Liquid water is not confined to Earth or to worlds close enough to the Sun for warmth. Several moons in the outer Solar System appear to contain oceans beneath ice shells, kept liquid by combinations of tidal flexing, radioactive decay, salts and ammonia.
The exact count depends on how strong the evidence must be. Europa, Ganymede, Callisto, Enceladus and Titan have long formed the core group. A 2024 analysis made Mimas a strong sixth case, while Triton and several moons of Uranus remain plausible additions.
None of these oceans has been photographed directly. Each has been inferred from a different combination of magnetic, gravitational, orbital, geological and chemical evidence.
An ocean world is an inference, not a visual category
NASA uses “ocean world” for a body with a substantial present-day liquid reservoir. The liquid does not have to reach the surface, resemble Earth’s seawater or sit directly on rock. Some proposed oceans contain salts or ammonia, and the largest moons may have water layers trapped between different forms of high-pressure ice.
Evidence can come from how a moon responds to its planet’s changing magnetic field. A conductive salty ocean can generate an induced magnetic signal. Gravity measurements and small wobbles in rotation can show whether an outer shell moves separately from the interior. Fractures, resurfacing and plumes provide geological and chemical clues.
These methods do not produce identical levels of confidence. Enceladus throws samples of its ocean into space, while Callisto’s ocean is inferred mainly from magnetic measurements and interior models.
Jupiter has three leading ocean moons
Europa has the most familiar case. Its young, fractured surface, induced magnetic field and tidal flexing support a global saltwater ocean beneath an ice shell. The ocean may hold more than twice as much water as Earth’s oceans combined and may touch a rocky seafloor, although its depth and ice thickness remain uncertain.
Ganymede also shows an induced magnetic response, complicated by the moon’s own internally generated magnetic field. Hubble observations of shifting auroral bands strengthened the ocean interpretation. Models suggest a water layer roughly 100 kilometres thick beneath about 150 kilometres of ice, possibly divided by high-pressure ice phases.
Callisto is the least active of the three. Galileo measured a variable magnetic field consistent with electrical currents in a salty liquid layer beneath its ancient cratered surface. ESA notes that evidence is less compelling for Callisto than for Europa or Ganymede, but it remains a principal target of the Juice mission.
Saturn contributes Enceladus, Titan and now Mimas
Enceladus supplies the most direct access. Cassini flew through plumes erupting from fractures near its south pole and measured salty ice grains, molecular hydrogen, phosphates and organic compounds. Gravity and rotational measurements indicate that the source is a global ocean beneath the shell.
Titan presents two different liquid environments. Methane and ethane fill lakes on its surface, but Cassini gravity data and radio measurements from the Huygens probe also support a deep internal ocean of water, probably mixed with salts and ammonia.
Mimas was the unexpected addition. It is heavily cratered and lacks the obvious fractures or plumes associated with active ocean moons. A 2024 Nature paper led by Valéry Lainey reanalysed Cassini measurements of the moon’s orbit and rotation. The team concluded that a global ocean lies 20 to 30 kilometres beneath the surface and probably formed less than 25 million years ago.
The Mimas result is a model-based inference from one detailed analysis, not a drilled sample or direct detection. Its importance lies partly in showing that a moon can hide an ocean without advertising it through recent-looking terrain.
Triton and the Uranian moons keep the boundary open
NASA also includes Neptune’s moon Triton on lists of moons that may host oceans today. Voyager 2 saw nitrogen geysers and a young, altered surface during its 1989 fly-by. Triton’s backward orbit indicates that it was captured by Neptune, a disruptive history that could have produced intense tidal heating. A present ocean is considered possible but unconfirmed.
The five large moons of Uranus have moved higher on the candidate list as well. Voyager 2 supplied the only close images in 1986. Later models combining those measurements with improved knowledge of composition and heat flow suggest that Titania and Oberon, and perhaps Ariel and Umbriel, could retain liquid layers.
Ariel’s deep canyons and signs of relatively recent resurfacing have prompted proposals for either a surviving ocean or one that froze in the geological past. Without an orbiter making gravity and magnetic measurements, distinguishing those histories remains difficult.
Old records improve while new spacecraft close in
Mimas demonstrates why archived mission data remain productive. Cassini ended in 2017, but better orbital solutions, shape models and computational methods allowed researchers to extract an interior signal years later. The Galileo and Voyager archives are undergoing similar reanalysis as laboratory measurements and geological models improve.
New spacecraft will provide more decisive tests at Jupiter. Europa Clipper is scheduled to reach Jupiter in April 2030 and make 49 close Europa fly-bys, using radar, magnetic, gravity and compositional measurements to test the ocean and ice-shell models.
ESA’s Juice spacecraft is due to arrive in July 2031. It will compare Europa, Callisto and Ganymede before becoming the first spacecraft to orbit an outer Solar System moon at Ganymede. Neither mission is designed to detect life directly, but both can establish where water lies, how deep it is and whether materials move between the ocean, ice and surface.
The growing catalogue does not mean every frozen moon contains a sea; it means an undisturbed-looking surface is no longer enough to rule one out.
As old measurements become more precise and new missions reach the outer planets, the number of credible ocean worlds may rise again.
