≡ Menu

Evidence Mounts for Ganymede’s Ocean

Yesterday’s discussion of hydrothermal activity inside Saturn’s moon Enceladus reminds us how much we can learn about what is inside an object by studying what is outside it. In Enceladus’ case, Cassini’s detection of tiny rock particles rich in silicon as the spacecraft arrived in the Saturnian system led to an investigation of how these grains were being produced inside Enceladus through interactions between water and minerals. If correctly interpreted, these data point to the first active hydrothermal system ever found beyond Earth.

Now Ganymede swings into the spotlight, with work that is just as interesting. Joachim Saur and colleagues at the University of Cologne drew their data not from a spacecraft on the scene but from the Hubble Space Telescope, using Ganymede’s own auroral activity as the investigative tool. Their work gives much greater credence to something that has been suspected since the 1970s: An ocean deep within the frozen crust of the moon.


Image: NASA’s Hubble Space Telescope observed a pair of auroral belts encircling the Jovian moon Ganymede. The belts were observed in ultraviolet light by the Space Telescope Imaging Spectrograph and are colored blue in this illustration. They are overlaid on a visible-light image of Ganymede taken by NASA’s Galileo orbiter. The locations of the glowing aurorae are determined by the moon’s magnetic field, and therefore provide a probe of the moon’s interior, where the magnetic field is generated. The amount of rocking of the magnetic field, caused by its interaction with Jupiter’s own immense magnetosphere, provides evidence that the moon has a subsurface ocean of saline water. Credit: NASA, ESA, and J. Saur (University of Cologne, Germany). Ganymede Globe Credit: NASA, JPL, and the Galileo Project

The early work on a Ganymede ocean grew out of computer models of the interior, but the Galileo spacecraft was able to measure the moon’s magnetic field in 2002, offering enough evidence for an ocean to keep the idea in play. The problem was that the Galileo measurements were too brief to produce an overview of the field’s long-term cyclical activity.

It was Saur’s idea to look at the idea afresh. Given that Ganymede is deeply embedded in Jupiter’s magnetic field, the aurorae that are produced in its polar regions are going to be influenced by any changes to that field, changes that produce a ‘rocking’ movement in the aurorae. These movements, Saur reasoned, would be a useful marker, one that, like the silica grains near Enceladus, could tell a story about activity deep below the surface. Says Saur:

“I was always brainstorming how we could use a telescope in other ways. Is there a way you could use a telescope to look inside a planetary body? Then I thought, the aurorae! Because aurorae are controlled by the magnetic field, if you observe the aurorae in an appropriate way, you learn something about the magnetic field. If you know the magnetic field, then you know something about the moon’s interior.”

The ‘rocking’ of the aurorae on Ganymede depends upon what’s inside the moon, and by the researchers’ calculations, a saltwater ocean would create a secondary magnetic field that would act against Jupiter’s field, tamping down the motion of the aurorae. The Hubble data show us that this is happening, for Saur’s models indicate the auroral activity is reduced to 2 degrees as opposed to the 6 we would expect if an ocean were not present. Ganymede thus joins Europa and Enceladus as an outer planet moon with increasing evidence for an ocean.

Does the likelihood of an ocean now mean we’ll shift more resources toward Ganymede as a possible venue for life? Remember that the European Space Agency’s JUICE mission (Jupiter Icy Moons Explorer) is still on track for a possible 2022 launch. Current planning calls for flybys of Callisto and Europa followed by an extended period of orbital operations around Ganymede. The mission would reach Jupiter in 2030, if these plans come to fruition.

For all its interest, though, Ganymede’s ocean seems less accessible than Europa’s, as it’s evidently sheathed in a crust of rock and ice that is 150 kilometers thick. Beneath that crust is an ocean scientists believe to be as much as 100 kilometers deep. Ganymede is a world that may well hold more water than all the water on the surface of our planet. But even if we could get to it, it’s also an ocean still thought to be trapped between two layers of ice, meaning the interesting interactions with the rocky core (as at Enceladus) would not be occurring.


Image: This is an illustration of the interior of Jupiter’s largest moon, Ganymede. It is based on theoretical models, in-situ observations by NASA’s Galileo orbiter, and Hubble Space Telescope observations of the moon’s aurorae, which allows for a probe of the moon’s interior. The cake-layering of the moon shows that ices and a saline ocean dominate the outer layers. A denser rock mantle lies deeper in the moon, and finally an iron core beneath that. Credit: NASA, ESA, and A. Feild (STScI).

What excites us about Enceladus is the prospect that hydrothermal vents at the ocean floor could be producing an environment with enough sources of energy and nutrients to make life possible. We know that the dark seafloor vents on our own planet are now considered a serious candidate for the place where life originated. If our current models are correct, Ganymede would lack the ability to develop this kind of ecosystem, but we still have much to learn about all these icy moon environments and the oceans they apparently conceal.

And as a draft of the paper on this work notes, whether Ganymede is life-bearing or not is of less consequence than the method employed here, which has implications not only within our own Solar System but elsewhere:

The method introduced here can also be applied to other planetary moons and planets to study their electrical conductivity structure. If these bodies are exposed to time-variable magnetic fields and exhibit auroral emissions, then their auroral patterns will be modified by any electrically conductive layers. Observations of the auroral emission responses combined with appropriate models for the responses will provide valuable information about these conductive layers, such as subsurface oceans. The method might one day even be applicable to exoplanets (and exomoons), once appropriate objects and associated magnetic fields are observationally confirmed.

The paper is Saur et al., “The Search for a Subsurface Ocean in Ganymede with Hubble Space Telescope Observations of its Auroral Ovals,” accepted at the Journal of Geophysical Research. Full publication information and links as soon as I have them. Meanwhile, this news release is helpful.


Comments on this entry are closed.

  • Ashley Baldwin March 13, 2015, 16:45

    I’m no cynic but all this exciting icy moon publicity is making a great case for the flagship Europa Clipper. Added to JUICE that’s Jupiter covered . What about Saturn ? One of the good things about “lesser” NASA missions is the facility to use part of any savings made by using a cheaper than prescribed launch vehicle to top up the funding cap. For a $0.5 million Discovery mission this can be up to $16 million. For $1 billion New Frontiers it will be much more. The next New Frontiers round, No. 4 ,starts this year for a 2021 launch. There are five bidders including a Saturn atmosphere probe and carrier/relay spacecraft. Everyone knows Falcon Heavy will be available soon and that it can deliver quick, direct flights to outer solar system destinations. It is also less than half the cost of the conventional Delta IV or Atlas V vehicles that would normally be used. If I was the Lead for the Saturn probe I would be doing a feasibility study of adding an Enceladus lander ,funded with the savings from using a Falcon. Battery powered for a few days compared to just an hour or so for Huygens it would be easily based on the Philae lander from Rosetta ,further saving cost , given the similar operating environment. A whole lot of extra science for little more money and in short order too. From an uncertain future the 20s could yet be reminiscent of the halcyon interplanetary days of Viking and Voyager. Meanwhile the October Enceladus Cassini flyby comes within just 30 miles of the moon and will pass straight through its plumes . What with Dawn at Ceres and New Horizons at Pluto too it’s going to be the year of the icy body.

  • Ashley Baldwin March 13, 2015, 19:11

    If Ganymede was moved with Jupiter to just one AU from the sun its ice would melt and escape to space as volatiles because it doesn’t have enough mass and gravity to hold onto them . Rene Heller’s team have shown that the largest ‘regular’ ( formed with its planet) exomoon possible, around a ten mass Jupiter gas giant , can be up to about 0.2 mass of the Earth or roughly twice that of Mars. That’s enough to hold onto an atmosphere . Just. As gas giants must form by definition beyond the ‘ice line’ in a star system ,their regular moons can be expected to be similar to Ganymede. If the big planet and exomoon migrated into the the habitable zone of the star , once again the moon’s ice would melt but might be retained as a moonwide Ocean. No tectonics , carbon cycle or secondary atmosphere though. So it may be that the best bet for a habitable exomoon is an ‘ irregular ‘ one , like Triton a stray body captured by a large planet. Unlike Triton, rather than an icy Kuiper Belt Object, it would need to be captured inside the ice line so it is more likely to be rocky and thaw out into a terrestrial and hopefully habitable body . Its parent planet should attract plenty of comets and asteroids ( hopefully not too often !) to provide the volatiles to start an atmosphere and more importantly lubricate tectonics. Unlike regular moons , irregular moons don’t have a clear upper limit in mass and could be as big as the Earth even. The only issue left then is that they orbit far enough out from the planet to avoid dangerous runaway greenhouse creating tidal heating ( think Io) and rotate quickly enough to create a protective magnetic field . They would need to be too close in to be protected by the planet’s own field .

  • Michael March 14, 2015, 12:43

    @Ashley Baldwin

    There are issues with large Jupiter planets having habitable moons, for one they rotate fast and can generated large magnetic fields that would sweep any moons in orbit at high velocity. If the upper atmospheres of these moons becomes ionised with say UV from their Sun or by particles that are trapped in the parent planets magnetic field they would erode a moons atmosphere quite rapidly. Second the orbital velocity of the moons and the gravitational pull of these massive planets would have asteroids and comets hitting the surfaces of these moons at quite high velocities making an atmosphere less likely to remain in place over long periods of time.

  • xcalibur March 14, 2015, 15:49

    I think that if we colonize the moons of the outer planets, a viable option would be to dig caves in the ice crusts. Ideally this would provide access to a surface spaceport and to the inner ocean, while the ice would be a natural radiation shield. Hopefully I’m on the right track.

  • Tom Mazanec March 14, 2015, 16:42

    Could Callisto have an ocean?

  • Larry Kennedy March 14, 2015, 19:00
  • djlactin March 15, 2015, 4:17

    Is it a layer of water (with solutes) between layers of other substances? Or is it a thick water-saturated porous layer of some other substance? (Something intermediate?) Could be distinguish between these tow extremes?

  • Ashley Baldwin March 15, 2015, 7:32

    I couldn’t agree more Michael. Trying to image a habitable moon is difficult . There are so many variables. I think the best bet is probably an “irregular” moon , i.e another body or planet , picked up gravitationally by a larger planet as it migrates into the habitable zone. Rene Heller’s many simulations have shown that moons the size of Mars or twice as big can form around large gas giants but as you point out they would be subject to the same bombardment as the gas giant itself, and with Shoemaker-Levy we have seen first hand examples of the effect of cometary bombardment. Smaller planets , Neptune sized ( the smaller the better really), have been modelled to gravitationally capture up to Earth sized bodies , because of the lower closing velocities involved because of their smaller mass , and they would be less likely to “draw”fire , but you are right to say that such moons would still be vulnerable to impactors attracted by the parent planet. Hopefully a larger irregular moon could sustain tectonics long enough to produce a secondary atmosphere once the early system bombardment has ended but it would remain more vulnerable than a conventional terrestrial planet . As with Jupiter in the solar system, the presence of a large gas giant somewhat further out would help mitigate this effect ( hopefully even more so!) ,by attracting or deflecting potentially dangerous debris, hopefully enough long enough to allow a biosphere to develop . An intelligent civilisation would obviously require much longer of course and indeed be vulnerable to any impact . Such is the problem with “habitable’ moons.

  • ljk March 16, 2015, 9:10

    Have we got Solar System Habitability Backwards?

    By Caleb A. Scharf | March 13, 2015

    Enceladus, Europa, Ganymede, Titan, Triton, Pluto, Eris…they may all have, or have had, large oceans of liquid water trapped beneath a frozen crust. That poses some interesting questions.

    I’ve written before on these pages (and elsewhere) about the wealth of evidence for internal bodies of liquid water in our solar system. Since the Pioneer, and then Voyager and Galileo missions, the icy shrouded moons of the giant planets have prompted the notion of layered interiors that could include oceans.

    Relatively simple physical modeling of the balance between self-gravity and multi-phase material pressure inside these objects suggests that internal ocean zones can happen. And there is a wealth of pieces of evidence to support this, which – taken altogether – point to an all but unassailable truth.

    Full article here:


    To quote:

    One of the hurdles for the broader community to get on board may be, in my opinion, that we’ve been so indoctrinated with the notion of Earth as a cosmic ‘oasis’ that it’s hard to see beyond this. Our planet is indeed a jewel, but it’s a jewel who’s worth is somewhat in the eye of the beholder. If we came from, say, Mars, we’d probably think that the thin atmosphere and exquisitely caustic dryness of martian soil were lovely, beautiful things (well, I guess they are, but not immediately for most of us).

    It really is possible that we have the habitability of our solar system backwards. While it’s true that the internal oceans of the outer solar system may mostly be places of low energy flux, with slow-living organisms more akin to the deep terrestrial biosphere, it’s also possible that hydrothermal environments are extensive and capable of supporting more vigorous ecosystems. And what the dark oceans lack for in energy they may make up for in sheer bulk.

    It would be truly ironic and, dare I say, Copernican, to discover that even in our home system the phenomenon of life is not actually centered on the Earth.

  • Lionel March 16, 2015, 10:49

    @Ashley and Michael

    Habitable exomoons around gas giants are a fascinating topic and I agree with the points that you’ve made above. How all the variables interact is super hard to predict, for example who would have predicted tidal forces acting on Enceladus could heat an underground sea? I think the tidal forces are going to be really big effect and on some bodies could sustain thicker atmospheres than we would otherwise expect. For example, I’m wondering about what a Titan or larger sized moon at an Enceladus distance from a Saturn sized planet in the habitable zone would be like. (Maybe Rob Henry who knows lots about tidal forces could give his informed opinion)

    Btw, I believe both of you are in the UK, fancy a pint or a coffee some time?

  • Michael March 16, 2015, 15:37


    As you say there are many possible scenarios that could play out, part of the fun really. I would really like a habitable moon around a gas giant as the views would be astounding but they are dangerous places to be near and a ‘technoalien’ species would struggle getting out of their gravitational grip and radiation environments.

    As for a pint or coffee I am up for it, I live near Chelmsford and London is not far away.

  • RobFlores March 16, 2015, 18:04

    Actually, Finding Life in the oceans of any ONE Ice worlds, ( Moons, or Dwarves such as Ceres), would tend to indicate that they may ALL Harbor life if they have Liquide oceans.. The liquid oceans, maybe similar enough that biological interchanges between them, due to asteroid strikes, over billions years have seeded them with living cells. If this is the case, they are probably all related to one another.
    It would create pandemonium in astrobiology if all the Ice world oceans’ in the solar system had living organisms and they were totally unrelated to
    each other.

  • Wojciech J March 17, 2015, 8:57

    One interesting aspect of possible exomoons hosting intelligent life(although that indeed seems to be a difficult scenario), would be that due relative easy access to other moons, civilization would be able to quite quckly develop into interplanetary one.With so many targets in nearby reach it would be possible to pursue colonization and mining much faster and easier than we can.

    Btw I am in London as well, and could attend a meeting after 25th.

    Kind Regards

  • Steve Bowers March 17, 2015, 16:26

    A life-bearing ocean that far under the ice (50km) would be a challenging target for investigation.
    Even worse if by any weird chance there was intelligent life down there. The ice above their heads would be an almost impenetrable barrier – especially since any borehole they could make upwards would tend to freeze solid as they made it.

  • Lionel Ward March 17, 2015, 20:29

    @Michael and Wojciech J

    Fantastic – some time during the first half of April would be good for me. Anyone else for a meetup in London?

  • Michael March 18, 2015, 10:42


    First week in April is fine for me, as at the end of march I am going to see a rock band and I will need plenty of time to recover from the ‘mosh pit’.

  • Wojciech J March 18, 2015, 18:28

    I would like to propose 11th April as the date of the meeting. What do you think?

  • Lionel March 18, 2015, 19:56

    nice :)

    @Wojciech J
    It looks like I might be joining a NASA Space Apps Challenge hackathon in Old Street that weekend, it runs most of Saturday and on Sunday

    Wonder whether you and Michael would either (i) like to meet the previous (Easter) weekend, which is 4 days long (I can do any time)
    or (ii) are up for joining the hackathon? Details on https://2015.spaceappschallenge.org

  • ljk March 19, 2015, 9:06

    Ganymede may have a lot of cousins in the Milky Way galaxy:


  • Michael Million March 19, 2015, 15:05


    ‘Wonder whether you and Michael would either (i) like to meet the previous (Easter) weekend, which is 4 days long (I can do any time)
    or (ii) are up for joining the hackathon? Details on https://2015.spaceappschallenge.org

    I wouldn’t mind going to go to the hackathon in London on the 11th April actually should be a good day out. I have just brushed through the site, is there anything special we need to do like register as it mentioned ‘participant’ when attempting to?

  • Ashley Baldwin April 2, 2015, 19:26

    Dear Lionel. Sorry to miss your meet up! Pardon my rudeness. Only just seen email as bogged down at work . I’m based near Southport in the NW. Happy to meet with you and Lionel at a later date though. My work often takes me to London , next at end of April.
    One thing Paul mentioned in this article was the inability of exchange between the mantle and ocean due to the intervening high pressure ice layer. This sounds bad news, a deal breaker indeed. Interestingly this issue has been addressed previously in relation to ‘water planets’ in the HBZs of stellar systems. Kaltenegger et al and Levi et Al both published in 2013 ( available on arxiv ) on volatile transfer across dense ice layers beneath deep surface oceans on ‘super Earths’ but I would speculate the same applies in this instance.

  • Michael April 7, 2015, 13:06

    @Ashley Baldwin April 2, 2015 at 19:26

    ‘One thing Paul mentioned in this article was the inability of exchange between the mantle and ocean due to the intervening high pressure ice layer.’

    If no carbonates can be recycled through the rocky/ice mantles then we may still be in luck. There will always be a rain of comets and asteroids in solar systems to add these components to their atmospheres although at a reduced rate.