Help Shape the Future of Space Exploration

Join The Planetary Society Now  arrow.png

Join our eNewsletter for updates & action alerts

    Please leave this field empty

Headshot of Emily Lakdawalla

An internal ocean on Ganymede: Hooray for consistency with previous results!

Posted by Emily Lakdawalla

12-03-2015 19:25 CDT

Topics: Jupiter's moons, Hubble Space Telescope, Ganymede, explaining science

NASA held a press briefing today to discuss a newly published paper that confirms a subsurface ocean at Ganymede. The idea that Ganymede has a global layer of liquid water is not a new one. Margaret Kivelson and coworkers described evidence for it from Galileo magnetometer data in 2002. Geophysically speaking, it's not hard to make an internal ocean at Ganymede, either. Ganymede is a very large world -- it's the biggest moon in the solar system, larger even than Mercury, and it gets a boost to its internal heat production from tidal forces from Jupiter and the other moons. Still, it's very cool to have an independent way of confirming that Ganymede does have an internal ocean, and that's what was published today, in an article by Joachim Saur and 13 coauthors. They used the Hubble Space Telescope's ability to see in the ultraviolet to watch how Ganymede's aurorae shift as Jupiter's tilted magnetic field rotates with each Jupiter day. It's wonderful that we can use an Earth-based telescope to look at the aurorae of a giant planet moon to deduce the presence of an ocean that we can't see.

Other writers have explained the work in more detail (try Deborah Netburn at the L.A. Times, or Jonathan Amos at the BBC, or just go to the source and read the writeup at the Space Telescope Science Institute website), so I won't go into that here. What I am interested in is: what does Saur's paper have to say about the internal structure of Ganymede? I've looked into it, and I have good news: this latest result is entirely consistent with past work on the internal structure of Ganymede. In fact, I don't even have to change this chart that Doug Ellison and I made back in 2010 comparing Ganymede to other ice-rich worlds of the solar system that could also contain oceans. (That being said, I should have updated it in 2011 to reduce Eris' size to match Pluto's -- I promise to do that after the New Horizons flyby when we know exactly how big Pluto is.)

Interiors of icy bodies in the solar system (as of 2010)

Doug Ellison, Emily Lakdawalla, and Bob Pappalardo. Internal models based mostly on Hauke Hussmann, Frank Sohl, and Tilman Spohn, 2006, except for Titan (Giuseppi Mitri et al., 2010) and the Galilean moons (Bob Pappalardo, personal communication).

Interiors of icy bodies in the solar system (as of 2010)
Many bodies in our solar system may contain oceans. Jupiter's ice-coated moons (Europa, Ganymede, and Callisto) probably contain internal saltwater oceans. The more distant large and medium-sized icy moons, the icy dwarf planets, and tiny Enceladus may contain colder ammonia-water oceans. Oceans in icy satellites could be a common feature throughout the universe, while Earth's surface ocean may be the more unusual case. These interior models were developed by many geophysicists from data about the sizes, densities, and orbital characteristics of each body.

The Hubble results don't precisely pin down the size and depth of an ocean at Ganymede, but the range of possible ocean thicknesses and depths is small. The observed behavior of the aurorae at Ganymede are consistent with a liquid ocean that is a relatively small fraction of Ganymede's radius. Of course, Ganymede is a very big world, so "relatively small" in this case can mean an ocean up to about 100 kilometers deep. According to the paper, some possible oceans that are consistent with the results include, but are not limited to:

  • A liquid layer from 150 to 250 kilometers depth with relatively low salinity
  • A liquid layer from 190 to 210 kilometers depth with relatively high salinity
  • A liquid layer below 330 kilometers depth with relatively high salinity

Water in one form or another at Ganymede goes from the surface to a depth of about 720 kilometers, at which point you reach Ganymede's rocky outer core. So I think we can generalize these all to say: there is an ocean of less than 100 kilometers thickness located within a few hundred kilometers of the surface of Ganymede. In all possible cases, the liquid layer would be perched within the solid icy mantle, with solid ice above it of the form we have on Earth, and ice below it in a high-pressure crystalline form. (For a lot more on the funky layered forms that liquid and solid water could take at Ganymede, read my writeup of Steve Vance's work on Ganymede's internal layering, which he presented at the Lunar and Planetary Science conference two years ago.)

That is a lot of liquid water, but it's in a very different place within Ganymede than the liquid water that we think exists at Europa and Enceladus. Europa has a much higher proportion of rock to ice than Ganymede does, and is also warmer because of greater tidal friction; the same physics that predicted Ganymede's perched-ocean-within-an-icy-mantle predicts that Europa's liquid water ocean is in direct contact with its warm rocky core. Warm liquid water percolating among warm rocks is, by definition, hydrothermal activity. Hydrothermal zones are places that exobiologists imagine life might happen, because you have lots of energy and you have rich chemistry created in all that warm liquid water eating away at rocks. Earlier this week, one of the Cassini instrument teams announced that they had detected rock particles from just such an environment that originated within Enceladus.

This ocean at Ganymede is not the same. It is contained entirely within an icy mantle. And it is very, very, very far from warm rocks; Ganymede's rock is at least 400 kilometers deeper than its ocean.

But if you don't insist that everything have to do with life to be cool, then this is a very cool result. It's reassuring to know that relatively straightforward physics -- differential equations describing how heat flows through a layered planet -- can fairly accurately predict the structure of a planetary interior. Look back at that chart at the top of this page and think about how many oceans there may be, all over the solar system. Maybe there's one at Pluto. Now wouldn't that be cool?

Three simplified models for Pluto’s interior

Barr and Collins 2015

Three simplified models for Pluto’s interior
A) Solid, icy crust, a liquid ocean, and a rocky core. B) Solid, icy crust directly above a rocky core. C) Uniform-density interior comprised of a homogenous ice/rock mixture.

See other posts from March 2015


Or read more blog entries about: Jupiter's moons, Hubble Space Telescope, Ganymede, explaining science


joel : 03/13/2015 08:17 CDT

Good news indeed especially for ESA's JUICE mission that will launch in 2022 and orbit Ganymede. If this could validate thermal models predictions that foresee such liquid layers for Callisto Ceres Titan Triton and for icy bodies with density abobe 1.5 and diameters above 1000km that would be great! Callisto would be good to add to the list because this would mean radiogeny alone is enough to produce liquid water layer so without depending upon tidal energy. So concerning a much larger number of objects in the universe. Even better on smaller objects lile Pluto or Ceres because they are more numerous and above all have less gravity so lower pressures hence no hot ice layer insulating the ocean from the rocky core. It would be interesting too to see what thermal evolution models predict for a big fast rotating prolate object like Haumea: how are layers structured on a 1000x1500x2000km non spherical object?

Paul_Wi11iams: 03/13/2015 12:46 CDT

As usual, I enjoyed your article. But am I misunderstanding, or were those three cutaways of Pluto about the liquid state that may have existed in the past during an ancient tidal evolution that got Charon to where it is today ? I'll try to wade through the paper you referred to by Amy/Barr/Geoffrey/Collins. A presently existing liquid ocean would be exciting for many, and the slow radioactive decay necessary for it, may also be needed to explain the behaviour or Ceres. so if that kind of heating were to be happening everywhere in the solar system, why not !

Paul_Wi11iams: 03/13/2015 01:27 CDT

Skipping down to section 5 the end of the pdf, there is a sort of timid mention of the possibility of an existing liquid ocean on/in Pluto. ============= [quote] The New Horizons mission has sparked new interest about the thermal evolution of Pluto and its moon Charon, in particular, the possibility of tectonic resurfacing on both bodies and the likelihood that these bodies may have had liquid water oceans, either now or in the past [/quote]

Torbjörn Larsson: 03/13/2015 02:52 CDT

I remember that image! It made me aware of the biosphere potential of ice moons at the time. Cool as Ganymede is, I'm much more excited by the implications if Enceladus ocean has contact with the rocks. And the silica reference speaks of consistency too, since the simultaneous methane explanation may be vents as well. And especially these methane (and silica) producing alkaline vents are suggested as nurseries for life. ["The Drive to Life on Wet and Icy Worlds", Russell et al, aAstrobiology.] The Enceladus ocean would be old enough, at least 0.5 Myrs IIRC. The Enceladus observations may mean that the image text may need adjustment as well: "The more distant large and medium-sized icy moons, the icy dwarf planets, and tiny Enceladus may contain colder ammonia-water oceans." If the methane source is serpentinization, driving the alkaline vents, you have a highly exothermic reaction adding to the tidal forces maintaining the liquid state. The silica production by itself puts a lower temp of 90 degC in the vents. So the need for ammonia may go away. (Did Cassini find any ammonia in the E-ring/plumes? I can't remember. But evidently the new results suggest a low salinity ocean FWIW.)

Torbjörn Larsson: 03/13/2015 03:51 CDT

Ah, no, unless I am mistaken, ammonia would be confused with water in a low resolution masspec (16-18 Da for dissociated ammonia vs 18 Da for water; and see the Cassini mass spectrograms below). I note in passing that the vent find of direct communication with the ocean and its methane could move the early Enceladus plume mass spectrum away from "comet-like chemistry" to "how was the organics produced". [ ; ; ]

Guest123: 03/13/2015 09:26 CDT

Paul McCarthy: 03/13/2015 11:28 CDT

The Late Proterozoic "Snowball Earth" would have fitted the graphic perfectly.

visenna: 03/14/2015 06:23 CDT

This is a nice comparison figure, it would be cool to see one for all the layed bodies (also other than icy) in Solar System.

Leave a Comment:

You must be logged in to submit a comment. Log in now.
Facebook Twitter Email RSS AddThis

Blog Search

Planetary Defense

An asteroid or comet headed for Earth is the only large-scale natural disaster we can prevent. Working together to fund our Shoemaker NEO Grants for astronomers, we can help save the world.


Featured Images

Jupiter from Juno at Perijove #4
Jupiter in approximate true color during Juno perijove 4
More Images

Featured Video

Class 9: Saturn, Uranus, and Neptune

Watch Now

Space in Images

Pretty pictures and
awe-inspiring science.

See More

Join The Planetary Society

Let’s explore the cosmos together!

Become a Member

Connect With Us

Facebook, Twitter, YouTube and more…
Continue the conversation with our online community!