Formation of Water-rich Giant Planet Satellites at Decretion Disk Ice Lines

Abstract

The volatile budgets of giant planet satellites are critical to unraveling the origin of their building blocks within the circumplanetary disks that hosted them. The Galilean moons Ganymede and Callisto, as well as the Saturnian moon Titan, are known to be anomalously water rich on the basis of their mean densities and interior models informed by gravity data from Galileo and Cassini, characterized by ice-to-rock ratios around unity. Here, we show that the water-ice sublimation line in a decreting circumplanetary disk lends itself to the formation of a water-rich solid reservoir, serving as a natural site for the birthplace of icy satellites. Fundamentally, this reflects how interior to the ice line, water vapor is advected outward, while beyond it, water ice drifts inward as pebbles. Using a semi-analytic model for dust and vapor evolution, we simulate vapor and ice accumulation at the ice line, showing that solids just beyond it achieve steady-state ice-to-rock ratios a factor of a few higher than elsewhere in the disk. For typical disk parameters, this ice buildup occurs within a timescale of a few thousand years. We propose this as a first-order process that explains, at least to some extent, the compositions of three aforementioned satellites. We explore the impact of uncertain turbulence parameters on our results, namely the turbulent Schmidt number and Shakura-Sunyaev alpha, before discussing them in the context of icy satellite D/H ratios. We conclude by evaluating alternative scenarios for explaining water-rich satellites, based on the conversion of CO to CH4, with water as a by-product.

Figures (9)

• Figure 1: Gas surface density $\Sigma_g(r)$ (black) and mass decretion rate $\dot{M}(r)$ (red) profiles in our circumplanetary (Jovian) disk model ($\alpha = 10^{-4}$). The disk inner edge at $\sim 5$ times Jupiter's (primordial) radius $R_{J,pr} \simeq 2 R_J \simeq 1.4\times 10^8$m (Batygin & Adams, 2025) is set by magnetospheric truncation. The disk outer edge at $\sim0.4 R_{Hill}$ is set by tidal truncation (Martin & Lubow, 2011) from perturbations by the (proto-)Sun. The ice-line at $T(r_i)=170K$ is situated at $r_i\simeq 0.23 R_{Hill}$, where $\Sigma_g(r_i)\simeq 2.2\times 10^4$ kg/m$^2$ and $\dot{M}\simeq 0.25M_J/Myr$. See text in Section 2.1 for details.
• Figure 2: Qualitative sketch of water-ice accumulation at the decretion disk ice-line.(a) Small and tightly coupled rocky particles, along with water vapor, are advected across the ice-line with disk gas. (b) Beyond the ice-line, water-ice condenses onto the rocky particles, forming larger icy particles that drift back across the ice-line. (c) Over many iterations of (interchangeable) steps (a) and (b), the rocky particles grow among themselves, such that they eventually drift back to the giant planet. The water that originally accompanied these particles remains at the ice-line, as released vapor is continually advected (and diffused) outwards. See text in Section 2.2.
• Figure 3: Schematic of "two-bin" model developed to simulate dust evolution (i.e., ice buildup) at the decretion disk ice-line, including key model parameters. Dust and ice transport is driven by advection with the surrounding, decreting gas, and drift resulting from headwind drag. Vapor transport occurs by advection and diffusion. See Section 2.3 for an in-depth description.
• Figure 4: Evolution of the (a) water vapor mass fraction $f_{vap}$ interior to the ice-line, and (b) ice-to-rock ratio $R_{ice}$ beyond it, for different ice mass fractions of inward drifting icy pebbles from the outer disk $f_{ice,out}$. As is apparent, both $f_{vap}$ and $R_{ice}$ grow rapidly on a timescale of a few hundred years, reaching their steady-state values after $\sim 10^3$ years. Moreover, a range of $f_{ice,out}$ roughly between $\sim 0.25$ and $0.4$ ($R_{ice,out} = f_{ice,out}/(1-f_{ice,out})$ between $\sim 0.1$ and unity) yields final $R_{ice}$ values concordant with those estimated for the water-rich giant planet satellites. See text in Section 3.1.
• Figure 5: The final (i.e., steady-state) ice-to-rock ratio $R_{ice}$ as a function of the mass fraction of advected water vapor from the inner disk $f_{vap,in}$, for various $R_{ice,out}$ of icy pebbles drifting inward for the outer disk. The vertical dashed line at $f_{vap,in}$ corresponds to the simulation results displayed in Fig. 4. Clearly, the primary control on the final $R_{ice}$ of the dust reservoir beyond the ice-line is $R_{ice,out}$, not $f_{vap,in}$, indicating most of the water delivered comes from the outer disk in the form of pebbles. See text in Section 3.1.