Ganymede gobal natural colour image (Source: NASA)

Ganymede – ice giant

Ganymede is next out from Jupiter after Europa and it’s an icy behemoth: larger than any other moon in the Solar System, larger even than Mercury.  ESA are gearing up to send JUICE (the Jupiter Icy Moon Explorer) out to the Jovian system and it will concentrate on Ganymede.  So what’s so exciting about this moon that it should be a bigger draw on our attention than Europa?

The light and the dark

Zooming in on Uruk Sulcus, an area of grooved terrain in a bright region of the surface of Ganymede (Source: NASA/JPL)

The surface of Ganymede is pretty fascinating to look at: it’s made up of huge dark fragments surrounded by bright, less cratered areas, like the breakup of a dirty icepack.  But when you look closer, it gets more complex: those lighter areas are a mass of parallel grooves, with different areas having grooves at different orientations, as if even that has broken up.

The general theory is that the dark areas are the original, older, surface and that the bright areas formed about 2 billion years ago during a period when the whole moon expanded.  The grooves show that slow expansion and can be compared to rifts on Earth where parallel extensional faults form as the surface is pulled apart.  As on Europa, some have said that cryovolcanism occurred at these fractures1, but it may be that they’re purely tectonic2.

The cause of the breakup

Our current best-guess at the internal structure of Ganymede (Source: NASA/JPL)

There are several theories as to why Ganymede expanded.  To understand them, we need to look inside the moon.  Its present structure is probably as in the diagram on the right: an iron-sulphur core, some or all of it probably molten, a silicate outer core and an icy mantle within which there is a thin liquid water ocean, about 200km below the surface.  The ice crust is normal ice (ice I), but deeper down there are layers of denser types of ice (polymorphs) due to the great pressure at those depths.  This structure explains several observations:

  • Ganymede has a magnetic field: it’s generally thought that this is generated in a conducting fluid within planetary bodies, so a liquid core is likely.
  • Its density shows it’s 50/50 ice and rock, so it has to have a huge depth of ice, probably 800-1000km thick.
  • There’s less difference to the cratering on the Jupiter side and the non-Jupiter side than you’d expect. That suggests there’s been some non-synchonous rotation of the surface as on Europa, and that in turn suggests there’s a liquid ocean where the surface and the depths decouple3.
The liquid core and ocean present a problem, though.  Ganymede has a less eccentric orbit than Europa and Io, so it doesn’t receive much tidal heating.  Callisto, the next moon out, is a similar size and composition, but is much more homogeneous and froze solid long ago.

A popular theory is that, in the process of getting to their present orbital resonance with each other, Ganymede, Io and Europa have passed through different resonances. One or more of these may have caused more eccentricity in the orbit of Ganymede and thus more tidal heating.  Depending on the state of Ganymede at that time, it could have led to expansion in two main ways:

1. Through delayed internal differentiation: the moon may initially have been as undifferentiated as Callisto, but then heating during a period of resonance caused it to melt internally, forming the liquid core. Differentiation causes expansion, so the bight grooved terrain would have formed at that time. This explains why it still has a molten core – the core’s only about 2 billion years old so it hasn’t had time to cool and freeze.

2. Through melting of its icy mantle: if Ganymede was already differentiated and had an icy mantle, the tidal heating from resonance could have melted that, causing expansion. Though melting of normal ice would actually lead to contraction because it has a lower density than water, modelling suggests that it would be mostly the deeper, denser polymorphs that would melt. The resulting water is less dense and so takes up more volume, and again this would lead to the extension that made the grooved terrain.

Modelling has suggested that either one of these processes wouldn’t cause enough expansion to explain the amount of extension seen in the bright terrains, so it could be that both played a role4.

Another cradle of life?

An artist's concept of JUICE within Ganymede's magnetic field. (Source: ESA)

As on Europa, the heating of this ocean brings up the possibility of life.  Eruptions from the deep silicate layer would give out nutrients in the same way as the black smokers mentioned in my last post, and those would work their way up through the heavy ice to the liquid layer5.  A thick ocean receiving heat and nutrients from below? Not a bad starting point for the development of a biosphere.

Whether that life could have survived to the present is more uncertain.  Certainly there’s still a liquid water layer and measurements by the Galileo mission suggest it’s salt water.  But under the present conditions of less heating, are there still silicate eruptions replenishing the nutrients to it?  Also, can anything from below still get up to the liquid layer, now that the heavy ice layers are much thicker?

These are the kinds of questions JUICE will aim to look at, but I’m as yet unconvinced that this makes Ganymede a more interesting goal than Europa.  On the one hand, we have this huge moon which may well have had life in the distant past, but is now much less friendly to its survival.  On the other, there’s Europa, with ongoing tidal heating and possibly a very thin crust which we may be able to penetrate. Ganymede’s liquid layer is about 200km down, not exactly accessible.

As with every voyage of discovery, though, we don’t know what we’ll find until we find it, so I may eat my words.  It’ll be a long wait, though – JUICE, assuming its funding holds up, is due to go into orbit around Ganymede in 2032!


  1. Shenk et al. (2001) Nature: 410, 57-60 []
  2. Head et al. (2002) Geo. Res. Let.: 29, No.24, 2151. []
  3. Zahnle et al. (2003) Icarus: 163,263-289 []
  4. Bland et al. (2009) Icarus: 200, 207-221 []
  5. Barr et al. (2001) Lunar and Planetary Science XXXII []