Regular articleOceans in the icy Galilean satellites of Jupiter?
Introduction
Observations of electromagnetic induction signatures at Europa and Callisto have been interpreted as indicative of sublithosphere liquid water oceans in these satellites Khurana et al 1998, Neubauer 1998, Kivelson et al 1999, Kivelson et al 2000, Zimmer et al 2000. The magnetic field data gathered during several close fly-bys of Ganymede have also been proposed to be in part due to induction in an ocean in this satellite as well (Kivelson et al., 2002). If these interpretations are correct, then the magnetic field data will allow estimates of the thickness of the ice layers covering the oceans. For Europa, the strength of the induced signal suggests an ocean underneath a thin ice shell a few tens of kilometers thick. Hussmann et al. (2002) have shown that tidal heating in such an ice shell can keep a subsurface europan ocean from freezing. For Ganymede and Callisto the oceans are at greater depths of 100 km or more. For Europa, the existence of an ocean underneath tens of kilometers of ice is supported by geological evidence (Pappalardo et al., 1999) and some authors favor even thinner ice layers (e.g., Greenberg et al 1998, Kattenhorn 2002. Similar geologic evidence is not available for Ganymede and Callisto. At least for Callisto, the inferred existence of an ocean is surprising since this satellite is incompletely differentiated Anderson et al 2001, Sohl et al 2002 and shows little sign of past endogenic activity (as witnessed by its old surface).
Subsurface oceans on satellites are possible because of the anomalous melting behavior of ice I for which the melting temperature decreases with pressure until it joins the ice I/ice II transition and the ice III melting curve in a triple point at a pressure of 207 MPa and a temperature of 251.15 K (Chizov, 1993; Fig. 1 ). This triple point pressure translates into different depths for the three satellites because of their differing masses and, possibly, ice shell densities. For Europa, assuming an ice shell density of 1000 kg m−3, a depth of 160 km is obtained, about as deep as the thickness of the water layer, which is believed to be around 150 km Anderson et al 1998, Sohl et al 2002. The depth of the minimum melting temperature is about 145 km in Ganymede if an ice shell density of 1000 kg m−3 is assumed. This density is reasonable since Ganymede is most likely differentiated (Anderson et al., 1996). For Callisto, the ice shell density could be about 1600 kg m−3 if the satellite is undifferentiated or has a mostly undifferentiated, cold and stiff outer shell. The value is calculated from the density of ice I, a concentration of rock of about 50 wt%, and a rock density of 3500 kg m−3. With the latter ice shell density, the depth to the triple point pressure is about 104 km. If the outer shell in Callisto were pure ice, the depth to the triple point would be about 166 km.
It is possible that the melting point is even further depressed if the ice in the satellites is not pure H2O but contains other components such as ammonia, methane, and/or salts (e.g., Kargel, 1992). The phase diagram of the water–ammonia system is well studied (e.g., Hogenboom et al 1997, Sotin et al 1998 and can serve as a model. Ammonia hydrates are predicted condensates in the satellites of Jupiter (e.g., Lewis, 1971) and the concept of internal layers of ammonia–water liquid is well developed in the literature (see, e.g., Kargel 1998, Sotin et al 1998 for reviews). The water–ammonia liquidus temperature depends on pressure and on the concentration of ammonia in the water. The liquidus surface in a concentration–pressure–temperature phase diagram contains the points 5 wt% NH3, P = 0.1 MPa, and T = 266.92 K; the low-pressure peritectic at 32.1 wt% NH3, 0.1 MPa, and 176 K, and the high-pressure eutectic at 29 wt% NH3, 170 MPa, and 176 K. The solidus up to 170 MPa is an isothermal surface at 176 K to a good approximation.
The evolution of an ice shell and ocean will depend not only on the bulk concentration of ammonia but also, to some extent, on the initial conditions. If the shell grew on top of an ocean with some initial concentration of ammonia and began from zero thickness, then the shell will be pure water ice. The composition of the ocean will be determined by the mass of the water ice removed from the ocean and the constancy of the ammonia mass. The ice shell bottom temperature will be the liquidus temperature and will be determined by the composition of the ocean and the pressure at the bottom of the lid. As the shell thickens, the shell bottom temperature will describe a path on the liquidus surface. Grasset and Sotin (1996) have suggested a possible path on the liquidus surface starting at 5 wt% NH3, P = 0.1 MPa, and T = 266.92 K and continuing to the high-pressure eutectic at 29 wt% NH3, 170 MPa, and 176 K (compare Fig. 1). The pressure of 170 MPa is equivalent to a depth of about 130 km on Europa, 120 km on Ganymede, and 135 km on Callisto.
If the satellite were initially at subsolidus temperatures and was warmed to reach the ammonia hydrate solidus, a partial melt would form at a temperature of 176 K and mostly independent of pressure. The composition of the liquid, under these circumstances, will be around 30 wt% ammonia, decreasing slightly with pressure, because the composition along the eutectic/peritectic is almost constant. The solid phase in the partial melt will be H2O ice. Since the ice is less dense than the fluid, the ice will tend to float on top of the forming ocean. The rate at which the ice will separate and float will depend on the rheology and the concentration of the melt in the solid. As the temperature increases, water will dilute the melt. The increasing concentration of the melt will further the formation of an ocean by facilitating the gravitational separation of ice from the melt. The composition of the ocean once formed and the temperature at the base of the ice shell will be determined by the pressure at the base of the ice shell, the bulk concentration of ammonia, and the liquidus surface.
Since we are interested in equilibrium configurations, we will neglect any rates of separation of the ice from the liquid. Rather, we assume that ice and melt have separated and that the composition of the ocean and the temperature at the bottom of the ice shell are given by the Grasset and Sotin (1996) liquidus.
For Europa, it is possible to attribute the existence of an ocean to tidal heating. This is not possible for Ganymede and Callisto; tidal heating is negligible at their present orbital distances and eccentricities. On these satellites, oceans must be due to radiogenic heating or to heat buried at depth and released through satellite cooling. It has been speculated (e.g., Malhotra 1991, Showman and Malhotra 1997, Showman et al 1997 that Ganymede went through a phase of intense tidal heating perhaps as little as 1 Gyr ago when it passed into the present resonance through a 3:2 resonance. During such a transition, the eccentricity would have increased by as much as a factor of 10, which would have increased the heating rate by approximately a factor of 2 above the radiogenic heating rate (Spohn and Breuer, 1998).
In this paper, we will investigate the possibility of present oceans in the Galilean satellites as a consequence of an assumed heat flow from below the ocean. That heat flow may be due to radiogenic heating, tidal heating, and/or satellite cooling, but the model is general and does not a priori attribute the heat flow to a particular mechanism.
Section snippets
The model
The model is shown in Fig. 2. An ice layer of thickness D lies on top of an ocean of thickness Do. The basal temperature of that ice layer is Tb. The thickness of the ocean is determined by equating Tb with both branches of the ice liquidus that define the dip in the ice melting temperature profile with pressure (depth) as shown in Fig. 2. The heat flow from the deeper interior is q. The existence of an ocean is dependent on the rate of heat transfer through the top ice layer of thickness D.
Parameter values
The most important model parameters are the thermal conductivity of the ice, k, the convective heat transfer parameters a and β, the rheology parameters, and the heat flow from below. The thermal conductivity of ice is known to be an inverse function of temperature. For compact, hexagonal crystalline ice Ih the empirical relation is Klinger 1973, Ross and Kargel 1998 where temperature T is in kelvins. In the temperature range of interest, between 100 and 300 K, this translates into a
Results
The results of the calculations are shown in Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7. Fig. 3, Fig. 4, Fig. 5 show results for H2O ice while Fig. 6 shows results for NH3–H2O ice. Plotted in Fig. 3 are lines of constant ice shell thickness scaled by the depth l0 to the triple point in a plane defined by the scaled heat production rate Θ and the scaling Nusselt number Nu0. To be exact, the lines of constant ice shell thickness are strictly valid only for Ganymede. However, the error introduced by
Discussion and conclusions
In the following we will defend a main conclusion that we draw from our model calculations: While an ocean can hardly be doubted for Europa, oceans are possible for Ganymede and a (partially) differentiated Callisto because we find that equilibrium states plot close to the transition line between oceans and complete freezing even for pure H2O. Pure H2O models are extreme because they have the highest melting temperature. Equilibrium models are extreme because satellite cooling should contribute
Acknowledgements
Part of this research was carried out while T.S. spent a sabbatical semester at UCLA and during a later visit. T.S. thanks M. Kivelson, K. Khurana, and R. Walker for fruitful discussions and their hospitality. This research was supported by the Deutsche Forschungsgemeinschaft and by NASA Planetary Geology and Geophysics under Grant NAG 5-3863.
References (56)
- et al.
Shape, mean radius, gravity field and interior structure of Callisto
Icarus
(2001) - et al.
The cooling rate of a liquid shell in Titan’s interior
Icarus
(1996) - et al.
Tectonic processes on Europatidal stresses, mechanical response, and visible features
Icarus
(1998) - et al.
Magnesium sulfate-water to 400 MPa using a novel piezometerdensities, phase equilibria, and planetological implications
Icarus
(1995) - et al.
The ammonia–water system and the chemical differentiation of icy satellites
Icarus
(1997) - et al.
The effect of pressure on the thermal conductivity of silicate rocks up to 12 kbar
Phys. Earth Planet. Inter.
(1989) - et al.
Thermal equilibrium states of Europa’s ice shellimplications for internal ocean thickness and surface heat flow
Icarus
(2002) Ammonia–water volcanism on icy satellitesphase relations at 1 atmosphere
Icarus
(1992)- et al.
Rheological properties of ammonia–water liquids and crystal–liquid slurriesplanetological applications
Icarus
(1991) Nonsyncronous rotation evidence and fracture history in the bright plains region, Europa
Icarus
(2002)
The permanent and inductive magnetic moments of Ganymede
Icarus
Thermal history of the Moonimplications for an early core dynamo and post-accretional magmatism
Adv. Space Res.
Satellites of the outer planetstheir physical and chemical nature
Icarus
Tidal origin of the Laplace resonance and the resurfacing of Ganymede
Icarus
Mystery of Callistois it undifferentiated?
Icarus
Line heat-source measurements of the thermal conductivity of porous H2O ice, CO2 ice and mineral powders under space conditions
Planet. Space Sci.
Tidal evolution into the Laplace resonance and the resurfacing of Ganymede
Icarus
Coupled orbital and thermal evolution of Ganymede
Icarus
Implications from Galileo observations on the interior structure and chemistry of the Galilean satellites
Icarus
The longevity of lunar volcanismimplications of thermal evolution calculations with 2D and 3D mantle convection models
Icarus
Subsurface oceans on Europa and Callistoconstraints from Galileo magnetometer observations
Icarus
Gravitational constraints on the internal structure of Ganymede
Nature
Europa’s differentiated internal structureinferences from four Galileo encounters
Science
Three-dimensional thermal convection in a spherical shell
J. Fluid Mech.
Transient high-Rayleigh-number thermal convection with large viscosity variations
J. Fluid Mech.
Inversion of two-dimensional numerical convection experiments for a fluid with strongly temperature-dependent viscosity
Geophys. J. Int.
Thermal convection in the outer shell of large icy satellites
J. Geophys. Res.
Cited by (237)
The COSPAR planetary protection policy for missions to Icy Worlds: A review of history, current scientific knowledge, and future directions
2024, Life Sciences in Space ResearchChasing nomadic worlds: A new class of deep space missions
2023, Acta AstronauticaOnset of convection in internally heated, temperature-dependent, power-law viscosity fluids at large viscosity contrasts
2023, Physics of the Earth and Planetary InteriorsJupiter’s ocean worlds: Dynamic ices and the search for life
2023, Ices in the Solar-System: A Volatile-Driven Journey from the Inner Solar System to its Far ReachesFrom science questions to Solar System exploration
2023, Planetary Exploration Horizon 2061: A Long-Term Perspective for Planetary Exploration