Four of Uranus’s five icy moons likely contain a thin layer of briny (or otherwise enriched) water, astronomers have concluded from Voyager 2 data.

Water is everywhere. We think of it as a liquid, but in most of the solar system, water is frozen as hard as rock, forming the crystalline surfaces of moons, comets, and other wandering bodies. Some icy moons, like Europa and Enceladus, famously host global layers of liquid ocean deep beneath their frozen exteriors. Ganymede, Callisto, and Titan likely do as well. How many other moons are in the ocean club?

A new paper re-analyzing Voyager observations suggests that four of Uranus’ five icy satellites also host oceans: Ariel, Umbriel, Titania, and Oberon. (Only small Miranda, intermediate in size between Saturn’s Mimas and Enceladus, appears not to.) The oceans are desperately thin: less than 30 kilometers (20 miles) thick inside Ariel and Umbriel (both of which are about 1,000 kilometers across, similar in size to Saturn’s Tethys and Dione), and less than 50 kilometers thick within Titania and Oberon (which are larger at about 1,500 kilometers, similar to Saturn’s Rhea and Iapetus).

Grayscale photos of six of the moons around Uranus
Voyager 2 captured the images that were used to create these mosaics of some of the moons of Uranus during its closest approach to the planet on January 24, 1986.
NASA / JPL-Caltech / Ted Stryk / CC BY-NC-ND 3.0

If the oceans exist, they would be left over from much larger liquid layers that formed when the moons first formed. The remnant liquid would be snuggled close to the waning heat of the moons’ rocky cores, sheltered beneath hundreds of kilometers of solid ice. They’d be extremely briny, hyper-concentrated with whatever dissolved materials helped to lower the temperature at which water would otherwise freeze. There are two candidate materials: salt and ammonia.

To validate these models, we would need to send a spacecraft. A salty ocean would be detectable from a spacecraft equipped with a magnetometer, while an ammonia-water ocean would not. But even an ammonia-water ocean remains detectable, because a global layer of liquid would mechanically disconnect the icy mantle from the rocky core. Careful tracking of the motion of surface features as the moons nutate in their elliptical orbits around Uranus could reveal that the icy moons’ outer layers are decoupled from their cores.

New computer models show that there likely is an ocean layer in four of Uranus’s major moons: Ariel, Umbriel, Titania, and Oberon. Salty — or briny — oceans lie under the ice and atop layers of water-rich rock and dry rock. Miranda is too small to retain enough heat for an ocean layer.
NASA / JPL-Caltech

There’s hope that we’ll get to test the conclusions of this paper. A once-a-decade survey of the scientific community conducted by the National Academy of Sciences determined that an orbiter and probe to the Uranus system is the top scientific priority for the next new flagship mission, now that the Mars Sample Return and a Europa-focused Jupiter mission are in process. Learn about the potential of the Uranus Orbiter and Probe mission in the July issue of Sky & Telescope magazine, on newsstands in a couple weeks. Do lots of icy moons host oceans? The only way to know for sure is to visit them!




Image of Yaron Sheffer

Yaron Sheffer

May 6, 2023 at 5:12 pm

These moons are 10 times smaller than Earth, yet their oceanic layers are 10 times thicker than the average thickness of the terrestrial ocean (40 vs 4 km). I would, therefore, *not* characterize those faraway oceans as "desperately thin"...!

You must be logged in to post a comment.

Image of Ludovicus


May 12, 2023 at 5:55 pm

We certainly DO need to visit these planets, to get a better understanding of the geo-dynamics involved. Likely not a huge candidate for life necessarily, but the cryovolcanism involving ices, brines, and rock certainly are interesting!

I suspect a significant NH3-H2O mixture may be present, since the parent planets have these in some abundance.

The only thing is we need be careful of is overly estimating the nature of these objects, since we have never landed on them! If we cannot even properly predict weather, yet, so how can we assume our computer models are even on the mark?! A model is only as good as its assumptions...

So - looking forward to having these investigated. (Though, I wish they'd start sending probes out faster - taking 12+ years to get to a planet is slow going... I'll be in the retirement home by then! Almost...)

More time with the moon, more time with the ice giants, enough with the gas giants, other than the Galilean investigations. Since the Exoplanet foray has lots of Neptune-sized planets, we need to understand these worlds better - focusing on these icy giants is paramount to us getting a better handle on how they form, and what can be expected.

You must be logged in to post a comment.

You must be logged in to post a comment.