Near-infrared spectra of the leading and trailing hemispheres of Enceladus
Introduction
Enceladus displays remarkable diversity in geologic terrains for a small icy satellite. In addition to heavily cratered regions, images returned by both Voyager 2 (Smith et al., 1982) and the Cassini spacecraft currently in orbit around Saturn show areas with few impact craters and ubiquitous grooves and ridges reminiscent of those on Ganymede, suggesting that these may have been tectonically resurfaced (Porco et al., 2006). The explanation for such geologic activity has remained a mystery as there is no obvious source of internal heat to drive the purported resurfacing events. With a mean radius of 252 km and mean density of 1.61 g/cm3 (Porco et al., 2006), Enceladus is too small and too ice-rich to generate sufficient radiogenic or accretional heat (Ellsworth and Schubert, 1983, Squyres et al., 1983) to melt either pure water ice or ammonia hydrate. Recently, Wisdom (2004) investigated the possibility that Enceladus may have been locked in a period-three librational resonance which implies large tidal heating. While its eccentricity is forced by an orbital resonance with Dione, heat generated from tidal dissipation is insufficient to melt a homogeneous Enceladus (Poirer et al., 1983, Squyres et al., 1983); however, tidal heating could melt a multilayered viscoelastic Enceladus in which an internal ocean of ammonia hydrate (Ross and Schubert, 1989) is overlain by a thick (100 km) crust of pure water ice or a thinner (25 km) crust of ammonia hydrate (Kargel and Pozio, 1996). Ammonia hydrates, either in the monohydrate phase, NH3⋅H2O, or dihydrate phase, NH3⋅2H2O, have long been invoked as possible agents of cryovolcanism in the outer Solar System (Lewis, 1972, Consolmagno and Lewis, 1978, Prinn and Fegley, 1981) because their presence in water ice lowers the melting temperature of the peritectic mixture (to ∼176 K for the dihydrate and ∼194 K for the monohydrate (Croft et al., 1988)).
The surface of Enceladus is principally comprised of remarkably bright, pure water ice; its visible geometric albedo (), the highest of any satellite in the Solar System (Verbiscer et al., 2005a), precludes the existence of virtually any non-ice impurities on its surface (Clark, 1982). The fact that water ice dominates the near-infrared spectrum of Enceladus is well established (Cruikshank, 1980, Clark et al., 1983, Cruikshank et al., 1998a, Cruikshank et al., 2005, Emery et al., 2005). Despite the observed diversity in geologic terrains, its global photometric properties are quite uniform (Buratti, 1988, Buratti et al., 1990, Verbiscer and Veverka, 1994), suggesting that the entire satellite may be covered by a thin veneer of water frost. In fact, prior to the Voyager flybys, the first indication that Enceladus was an intriguing world came with the realization that the densest part of Saturn's tenuous E ring coincides with the orbit of Enceladus at (Baum et al., 1981), suggesting an intimate connection between the satellite and ring.
Both pure ammonia and its hydrates have absorption features in the near-infrared (Schmitt et al., 1998). Ammonia monohydrate has possibly been seen in the near-infrared spectra of Miranda (Bauer et al., 2002), Charon (Brown and Calvin, 2000), and the Kuiper belt object Quaoar (Jewitt and Luu, 2004). Because of the decreased melting temperatures, the detection of ammonia and/or its hydrates on the surface of any body in the outer Solar System is of particular interest because its presence enables geologic activity. (The melting temperature of pure ammonia at 1 bar is ∼195 K (Croft et al., 1988), about the same as that of the monohydrate.) According to Consolmagno and Lewis (1978), photolysis would deplete a 10-μm surface layer of ammonia rich ice on a saturnian satellite in only 100 years of exposure to the Sun. In addition Lanzerotti et al. (1984) find that energetic particle bombardment, or sputtering, would deplete the uppermost surface layer of any ammonia molecules on time scales of less than 106 years. Therefore, the presence of ammonia or its hydrates on the surface of an icy satellite implies recent emplacement, possibly by cryovolcanic activity. Hapke (1986b), however, dismisses this argument since sputtering would produce an ammonia-depleted layer too thin to hide any ammonia near the surface. Neither hydrate has yet been detected on the surface of any saturnian satellite, although Emery et al. (2005) reported a possible detection of NH3 in their near-infrared spectrum of the trailing hemisphere of Enceladus.
In an effort to search for these volatiles as well as any hemispherical differences in surface composition, we obtained near-infrared spectra of Enceladus at both eastern and western elongation with the CorMASS spectrograph (Wilson et al., 2001) on the 1.8 m Vatican Advanced Technology Telescope (VATT) at the Mt. Graham International Observatory. The spectral resolution () of these CorMASS data () is comparable to but still higher than that of the Visual-Infrared Mapping Spectrometer (VIMS) () onboard the Cassini spacecraft. Previous high resolution () near-infrared spectra of Enceladus acquired by Emery et al. (2005), also reported by Cruikshank et al. (2005), provided only observations of the trailing hemisphere. Grundy et al. (1999) observed both hemispheres, but their spectral range was limited to 1.2–2.35 μm. Cruikshank et al. (2005) also provided a spectrum of the leading hemisphere, limited to the wavelength range between 1.4 and 2.1 μm. The CorMASS observations reported here represent the first complete spectra at of both the leading and trailing hemispheres of Enceladus spanning 0.8 to 2.5 μm. CorMASS delivers simultaneous observations of this wavelength range via five cross-dispersed orders, mitigating the systematics that would ordinarily arise when splicing independently acquired spectral intervals. We apply Hapke (1993) spectrophotometric theory to produce model fits to each spectrum which enable us to characterize the physical properties of regolith particles on each hemisphere. Such model fits also constrain the amount of ammonia hydrate which may be present in the optically active surface of Enceladus.
Section snippets
Data reduction
Source and spectrophotometric standard data were obtained for Enceladus' leading hemisphere on December 4, 2003 and trailing hemisphere on October 23, 2004 (see Table 1). The CorMASS spectra were reduced using an Interactive Data Language (IDL) package, Cormasstool, based on Spextool (Cushing et al., 2004). Each source and standard image was divided by a normalized flat field which was created by median combining numerous images taken of a continuum lamp which filled the entire slit. Because
Hapke modeling
The Hapke, 1981, Hapke, 1984, Hapke, 1986a, Hapke, 1993, Hapke, 2002 spectrophotometric theory consists of a series of equations, based upon radiative transfer, which relate the reflectance of a particulate surface to the physical properties of its constituents. These properties include particle size, albedo, opacity, and index of refraction (both real and imaginary parts). In addition to the physical properties of individual particles, the Hapke equations also describe how these particles are
Water only mixtures
We fit both hemispheres with two types of areal (checkerboard) mixtures of water ice. The first is a “simple” mixture whose end members consist of uniformly sized particles. The second is a “complex” mixture whose end members are intimate mixtures of two particle sizes. All pure water ice mixture components are at a uniform temperature of 80 K and have a density of 0.933 g/cm3. The results, summarized in Table 3, indicate that the leading hemisphere is best fit by a simple areal mixture of 70%
Uniqueness of solutions
The number of fits presented in Table 3, Table 4 attests to the non-uniqueness of the solutions presented. Certainly alternate combinations of particle sizes and distributions will produce equally acceptable fits to each hemisphere. We can, however, demonstrate that the spectral properties of the leading hemisphere are distinct from those of the trailing hemisphere by attempting to fit the leading hemisphere data with the model which best fits the trailing hemisphere. Similarly, we try to fit
Conclusions
From our analysis of the near-infrared spectra 0.8–2.5 μm of the leading and trailing hemispheres of Enceladus, we find that the trailing hemisphere is dominated by particles which are smaller and/or microstructurally more complex than those on the leading hemisphere. While we do not claim to have found unique solutions which fit each spectrum 0.96 to 2.5 μm, our solutions demonstrate that the two hemispheres are spectrally distinct. The trailing hemisphere is exposed to not only increased
Acknowledgements
We thank Drs. Will Grundy and Ted Roush for generously providing the optical constants for water ice and ammonia hydrate, respectively. We also thank Joshua Emery and an anonymous referee for their careful reviews of the manuscript. P. Helfenstein gratefully acknowledges support from NASA Grant NAG5-11492.
References (62)
- et al.
Saturn's E ring. I. CCD observations of March 1980
Icarus
(1981) - et al.
Edge-on observations of Saturn's E and G rings in the near-IR
Icarus
(1997) - et al.
The near-infrared spectrum of Miranda: Evidence of crystalline water ice
Icarus
(2002) - et al.
Search for volatiles on icy satellites. I. Europa
Icarus
(1988) Enceladus: Implications of its unusual photometric properties
Icarus
(1988)- et al.
Albedo and color maps of the saturnian satellites
Icarus
(1990) - et al.
Near-infrared photometry of the saturnian satellites during ring plane crossing
Icarus
(1998) Implications of using broadband photometry for compositional remote sensing of icy objects
Icarus
(1982)- et al.
Saturn's satellites: Near-infrared spectrophotometry (0.65–2.5 μm) of the leading and trailing sides and compositional implications
Icarus
(1984) - et al.
The evolution of icy satellite interiors and surfaces
Icarus
(1978)