Could Jupiter's moon Ganymede host life? Layer-cake model advances the idea.

A new modeling study of the interior of the Jupiter moon Ganymede reveals four layers of ice and oceans, starting with a water layer – not ice – at the seafloor.

Jupiter's moon Ganymede has long been thought to host an ocean beneath its icy surface – but four oceans?

They lurk in layer-cake fashion, separated by layers of increasingly dense ice, according to a new modeling study of the moon's interior conducted by researchers at NASA's Jet Propulsion Laboratory, the agency's Ames Research Center, and École Normale Supérieure de Lyon in France.

If this picture of the moon holds up in the face of measurements at the moon itself – at such time as those measurements can be made – it would add Ganymede to the list of icy objects in the solar system potentially capable of hosting at least simple forms of organic life. The others are Europa and Callisto, two other moons of Jupiter, and Saturn's moons Titan and Enceladus.

Until now, Ganymede was thought to have a single under-ice ocean. Unlike Europa's ocean, Ganymede's wasn't considered a potential habitat.

"The story at Europa is that you can have water in direct contact with rock" at the seafloor, says Steven Vance, the lead author of the paper describing the results and accepted for publication in the journal Planetary and Space Science.

If the seafloor has hydrothermal activity, as Earth's ocean floors do, "you can have life," Dr. Vance explains.

"The story with Ganymede has been: You can't have life on Ganymede because the pressures are so high. And even if life could live at such high pressure, you don't have a lot of water-rock interactions because you have these high-pressure ices that blanket the seafloor," he says.

High pressure indeed. On Ganymede, pressures at the seafloor would top 12,000 atmospheres. Key biological processes, such as the folding of proteins, shut down at pressures of about 8,000 Earth atmospheres, researchers have found.

That still could be a problem, but is the bottom really coated with a layer of ice? The team's modeling suggested that this might not be the case – just add sufficient amounts of salt. In this study's case, the salt is dissolved magnesium sulfate, which would be readily available through hydrothermal activity at the water-rock interface, if the activity existed.

Salt not only lowers the freezing point of water. It also makes the water denser than fresh water. Experiments have shown that under extreme pressures, this process goes into overdrive. Water at such pressures can freeze, even though it's at hot-tub temperatures. The resulting ice is so dense that it remains at depth, rather than rising to the surface, as ice cubes in a glass of water do.

Vance and colleagues plugged the physics behind this into their model. The model also included pertinent information, such as Ganymede's gravity field and temperatures from the surface to the rocky mantle, which is heated by a combination of radioactive decay in elements in the mantle and tidal heating. These are based on measurements taken by NASA's Galileo spacecraft, which routed the Jovian system for nearly eight years.

The results yield an overall depth of the layered ice and oceans as some 900 kilometers (558 miles), with four layers of ice and oceans, starting with a water layer at the seafloor.

The model presents a snapshot of the interior based on today's conditions. It doesn't attempt to explain how the layering occurred – a task that would require additional refinements.

But conceptually, Vance says, such layering could occur as a once-mini-water-world cools and the deepest, densest layers of ice form. Perhaps a billion years later, heat from radioactive decay would work its way from the molten core and through the rocky mantle to reach the lowest, densest ice layer.

This lowest layer would melt from underneath forming a liquid layer that, in the presence of salts, would resist refreezing. In addition, any salts in the original ocean and captured in the ice would be released into this fledgling sea.

At first, the ice layer is thick enough to act as an insulator. This traps heat from the crust and, through additional melting, deepens the liquid layer.

At some point, however, the ice would grow thin enough to allow some heat to escape, reaching a kind of equilibrium with the ocean below that stabilizes the ice's thickness. The process also would have allowed the new ocean to become salty enough, hence dense enough, to float the ice that gave it birth.

That process would repeat itself at lower temperatures and ice densities up to the moon's ice surface crust.

Intriguingly, the model also generates an odd icy "snow" in the ocean nearest the surface – in which the snow rises, rather than falls. The snow gives parts of this ocean a slushy consistency.

The model has elements that still keep Vance awake at night.

For instance, it's not clear how heat in the water underlying each ice layer would rise through the ice in ways that allowed the layer to reach a relatively stable thickness. Earth's crust transfers such heat through a process known as solid-state convection, which gives the planet its plate tectonics. This also happens, researchers have posited, on the ice covering Europa, where a block of relatively warm, more buoyant ice rises from below to poke through the surface, releasing heat in the process.

Still, the results are intriguing enough to push the team to explore the effects of possible salts other than magnesium sulfate on the modeled system and to work out the evolution of layered oceans and ice over Ganymede's history, and how long such a layer-cake arrangement could exist.

Some help could come from the European Space Agency's Jupiter Icy Moons Explorer mission, currently envisioned for a 2022 launch, with arrival at Jupiter in the 2030s. The mission not only could help identify sites where later missions could deposit seismographs for probing the moon's interior; it also will return data that could enable researchers to determine how the ocean's density changes with depth, which could provide additional evidence as to whether the layer-cake model is plausible.