Particle deposition on the saturnian satellites from ephemeral cryovolcanism on Enceladus
Abstract
The geologically active south pole of Enceladus generates a plume of micron-sized particles, which likely form Saturn's tenuous E-ring extending from the orbit of Mimas to Titan. Interactions between these particles and satellites have been suggested, though only as very thin surficial phenomena. We scrutinize high-resolution images with a newly developed numerical shape model of Helene and find that the leading hemisphere of Helene is covered by thick deposits of E-ring particles, which occasionally collapse to form gully-like depressions. The depths of the resultant gullies and near-absence of small craters on the leading hemisphere indicate that the deposit is tens to hundreds of meters thick. The ages of the deposits are less than several tens of My, which coincides well with similar deposits found on Telesto and Calypso. Our findings as well as previous theoretical work collectively indicate that the cryovolcanic activity currently occurring on Enceladus is ephemeral.
Key Points
- Helene's leading hemisphere is covered by thick deposits of E-ring particles
- The E-ring deposits are significantly young (less than several tens of My)
- Young deposits on small satellites indicate Enceladus' activity is ephemeral
1 Introduction
Mid-sized satellites in the E-ring system, such as Tethys and Dione (co-orbital moon of Helene), show almost bimodal distributions of albedo [Buratti et al., 1990; Verbiscer and Veverka, 1992] and visual and infrared spectra [Pitman et al., 2010]. For example, the leading hemisphere of Dione at visible wavelengths is 1.8 times brighter than the trailing hemisphere [Jaumann et al., 2009], which has been suggested to result from differential accumulation of E-ring material on its surface [Ostro et al., 2010; Verbiscer et al., 2007] and/or by bombardment of corotational plasma and energetic electrons [Schenk et al., 2011]. However, no depositional features have been reported on satellites in the E-ring region except for those on Enceladus (e.g., buried craters) [Kirchoff and Schenk, 2009]. We thoroughly examine all high-resolution images obtained by the Cassini spacecraft through November 2013 and find that possible depositional features ubiquitously exist on small satellites in the E-ring region. The best examples are found on Helene because it is the most extensively imaged small satellite in the E-ring.
2 Geological Study of Helene
Like most saturnian satellites, Helene has numerous craters. Using high-resolution images, we identify more than 70 craters (see Crater on Helene in supporting information). Interestingly, Helene shows a bimodal appearance—the heavily cratered trailing hemisphere exhibits a crater density ten times greater than the smooth-looking leading hemisphere (Figure 1). In the case of Enceladus, the deficiency of craters has been suggested to result from a combination of viscous relaxation and burial of craters by both the south polar plume and possibly E-ring material [Kirchoff and Schenk, 2009]. However, a small satellite, such as Helene, is not disturbed by endogenic activity, which provides an optimal situation to examine its possible interactions with E-ring materials.
We study 437 high-resolution images of Helene (500 m/pixel or better) obtained by the Cassini spacecraft through 7 flybys between 2006 and 2011 and find that Helene has numerous, sharply curved gully-like depressions (hereafter streaky depressions; Figure 2). Streaky depressions exist on slopes on the leading hemisphere typically as a group exhibiting similar orientations. These streaky depressions on the leading hemisphere can be identified on images whose resolutions are as high as 200 m/pixel; however, images of the trailing hemisphere, whose resolutions are better than 200 m/pixel, do not show similar features. This indicates that no streaky depressions exist on the trailing hemisphere (Figure 2d).
We construct a numerical shape model of Helene to calculate the local gravitational gradients (Figure 1; see Numerical shape model in supporting information) and critically compare these gradients to the distribution of streaky features (Figure 3). We find that streaky depressions exist only on slopes and strictly follow the local gravitational slopes (Figure 3), which indicates that the streaky depressions result from gravity-induced mass movements. Terrestrial gullies sometimes show structures—such as alcoves (funnelform depressions extending from the top of the slope), channels (linear depressions extending from the narrow stem of an alcove), and fans (cone-shaped deposits crossed by streams)—that indicate transport of material from the top to bottom of the slopes [McClung and Schaerer, 1993]. Analogously to such terrestrial gullies, streaky depressions on Helene sometimes exhibit alcoves and channels (Figures 2b and 2c), further supporting the idea that streaky depressions are formed by gravity-induced mass movements.
The slope angles where streaky depressions exist generally range from 7 ± 3 degrees to 20 degrees, which is significantly smaller than the friction angles of terrestrial rock materials (typically, 25–30 degrees or larger [Lambe and Whitman, 1979]). This observation indicates that streaky depressions form in material which collapses easily and exhibits a small friction coefficient, such as fine particulate material rather than massive ice blocks. Perhaps, such fine particles continuously collapse and form the streaky depressions. We conclude that the leading hemisphere of Helene is generally covered by such fine particles because (i) the leading hemisphere generally appears smooth as evidenced by the lack of shadows, even in areas with large illumination angles; (ii) almost no small craters can be identified on the leading hemisphere; and (iii) large craters exhibit flattened shapes. Together, these lines of evidence indicate that deposition of fine particles has modified or erased craters on the leading hemisphere.
3 Discussion
3.1 Plumes From Enceladus Feed Small Satellites
Spectral analysis [Filacchione et al., 2013] shows that the color of Helene is similar to that of Enceladus. Also, in orbits far beyond Enceladus, satellites are expected to overtake E-ring particles (i.e., Enceladus-derived particles), which would result in preferential deposition on the leading hemisphere of Helene. Moreover, E-ring particles are known to be fine particles [Kargel, 2006], consistent with the apparent fine particulate nature of deposits on Helene's leading hemisphere. These facts indicate that the deposit on Helene comes from the E-ring.
Our findings indicate that ice particles ejected from Enceladus are preferentially deposited on Helene's leading hemisphere, and that the resulting deposits occasionally collapse to form streaky depressions. Motivated by this finding, we identified similar deposits on other small satellites in the E-ring region, such as Telesto and Calypso (Tethys' Trojan satellites), Pallene, and Methone (Figure 2 and S1 in supporting information), where higher brightness (i.e., presumably higher densities) of the ring is reported [Verbiscer et al., 2007]. For example, craters on Telesto and Calypso generally exhibit softened, blanketed morphologies with indistinct rims (and sometimes such craters are almost entirely erased), similar to craters on the leading hemisphere of Helene. In fact, even large (>5 km diameter) craters on Telesto and Calypso appear to be buried just as small craters are on Helene. Also, Calypso exhibits streaky depressions (Figure 2e), which are similar to those on Helene. Spectral observations demonstrate the contribution of E-ring material on their surfaces [Buratti et al., 2010], which supports the view that, as with Helene, E-ring material accumulated on these satellites into thick deposits. Moreover, high-resolution images of Pallene (Figure 2g), one of the three Alkyonides satellites, indicate a circular outline with a sharp terminator line lacking any undulations even at the highest resolution of about 500 m/pixel. This implies that Pallene has a featureless spherical shape, which is unusual for a body of only ~2.2 km radius. Methone (Figure 2h), one of the other two Alkyonides satellites, is likely quite similar to Pallene in terms of size, shape, and smooth appearance [e.g., Thomas et al., 2013]. The shapes of these bodies are also possibly related to the accumulation of E-ring material on their surfaces, as such material can cover the original irregular topography.
3.2 Lack of Hemispheric Dichotomy for Telesto/Calypso
Unlike Helene, Telesto and Calypso appear to have smooth surfaces and few small craters even on their trailing hemispheres (though Calypso's trailing hemisphere has yet to be observed at high resolution), which suggests that Telesto and Calypso lack any dichotomy (see Figure S1 in supporting information). This diversity may be explained by the motion of the E-ring particles.
The mean velocity difference of a particle ejected from Enceladus relative to an encountered satellite will depend on distance from Saturn. This would naturally cause the impact distribution of particles onto the satellite surface to vary with distance from Saturn as well (see Lack of hemispheric dichotomy for Telesto/Calypso in supporting information). (i) Inside the orbit of Enceladus, particles generally move faster than encountered satellites, which results in deposition on the trailing hemisphere, as is observed for Mimas. (ii) In the orbit immediately beyond Enceladus, the mean relative velocity is modest but non-zero. Particles will move both faster and slower than encountered moons, but on average will be slightly slower. This can explain the small but not negligible albedo difference (1.1 times) between the leading and trailing hemisphere of Tethys (a co-orbital moon of Telesto and Calypso). Still, particles will encounter both the leading and trailing hemisphere of encountered satellites, and this can explain the deposition everywhere on Telesto and Calypso. (iii) Finally, in orbits far beyond Enceladus, the mean relative velocities may be large, with satellites rapidly overtaking Enceladus-derived particles, which results in preferential deposition on the leading hemisphere of Helene, Dione, and Rhea. We note that bombardment by corotational plasma or energetic electrons also influences the color and albedo pattern on mid-sized E-ring satellites [Schenk et al., 2011].
The lack of a hemispheric dichotomy on Telesto and Calypso may be due to not only E-ring particle dynamics but also to possible non-synchronous rotation of these satellites. Non-synchronous rotation re-orients the satellite, causing the leading and trailing hemispheres to migrate across the satellite figure and preventing preferential particle deposition on any specific hemisphere of the satellite. In addition to Telesto and Calypso, Pallene and Methone also appear to lack any hemispheric dichotomy. This may likewise be explained by E-ring particle dynamics or non-synchronous rotation; however, the depositional mechanics on Pallene or Methone may be more complex (see Deposition of ring particles on Pallene and Methone in supporting information).
3.3 Thickness of E-Ring Deposits
Shadows of streaky depressions indicate that their depths are typically a few tens of meters, which implies that the particle deposits in which the depressions form must be at least tens of meters thick. Moreover, the leading hemisphere shows the near-absence of small craters (less than ~3 km across; Figure 1c). This indicates that the deposits are unlikely to be more than a few hundred meters thick. Therefore, we estimate that the thickness of the deposits on Helene accumulated from the E-ring is between 10 and 300 m.
On the other hand, Telesto and Calypso almost completely lack large (>5 km diameter) craters. Moreover, the few existing craters are almost entirely erased. This fact suggests that E-ring deposits on Telesto and Calypso are roughly twice as deep as those on Helene. This difference may be due to the densities of E-ring. The E-ring is known to be denser closer to Enceladus [Verbiscer et al., 2007]. Therefore, the E-ring at the orbit of Telesto/Calypso exhibits higher brightness than at the orbit of Helene, which may explain the thicker deposits on Telesto/Calypso relative to those on Helene.
Unlike the deposits on these small satellites, the E-ring deposits on Tethys and Dione are probably quite thin because (1) these satellites' radar-optical albedo appears to decrease with distance from Enceladus [Ostro et al., 2010]; (2) crater statistics [Kirchoff and Schenk, 2010] indicate no deficiency of craters on either Tethys or Dione, in contrast to Helene and Telesto; and (3) high-resolution images of Tethys and Dione show no unambiguous evidence for thick deposits, such as streaky depressions. Nevertheless, albedo [Verbiscer et al., 2007], spectral [Filacchione et al., 2010], and thermal inertia [Howett et al., 2010] measurements indicate thin but non-zero deposits of E-ring particles on these mid-size satellites. Thus, the E-ring particles are likely deposited widely on Tethys and Dione, but the resulting deposits are much thinner than those on the small satellites. The reasons for this difference are unknown. Possible explanations are differences in the dynamics of E-ring particles among the satellites or higher impact velocity onto the mid-sized satellites.
4 Implication for Ephemeral Cryovolcanism on Enceladus
We perform an age estimate of the E-ring deposits based on the cratering rate. The high crater density on the trailing hemisphere of Helene indicates that Helene is basically an old object. Based on the crater chronology of the saturnian system [Zahnle et al., 2003], we estimate that the surface age of the heavily cratered terrain on the trailing hemisphere is ~4.0 Gy (and at least ~1.0 Gy), which coincides with the age estimated from the distribution of large (>5 km) craters (Figure 1c). We also find that the distributions of craters exceeding 5 km in diameter on Helene and Telesto are generally similar to those of Dione and Tethys [Kirchoff and Schenk, 2010] (Figure 1c), which indicates that, in general, the original crater densities of small satellites are similar for these mid-sized satellites. Thus, Helene, Telesto, and Calypso probably have similar formational ages, which are significantly older than that of the E-ring deposits.
On Helene's E-ring deposit, whose area is 1637 km2 estimated from our shape model, we identify five craters of ~200 m diameter but no craters exceeding 1 km in diameter. If we adopt standard cratering-rate estimates for the outer solar system [Zahnle et al., 2003], the best estimates for deposit ages are several tens of My or younger (see Age estimates based on cratering rate in supporting information). The crater size-frequency distribution of the trailing hemisphere of Telesto is similar to that of the leading hemisphere of Helene.
Interestingly, several lines of evidence suggest that Enceladus has not been active at its current level over solar-system history. For example, Kargel [2006] shows that no geological evidence exists to support a large change in its radius, which would be expected if the current mass-loss rates have been maintained through Enceladus' lifetime. Moreover, Roberts and Nimmo [2008] show that the current energy output of Enceladus is difficult to sustain over solar-system history. Enceladus' current heat flux greatly exceeds that occurring in equilibrium with its current eccentricity given the time-averaged Q of Saturn [Meyer and Wisdom, 2007], suggesting that the current activity may be greater than average. Recent global geophysical models show that Enceladus can exhibit episodic activity, with short periods of intense activity interspersed with long quiescent epochs [Showman et al., 2013]. These studies are consistent with our views. Thus, there is the possibility that the accumulation of E-ring material on small satellites in this region began several My ago as a result of the initiation of cryovolcanism of Enceladus.
Acknowledgments
This work is supported in part by Grant-in-Aid for JSPS Fellows (to NH), JSPS KAKENHI (to HM), and the NASA Origins program (to APS). We use the raw data freely available via NASA's Planetary Data System (http://pds.nasa.gov/).
The Editor thanks two anonymous reviewers for their assistance in evaluating this paper.
