Large-scale dynamics and the fate of salts in Europa's icy shell

During the past decade the large icy moons of Saturn and Jupiter have become one of the prime targets in the search for extraterrestrial life (Coustenis et al, 2013). The likely existence of subsurface oceans in those moons (e.g., Soderlund et al., 2020) provide a key ingredient supportive to life as we know it, which is liquid water (Chyba & Phillips, 2001). The Jovian moon Europa is of particular interest when it comes to the search for habitable environments. Europa has been and still is experiencing significant tidal heating due to the gravitational pull of Jupiter (Sotin et al., 2009) and its ice-water layer most likely contains a non-negligible amount of salt (Carlson et al., 1992; Kargel et al., 2000; Zolotov et al. 2009; Trumbo et al., 2019). Both factors slow down freezing of liquids, which might help to keep a present-day liquid salty ocean as well as liquid reservoirs in the icy shell forming the outermost layer of Europa (e.g., Sotin et al., 2002; Tobie et al., 2003; Hussmann et al., 2006; Kalousova et al., 2014).

Depending on the conditions present, liquid reservoirs in Europa’s ice shell could either form somewhere within the ice shell due to melting or be a relic of the process of freezing at the ice-ocean interface (mushy layer). Possible convection of Europa’s ice shell might allow for transporting liquid inclusions to shallow subsurface regions. If liquid inclusions exceed a certain volume fraction and thus become interconnected they will not only be passively advected by the solid but simultaneously percolate (two-phase porous flow). Potential shallow subsurface liquid reservoirs on Europa provide a unique chance for probing these regions in the context of conceivable future missions (Dachwald et al., 2016). Investigating the emergence and the subsequent transport of liquid reservoirs in the ice shell will provide insights into their current location. It is thus not only important for understanding the general habitability of Europa but it is also essential for designing missions for probing the subsurface.

In the present study, we want to set the stage for the concurrent modeling of solids and liquids in Europa’s ice shell in order to investigate the evolution of potential brine inclusions. As a first step, we conduct a parameter study for a completely solid ice shell with salt intrusions. The goal is to investigate the influence of different parameters on the transport of salt intrusions in a convecting ice shell. This study will serve as a basis for later, more complex models incorporating liquid intrusions.

We study ice layers that are pure water ice as well as a mixture of water ice and salt. For the pure water ice case, we investigate the effects of viscosity and thermal conductivity on the ability of the ice shell to convect. To this end, we use the mantle convection code GAIA (Hüttig et al., 2013) and perform simulations in a spherical annulus geometry (Hernlund & Tackley, 2008; Fleury et al., 2020). 

We test realistic material parameters such as a composite rheology (Goldsby et al., 2001) and compare the results to pure diffusion creep rheology. We also compare simulations that account for a temperature-dependent thermal conductivity (Hobbs et al., 2010; Petrenko et al., 1999) with results obtained for a constant thermal conductivity.. Fig. 1 shows the convection pattern in a 60 km ice shell employing a diffusion creep rheology with a reference viscosity of 1014 Pa s for a pure water ice scenario. Fig. 1a uses a variable thermal conductivity following the parameterization of (Petrenko et al., 1999) and Fig. 1b shows a case with a constant conductivity of 2.44 W/m/K. In the variable thermal conductivity case, convection is more sluggish and leads to a thicker immobile layer at the top of the convecting domain (the so-called stagnant lid) compared to the constant conductivity case.

Figure 1: Two snapshots at 48.3 Myr for a 60km thick convecting ice shell: a) constant thermal conductivity and b) temperature dependent thermal conductivity.

Furthermore, we systematically vary the thickness of the ice shell between 10 and 120 km, which influences the existence and - if present - the vigor of convection. Generally, cases using a variable thermal conductivity show less vigorous convection compared to the constant conductivity cases. This is illustrated by the Nusselt-Rayleigh diagram in Fig. 2. For a variable thermal conductivity, solid-state convection occurs for ice shell thicknesses larger than 34 km, while for constant conductivity scenarios, convection occurs already for an ice shell thickness of 27 km.

Figure 2: Nusselt number as a function of Rayleigh number for constant (filled circles) versus temperature dependent (empty circles) conductivity. Dashed lines indicate onset of convection.

Finally, considering the more realistic material parameters mentioned above, we add compositional heterogeneities (e.g., salt intrusions) considering different initial distributions and study their redistribution via convective transport in ice shells of different thicknesses. Our parameter study will show under which conditions salt intrusions can exist near the surface, can remain as compositionally distinct regions, or are evenly mixed within the ice shell.

References:

A. Coustenis et al., Life beyond Earth: The search for habitable worlds in the universe (Cambridge University Press, 2013).
M. Soderlund et al., Space Science Reviews 216, 1 (2020).
F. Chyba and C. B. Phillips, Proceedings of the National Academy of Sciences 98, 801 (2001).
C. Sotin et al., Europa, (2009).
R. Carlson et al., The Galileo Mission (Springer, 1992).
S. Kargel et al., Icarus 148, 226 (2000).
Y. Zolotov and J. S. Kargel, in Europa, (2009).
K. Trumbo et al., Science advances 5,(2019).
C. Sotin et al., GRL 29, 74 (2002).
G. Tobie et al., JGR: Planets 108 (2003).
H. Hussmann et al., Icarus 185, 258 (2006).
K. Kalousova et al., JGR: Planets 119, 532 (2014).
B. Dachwald et al., in Proceeding of the conference Enceladus and the Icy Moons of Saturn (2016).
C. Hüttig et al., PEPI 220, 11 (2013).
W. Hernlund and P. J. Tackley, PEPI 171, 48 (2008).