysseus and Penelope on Tethys, and Tirawa on Rhea. Ejection velocities and particle sizes are consistent with what is currently known about cratering physics. We consider impacts on competent surfaces (the spallation model of Melosh 1984, Icarus 59, 234-260) and into unconsolidated regolith (the gravity-scaling model of Housen et al. 1983, J.G.R. 88, 2485-2499). These two models should bracket the behavior of real impact ejecta. From each crater we launched 600 test particles according to each model at velocities comparable to or exceeding the satellite's escape speed. Most test particles are swept up by the source moon on time-scales of a few to several decades, and produce craters no larger than a few kilometers in diameter. A small but non-negligible fraction of material reaches satellites other than the source moon. Our models generate cratering patterns consistent with a planetocentric origin of most small impact craters on the Saturnian icy moons, but the predicted craters tend to be smaller than canonical Population II craters. We conclude that the known giant craters in the Saturnian system cannot account for the full range of observed Population II craters. A more complicated story, presumably one involving the disruption of co-orbital moons, is probably required to generate the larger Population II craters.