Icy Moons are Ocean Worlds

If you’ve ever spent time with an 11-year-old during their dinosaur phase, you know the feeling of having everything you thought you knew upturned by a pitiless pedant. Science moves on and leaves whatever we learned in school hopelessly out of date. The dinosaurs I grew up with were slow thinking and cold-blooded monsters. They came in muted shades of ugly brown and green, like Subarus. I was taught that the brontosaurus was so chunky it had to spend its life submerged in marshes to help buoy its weight, while the T-rex staggered around on its back feet like Godzilla, waving its little arms.

But starting in the 1990’s, dinosaurs started to get cooler—suddenly they were running fast, covered in feathers, hunting in packs. The T-Rex got upgraded to a high-speed, warm-blooded killing machine. The dinosaur entertainment complex rolled out a whole new set of small, intelligent hunter-killers. And at some point the brontosaurus got cancelled and doesn’t even exist anymore, subdivided into three new dinosaurs I had never heard of.

Something similar has happened to the frozen worlds of the outer solar system. In a series of glow-ups, they’ve gone from being a sort of Space Antarctica of interest only to the most spectrumy of ice nerds, to a series of water worlds that are the most likely environment in our solar system to harbor life. Even poor Pluto, demoted from planethood back in 2006, has been revamped to a candidate ocean world, with the implication that the thousands of Pluto-like objects still undiscovered in the Kuiper belt may be harboring secret seas of their own. Today you can hardly swing a telescope without pointing it at a celestial body hiding a warm underground ocean.

This remarkable transformation in our understanding is the result of just three missions: Voyager, Galileo, and Cassini—along with some computer modeling and hard staring by the Hubble and Webb space telescopes.

As of this writing, six icy worlds (Europa, Enceladus, Titan, Mimas, Callisto and Ganymede) are confirmed to have vast underground oceans of liquid water, and a bunch more (Dione, Pluto, Miranda, Ariel, Triton, Oberon) are on the waitlist.

Let’s meet the crew!

Meet the ocean worlds

Of the six worlds so far shown to have underground oceans, three are moons of Jupiter, and three are moons of Saturn.

Europa is the OG ocean world and a bit of a celebrity for that reason. When Voyager flew through the Jupiter system in 1979, scientists were amazed to discover active volcanoes on Io, confirming a prediction published just a week earlier (!) that tidal heating could substantially warm the inner moons of Jupiter. Since Europa was the next moon out from Io, it stood to reason that tidal heating might be at work there too, an impression reinforced by the craterless smoothness of its surface. But whether Europa was an actual ocean world, or just covered in warm convecting ice, was not definitively settled until 2000, when magnetic evidence for an ocean became overwhelming.

The radiation environment around Europa is punishing; an astronaut standing on the surface would get a fatal dose in about a day. But the same radiation means the ice crust is enriched in molecular oxygen and peroxides (created when water molecules are split by radiation) that may cycle down into the planetary ocean, creating a rich potential environment for life.

Hubble has seen plumes coming out of Europa, but the observations remain tentative. Hopefully Europa Clipper will settle the matter when it arrives at the moon in 2031.

Ganymede, the next furthest moon out after Europa, is kind of a reverse Pluto. By all rights it should be a planet—it’s bigger than Mercury and has its own magnetic field—but Fate has placed it in orbit around Jupiter, and that is where it is going to stay. Ganymede is not a tectonically active world like Europa, and its ocean is locked away under a hundred miles or more of ice. But despite the gruff exterior, the moon shows surface signs of an active past, and contains enough rock (with its cargo of radioactive elements) to stay toasty on the inside, maintaining not only a liquid core, but the largest known ocean in the Solar System.

Ganymede’s intrinsic magnetic field made it hard to apply the same techniques that proved the existence of oceans on Europa and Callisto. But careful observations of aurorae on Ganymede by the Hubble telescope in 2011 showed them to be shifting in a way that only the presence of a global ocean could explain.

Callisto is the outermost of the four large Jovian moons. Where Europa has some of the newest crust in the solar system, Callisto has the oldest, a surface so cratered there is simply no way to crater it further. Callisto is also poorly differentiated, meaning that its interior is a jumble of ice, rock, and small amounts of metal that have not settled into distinct layers. Voyager showed Callisto to be Jupiter’s punching bag, sitting out there in a cold orbit, absorbing impacts. But to everyone’s surprise, magnetic measurements by the Galileo orbiter showed evidence of a deep, liquid ocean, making Callisto the first candidate ocean world in 1998. Absent any sign of surface activity, the inaccessible ocean 150 kilometers under Callisto’s crust is one of the most isolated habitats in the Solar System, cut off from the outside for over four billion years. Whatever may be down there is not coming up without a fight.

And now for the three moons of Saturn:

Enceladus is far smaller than the Jovian ocean worlds, roughly the size of Ohio. But it is a much more exciting place than Ohio, and far more livable.

Enceladus rivals Europa as the most promising candidate for life in the solar system. The ‘tiger stripe’ features on its southern hemisphere send giant plumes of seawater into space, and in 2008 the Cassini probe was able to fly through one and taste the Enceladan ocean directly. Chemical analysis of the plumes and surface has found salty water, all six of the elements necessary for terrestrial life, phosphates, unidentified organics, hydrocarbons, and a kiss of cyanide. Cassini also detected silica dust and molecular hydrogen (potential microbe kibble) originating in undersea rock, the first direct detection of a water/rock interface. At this point the only way for Enceladus to be more habitable would be if we found dense, walkable neighborhoods and an IKEA.

Enceladus was shown to have at least a regional ocean in 2014, and observations upgraded this to global status the next year. It remains the only alien sea we have been able to sample directly.

Somewhat incredibly, there’s no mission in the pipeline to visit Enceladus, even though doing so would cost less than the $5B NASA will spend flying Artemis III to low Earth orbit in 2027.

Titan is the most enigmatic world in the solar system. Barely losing the ‘biggest moon’ contest to Ganymede, it has a nitrogen atmosphere dense enough that an astronaut wearing a wingsuit could fly around just by flapping. It’s worth stressing that, unlike the other moons in this list, Titan is astronaut friendly—the thick atmosphere is a better shield against radiation than what we have on Earth, the low gravity (0.14g) makes getting around a breeze. All you really have to do is remember to bring oxygen and a sweater.

Like those lottery scratch-off tickets that give you a second chance at winning, Titan offers two distinct chances at finding life. On the surface there is a very famliar landscape of rivers, streams, lakes and rainfall, except that the lakes and raindrops are made of liquid hydrocarbons like ethane, with water ice playing the role of rocks. If life exists in this complex surface environment, it resembles nothing we know or can easily imagine, which is part of what makes the prospect of finding it so exciting.

Underneath this remarkable landscape, Titan is an ocean world, with a salty subsurface sea that could be home to more recognizable forms of biochemistry, especially if it is able to interact with the organic-rich surface. The sea was first detected through orbital analysis in 2012, corroborating observations made in 2005, when the Huygens lander observed a radio wave resonance suggestive of an underground salty ocean. As of 2025, there is controversy over whether Titan has a genuinely world-spanning ocean, or whether it is more of a slush of ice and meltwater, but from the point of view of astrobiology, both options are exciting and livable.

What exactly is going on on Titan should become clearer when the Dragonfly probe lands on the moon sometime in 2034.

Mimas is the newest and least expected addition to the ocean world roster. A tiny moon notorious for looking just like the Death Star, it appears far too frozen and rough to sport a liquid ocean, which would be expected to soften its features, especially the giant marquee crater. For a while scientists were positing an alternative explanation for its orbital behavior (an oblong silicate core), but around 2024 they gave up and made peace with the ocean.

The conjecture is that the ocean on Mimas formed recently, a result of orbital changes within the last few million years, and the crust hasn’t had time to get the memo about the new interior. The existence of a surprise ocean raises the likelihood of finding more ‘stealth’ ocean worlds among the minor outer moons.

To put the known ocean worlds in perspective, here’s a schematic of what’s going on in the top 1000 kilometers or so of each of them:

As you can see, the ocean worlds divide into two groups.

Ganymede, Callisto, and Titan all have oceans deep enough for high-pressure ice to form along their bottom. This ice has no real counterpart on Earth—pockets of it may exist in the mantle, but our oceans would have to be some 50 km deeper before the pressure got high enough to form it. This ice layer has the effect of separating the liquid ocean from the rock underneath, which is bad—you want the two to mingle so all kinds of useful salts and minerals can leach into the water.

In addition to this high-pressure ice layer, Ganymede and Callisto also have an extremely thick upper crust that seems to be geologically stagnant, making them less appealing exploration targets.

Europa, Enceladus, and Mimas have a thinner ice shell on top and liquid water in direct contact with a rocky mantle (or core). From a habitability perspective, this is exciting, since rock/water reactions on Earth drive a lot of the chemistry necessary for life. And of course, the thinner shell creates the possibility that we could one day sample these oceans without having to figure out how to make a robot drill through a hundred miles of primordial ice.

But I should stress how provisional these models are. Everything in the diagram above is a guesstimate based on orbital perturbations and fairly primitive computer models. In particular, the models are very sensitive to the chemical composition of ocean water, something hard to observe with remote sensing, and critical to the inner structure of each moon.

For example, if you plug some realistic salt assumptions into the model for Ganymede, you get a ‘club sandwich’ version of the world, with multiple layers of high-pressure ice separated by thin oceans of brine, and a final liquid layer sitting on top of bedrock. Is this configuration realistic? Stable? Habitable? A lot of these questions have to wait until we can land a seismograph or two.

Candidate ocean worlds

In addition to the six known ocean worlds, there is a whole bestiary of candidates just waiting for us to come take a closer look.

Saturn’s moon Dione likely had an ocean in the past, but whether that ocean still exists or has frozen solid is not known.

Most of what we know about the moons of Uranus comes from a single flyby by Voyager 2 in 1986. That encounter showed the moon Ariel to have a very active surface, second in the solar system only to Enceladus. Ariel has spent time in the right kinds of orbit to experience tidal heating (the mechanism that melts Io), and during its flyby, Voyager 2 even detected material consistent with an Enceladus-like plume. But no spacecraft has visited Uranus since, and the data is just too sparse.

All the more so for Titania and Oberon. Models show that both worlds could have a subsurface ocean in contact with rock (good!), but at extremely cold temperatures that would require a lot of ammonia and other antifreeze compounds (bad!).

Neptune’s moon Triton is a close relative of Pluto that somehow got captured into a weird orbit around the ice giant. Voyager 2 observed active plumes on Triton, although these were likely shallow phenomena caused by sunlight shining on dark material through nitrogen ice. More exciting were the observations of a very young crust, showing that Triton still packs enough heat to resurface itself on the regular.

The same holds true for Triton’s cousin Pluto, which the New Horizons probe showed to be a geologically active world that almost certainly has a deep ocean. Modeling shows that a properly insulated ice layer on Kuiper Belt objects the size of Triton or Pluto could sustain liquid oceans for billions of years, with only radioactive rock to warm them.

This opens the door to some dizzying prospects. Rogue planets are believed to outnumber the stars in our galaxy by perhaps 10:1, and many of them could be traveling with ocean moon companions that could remain habitable for billions of years. Any life out there would exist in unimaginable darkness, finally breaking through a thick prison of ice only to discover itself orbiting a black planet under a sunless sky.

Closer to home, the Kuiper Belt is likely full of slowly freezing ocean worlds in the mold of Pluto, which could number in the hundreds.

What do we want from an ocean world?

With such a deep bench to choose from, which moons are worth exploring first? There are several criteria everyone agrees on.

1. Water in contact with rock. A lot of interesting things happen at rock/water interfaces, and several elements believed to be essential for life (phosphorus, sulfur, metal ions) need to leach out of silicate rock to be chemically available in ocean water. Rock/water interactions also create molecular hydrogen, which computer models have shown could be an abundant enough food source to sustain an Earth-sized ocean ecosystem for billions of years.That said, the high-pressure ice lining the ocean bottom on Ganymede, Titan, and Callisto might not be a showstopper—it’s possible that it convects, or that material erupts through it, preserving the connection between deep rock and ocean. Clearly at least some of it is getting through, since the oceans are salty. But given a choice, we want to prioritize worlds where we know liquid water flows through rock.
2. Antiquity. Since we have no idea how long it takes life to arise, it’s prudent to explore oceans that have been around for a few billion years. This is easier said than done. Moon orbits in the outer solar system are chaotic, and oceans may experience multiple freeze/thaw cycles over the aeons. This is the one place where Europa outshines Enceladus, since the former almost certainly has a primordial ocean, while the Enceladan ocean is of indeterminate age, and might be only a few hundred million years old.
3. Remodeling. The ideal is the smooth, almost craterless flatness of Europa or Triton, which suggests active turnover between the surface and interior. Jupiter has contenders at both extremes, with the crust of Europa being about 50 million years old (basically brand new), while Callisto is the most heavily cratered body in the solar system, a poor moon that has done nothing tectonically for its entire existence except serve as a punching bag for meteors. The appeal of a young crust is that it implies dynamic movement and recirculation from below, either through convection (warm ice flows readily) or by cryovolcanoes erupting and coating the surface. Both mechanisms bring material from deep underground to the surface (great for the search for life!) and by symmetry carry material from the surface down to the ocean, which plays an important role on radiation-fried moons like Europa, where the top layer of ice gets enriched with enough oxygen to fuel an entire ocean.
4. Radioactive rocks. Radioactive elements like potassium and thorium are Nature’s electric blanket, helping keep even the iciest worlds toasty at the core. Even on a tidally heated world like Europa, much of the nternal heat still comes from radionuclides, and they are the main sustainers of heat on remoter worlds like Pluto. The gold standard would be to find a moon that formed early enough to capture some aluminum-26, a short-lived isotope present during the early years of the solar system that would have really brought the heat.
5. Plumes. Plumes can save you big money on a lander by propelling the contents of an ocean directly into space, where it can be sampled by passing spacecraft. If we’re really lucky, a big vent may spew remnants of some space fish directly onto the surface ice, saving us a long and contentious search for life. Plumes and vents might also sustain some kind of weird ecosystem along their edges, the way subsea hydrothermal vents do on Earth. It’s not likely that an alien squid is going to hit the windshield of our next Enceladus orbiter, but it’s not out of the question, either, and that’s why everyone loves a plume.

Together these factors explain why Europa and Enceladus are such attractive targets for astrobiology.

But in starting to look for life on these moons, we have to be careful to not let our thinking get too uptight.

Unlearning some bad Martian habits