In the Star Trek: Voyager episode "Thirty Days", there was a planet made completely out of water, which is the inspiration for this question. However, that planet was held together by an artificial containment field. I'd like to know if a water planet would be possible that would be held together just by normal gravitation.

To be clear what the properties of this water planet should be:

  • The planet is completely made out of water (that is, any non-water substance makes up a negligible amount of its mass). There may be a non-water atmosphere above, but there's no stone or metal center; the water goes down to the center.
  • Most or all of the water on the planet's surface is liquid (there may be polar ice caps, but they shall cover less than half of the planet).
  • The planet is held together by its own gravitation.

Is such a planet possible? And if so, could there be a plausible mechanism how it could be created naturally?

  • 27
    $\begingroup$ Liquid water all the way down? Or is a water ice core ok? $\endgroup$ – Samuel Dec 4 '14 at 20:33
  • 7
    $\begingroup$ I believe the Star Trek episode you have in mind is VOY 5x09 Thirty Days, where Voyager encounters a ball of water held together by a containment field generated by an ancient field generator, which in turn was built by an earlier civilization which disappeared before the people who live there when Voyager encounters them arrive. Do note that this does not fit your criteria; it is not held together by its own gravity (in fact generator failure is a major plot device), and it is debatable whether it is made up completely of water. $\endgroup$ – a CVn Dec 4 '14 at 20:41
  • 7
    $\begingroup$ The water on Naboo goes all the way down through the core, doesn't it? If it's in Star Wars, it must be possible! $\endgroup$ – KSmarts Dec 4 '14 at 20:55
  • 13
    $\begingroup$ @GrandmasterB actually that is only normal ice, I think there are 10+ different types of ice, most are not possible here on earth, some of them would be created at great pressures such as being in the center of a planet size mass of water. They are called Ice because they are forms of solid water. $\endgroup$ – bowlturner Dec 4 '14 at 20:57
  • 17
    $\begingroup$ @bowlturner You really have to be careful with the ice-nine. $\endgroup$ – KSmarts Dec 4 '14 at 21:13

18 Answers 18


There have been a few exoplanets discovered that might be what you're after. Gliese 436_b might be close to what you are asking for here, but likely contains an (albeit tiny) rocky core.

The component here that makes these planets viable is what is known as 'hot ice' - water actually has around 10–12 solid states (only one of which is the ice we know). Under extreme pressure, water molecules take other forms, all of which could act as a solid core for a water planet. Think of carbon and the many states it can take under various pressures (from graphite to diamond) - water has some of the same properties.

If you can wade through the article, here is a water phase diagram that displays its various forms.

A liquid water surface with various forms of solid water beneath is more than feasible. Might even support a magnetic field.

Added: After reading that article more...there are atleast 15 ice polymorphs, a little more than my 10–12 estimate.

More added:

I had to research this a bit, but apparently some of these ice structures are more than capable of being magnetically conductive and should work as a metal core. It's more than possible that this ocean world could support a magnetic field strong enough to protect the world.

And yet more added:

http://www.cfa.harvard.edu/news/2012-04 Looks like we've found a few of these now.

These planets tend to form in the far reaches of a solar system where ice is more abundant. The planet then 'migrates' inwards and into the habitable zone. What qualifies as 'migrates' is a bit beyond me, but although it's unlikely to form in the habitable zone, it can move into it. Whether or not that's stable and for how long it will stay there is another question

  • 2
    $\begingroup$ I was looking for an article about the different phases of Ice, glad you found them. $\endgroup$ – bowlturner Dec 4 '14 at 21:00
  • 4
    $\begingroup$ @ckersch: I just noticed that the page you linked even contains a link to a complete Wikipedia page just for that type of planet. $\endgroup$ – celtschk Dec 4 '14 at 21:11
  • 6
    $\begingroup$ Considering the context, it might not be totally amiss to point out that the Ice-nine in that phase diagram has very different properties to the ice-nine in Vonnegut's Cat's Cradle. $\endgroup$ – Jon Hanna Dec 5 '14 at 4:27
  • 2
    $\begingroup$ The link I provided has this one ice-9 "Kurt Vonnegut's highly entertaining story concerning an (imaginary) ice-nine, which was capable of crystallizing all the water in the world [83], fortunately has no scientific basis (see also IE). Ice-nine, in reality, is a proton ordered form of ice-three, and only exists at very low temperatures and high pressures and cannot exist alongside liquid water under any conditions." $\endgroup$ – Twelfth Dec 5 '14 at 18:19
  • 6
    $\begingroup$ "wade through" +1 I had a giggle. $\endgroup$ – Gusdor Dec 8 '14 at 15:51

It's plausible to have a planet made almost entirely of water (the atmosphere is part of the planet). There is such a planet in the book Lockstep by Karl Schroeder. It's not a key plot device, but a great story anyway.

To have such a planet naturally occur is highly unlikely though. It's not impossible, the universe is so huge, highly improbable things happen all the time. I didn't really guffaw at it for the Lockstep story, but if it's central to your plot, readers may scrutinize more closely. More likely is that you'll have a rocky core.

I'm unsure you'll have polar caps. More likely it'll be all liquid or all frozen. With no land mass to anchor the ice it would flow freely and not accumulate in one place.

Sounds like a neat place though. Especially if you place a moon around it for a literal tidal wave circling the planet.

  • 8
    $\begingroup$ A tidal wave without land to measure it against is extremely boring. A civilization would probably live for a long time, before even noticing it. $\endgroup$ – Jens Schauder Dec 5 '14 at 11:28
  • 1
    $\begingroup$ @Samuel: It wouldn't be on the waves of a ship, I doubt any ship ever has noticed a "tidal" wave. $\endgroup$ – Mooing Duck Dec 5 '14 at 18:23
  • 5
    $\begingroup$ @Samuel The very plastic planet would take the shape of an ellipsoid with the long axis pointing at the moon. For a relatively speaking massive moon like with Earth the elongation would be a few meters, and you would rid it up during a quarter day, then the moon passes overhead, and you ride it down a quarter day. Some good instrumentation will be demanded to notice that. Even measuring the circumference of a planet using no instrumentation but with land to have good references yields many % of error. $\endgroup$ – David Tonhofer Dec 6 '14 at 19:49
  • 3
    $\begingroup$ @Samuel I am asserting this is the only possible configuration, which is the particular one we are talking about here, re: no land. If there is no land, there is only a single configuration that can occur: the one I described. "Tidal waves" on Earth are consequence of landmasses disturbing the flow of the water mass around the planet. $\endgroup$ – David Tonhofer Dec 8 '14 at 21:47
  • 4
    $\begingroup$ @Samuel that is like calling a hill an earthquake... $\endgroup$ – Aron Feb 13 '15 at 5:12

As long as you accept an ice core instead of liquid water core there is no issue with stability; once the water world exists it will be stable enough. Although anything that would normally rob a planet of its atmosphere would be a serious issue.

How can I make this claim without doing numbers or looking references? Well, earth has stable hydrosphere at earth normal gravity, from that it follows, unless I am misunderstanding how gravity works, that a lower density planet with approximately earth normal surface gravity could have a stable hydrosphere.

And there are gas giants with densities lower than water and surface gravity higher than earth so it isn't really necessary to do that math either.

The plausible mechanism is the hard part. Basically you are asking for there to be lots of oxygen in the orbit for the water, but nearly no carbon, silicon, aluminum, or other similarly common and similarly created elements that would create a solid core. If you accept the existence of significant amounts of methane, ammonia and carbon dioxide that would help slighly, but it wouldn't help with aluminum and silicon. For that matter, suphur, iron, and nickel would probably have to exist in significant quantities on something the size of the water world.

So essentially, this question could be rewritten as "Is there a way for a star to go nova or super-nova in a way that creates an abundance of oxygen, but insignificant amounts of other metals." (metal = not hydrogen or helium) Off hand, as a non-expert, I doubt it very much. The reactions are not really deterministic enough for that.

As for having those other elements being depleted just before planet forming... That I can see happening, but it would IMHO only get you down to very small core of "not water" at best. Even if you assume some freak incident removing everything you don't want, the unwanted elements would still be in the same star system and some of them would eventually return as dust, comets, and other similar debris falling time. So the state of no non-water core would not be stable over time.

Some have suggested biological removal of heavier elements. I considered this, but while it depletes the elements from the water solution, it actually converts them into a solid insoluble form that after the organism dies falls down. So rather than helping with getting rid of a solid core, it actually adds the requirement for volcanism or some other recycling method to get the elements back to the solution, if you want to have native lifeforms.

I should add that since the planet would have lower density than earth and needs similar surface gravity to retain water and avoid gathering hydrogen and helium and becoming a gas giant, it must necessarily have much larger radius and mass than earth. This is implied by that math I dismissed before as "not necessary to do". This in turn implies that the core has significantly higher pressure than our core does. This means that if metals are present, the core will be metallic. The "ice core" fails because the pressure will squeeze the water out from the core.

Of course, a small core would be covered with exotic ice, so the difference from ice core could be negligible in practice.

  • 1
    $\begingroup$ My thoughts exactly. Though I add that some stuff can be disolved or emulsified and not form a hard core, or float, or be expelled from the solid phase that exist in the core. That would have the best use for a story, IMO. The people could find lumps of material at the liquid/solid boundary even though that's near the center: but that's where life would arise too. $\endgroup$ – JDługosz Dec 6 '14 at 1:16

From http://www.expanding-earth.org/page_10.htm

The daily influx of meteorites and meteor dust is well known to scientists, but the total volume of mass daily added to Earth's surface is difficult to estimate and is not well documented. Estimates of total volume published by NASA vary widely (or wildly?) just for dust alone, ranging from as little as 1,000 tons/day (300,000 metric tons/yr, Dubin and McCracken, 1962) to 55,000 tons/day (20,000,000 tons/yr, Fiocco and Colombo, 1964). However, a more recent estimate puts the accreting dust volume at approximately 78,000 tons/yr, or 214 tons/day.

It's likely that your water planet will also be the resting place of large quantities of cosmic solid matter, and that the larger meteors would sink towards the centre. Dust-sized particles might remain in suspension, especially if the surface is turbulent.

The older your planet is, the more non-watery its centre is likely to become.


Obviously it is possible - its own gravity would hold water megadroplet together with no problem.

Problem: lack of the significant metal core ==> lack of own magnetic field ==> lack of magnethosphere ==> solar wind flares strips upper layers of your waterworld atmosphere, and droplet may evaporate in few hundred millions of years (will be losing mass constantly, and it will be competition between evaporation and space debris falling down).

Edit: Seems that water droplet big enough can compress water with it's own gravity hard enough to create rotating magnetic core. Glad that I was able to hint the right questions to be asked and contribute to best answer.

  • 2
    $\begingroup$ Your question made me research it, but oddly enough, several water solid states appear more than capable of being a conductive core, suggesting a pure water world could have a magnetic field. It's suggested that Neptune's magnetic field could be attributed to this $\endgroup$ – Twelfth Dec 5 '14 at 18:26
  • $\begingroup$ Conductive is not enough. You want also rotating as compared to surface. Neither Mars nor Moon have rotating magnetic core, and so solar wind was able to strip their atmosphere. $\endgroup$ – Peter M. Dec 5 '14 at 18:43
  • $\begingroup$ Agreed, but high pressure ice acts as metallic (note moon and mars are not high pressure enough). Neptune and uranus on the other hand physics.aps.org/story/v15/st12 "At high temperatures and under extreme pressures, water’s hydrogen nuclei can roam and conduct electricity, as electrons do in a metal... Through computer simulations and experiments, the team has found clear hints of a long-sought “superionic” solid phase of water. The discovery could help explain what powers the magnetic fields of Neptune and Uranus, planets that harbor large amounts of “hot ice” in their depths." $\endgroup$ – Twelfth Dec 5 '14 at 18:50
  • 1
    $\begingroup$ THen you just need the rotation $\endgroup$ – Twelfth Dec 5 '14 at 18:52

Well the ice giants in our solar system (Neptune and Uranus) are largely water and ice. Uranus in particular has a smaller rocky core, so an exaggerated version of Uranus could be your model. Of course that's quite different from a landless Earth with oceans all the way down.
For one thing the atmosphere is much thicker (though not nearly as thick as Jupiter and Saturn). That is something you might not be able to get away from, as the atmosphere would have to be heavy in light hydrogen yet heavy enough to cause pressures favorable to liquid water. Hydrogen will almost certainly be the most common element (with helium a distant second); the only reason we have little of it in our atmosphere is that the solar wind probably blow most of it away during formation. If the same happened to your water world, the early water vapor would probably go as well. Also, the format of the "water" is probably not what we are used to. While the "surface" off these planets are frigid, they heat up as you go deeper into the core. What you end up with is probably a combination of exotic forms of ice and superheated liquid. Probably not a place you want to practice your backstroke...


The only way you are going to have such a planet is if it's an artificial construct.

Lets suppose you gather enough water in one place somehow. Yes, you can have a body of H2O with enough self-gravity to act like a planet. You're not going to get a pure waterworld out of it, though:

1) The center is going to be solid. At the pressures involved the water will freeze. You can't overcome this with a hot core because the heat needed will cause the core to boil--major convection, the temperature difference drops.

2) Real planets are in environments with debris floating around. The dinosaur killer hits? You now have a small rocky core in your waterworld.

  • 1
    $\begingroup$ Not sure if this is correct. The center will be solid, but not ice as we know. Water has several different solid states, not just 'ice'. The dinosaur killer hits won't become a core as this water solid will be there instead...might have a layer of rock added to it though. For what it's worth, water can remain solid at temperatures well over 1000 Kelvin $\endgroup$ – Twelfth Dec 5 '14 at 18:24
  • 1
    $\begingroup$ @Twelfth: Over millions of years of tidal forces (from the sun if nothing else) heavy stuff will fall and light stuff will rise. Resulting in all of the asteroids, comets and junk ending up in the center acting as a core. $\endgroup$ – Zan Lynx Dec 5 '14 at 22:07
  • $\begingroup$ @ZanLynx - Agreed, but the alternate forms of water solids could be more dense than rock in the core of the planet. I really can't find much on densities of these various water solids $\endgroup$ – Twelfth Dec 5 '14 at 22:47
  • $\begingroup$ @Twelfth The question was a planet of pure water. Ice isn't water. And if the asteroid ends up spread over the core that means there's a core--not just water! $\endgroup$ – Loren Pechtel Dec 6 '14 at 3:34
  • $\begingroup$ @LorenPechtel although it hasn't been edited into the question, the OP states in the comments on the question that an ice core would be acceptable. Not all of the water needs to be liquid. $\endgroup$ – trichoplax Dec 6 '14 at 15:24

Remember that for the planet to be mainly water:

  • It must exist in the temperature sweet-spot between ice and water vapor. Not just at the surface, but for most of its depth. Seems counterintuitive. So that imposes big constraints on its sun in terms of distance and heat.
  • This isn't stable over time: its gravity must be sufficient to minimize atmospheric escape losses. That might imply a core of ice-IX, water-VI, -VII or somesuch - you can do the numbers.
  • But anyway, without a water source or internal radioactivity in the core, such a planet would have continuous atmospheric loss. Need it exist for 1 million years? 1 billion? more? Do we also consider this over the lifecycle of its sun?
  • I don't know if its surface boils or evaporates if that's worse for losses, but I imagine it would be.
  • If it's all water, there's an implicit assumption that it continuously spins, i.e. never has an icy darkside. But doesn't fluid dynamic viscosity kill rotation very quickly in a water sphere? Hence you surely end up with an icy darkside. Unless its "year" of orbiting the sun is so short that the darkside never ices up => imposes huge constraints on orbital period and radius. (But if orbital radius gets very small, it boils... the sun is pumping too much heat into the planet, and the atmosphere escapes)
  • So we require this whole pressure-thermal-gravitational-orbital spherical(/geoid) arrangement to be stable over time, and over the lifecycle of its sun, and for most depths on the planet. Intuitively this seems to be numerically implausible, before you write a single equation.
  • 2
    $\begingroup$ As for the temperature sweet spot (your first bullet point), the OP does specify that a frozen core is acceptable, and doesn't specify a maximum size of that core that I see. Hence the "most of its depth" should not be a major issue; you can make the liquid water as deep (or not) as you want. $\endgroup$ – a CVn Dec 5 '14 at 9:05
  • 3
    $\begingroup$ Water yes, liquid water no. Third comment from the top on the question, by OP: "A water ice core would be OK (I guess only liquid water would be impossible due to the huge pressure)." $\endgroup$ – a CVn Dec 5 '14 at 11:52
  • 1
    $\begingroup$ I didn't do the numbers, but given the possibility of convection processes and the high heat capacity of water, could the heat from the sun convect through the planet and keep it liquid? $\endgroup$ – Stefan Majewsky Dec 5 '14 at 15:46
  • 2
    $\begingroup$ @StefanMajewsky: Heat isn't the issue. Pressure is the issue. $\endgroup$ – Mooing Duck Dec 5 '14 at 18:25
  • 1
    $\begingroup$ Also note the sentence directly from my question: "Most or all of the water on the planet's surface is liquid" which, while not explicitly saying it, implies that the water that is not at the surface might not be liquid. Frozen water is still water. $\endgroup$ – celtschk Dec 5 '14 at 19:22

I had some thoughts.

First, you need to separate the water from other elements, and presuming that takes place in a condenced situation, then the water removed from that and finally the pure water used to make a single body.

It's been noted that a quarter-million miles off is a stange place for a planet to keep most of its lithosphere, to point out that the moon was formed from the ligher parts after the earth had fractionated. If a similar impact happened on a water world, it might not easily have the same effect. But that's the starting point and I elaborate on that basic idea.

Also, it could be a sattelite of a giant planet, fair enough? After all, Titan is called a terrestrial planet by those who study the conditions present on the surface, and where it's located does not come into that definition.

An important intermediate step is to have ice asteroids. We have bodies that have ice patches among chunks of different types. We just need such chunks to be alone.

(As an aside, note that Enceladus has giesers that expell water at orbital velocity, forming a tenuous water ring around Saturn.)

So first you get planetoids that are large enough to fractionate but small enough to cool completely and later get broken without totally vaporizing. Off-center collisions can create asteroids that are composed only of the icy outer layers.

Various ideas can be posed as to how they separate from the rocky fragments. One such is that the main body is held in a resonance and won't easily leave that orbit, even if peturbed. Only small-enough pieces knocked off of it will make an excursion and possibly be caught in a different resonance, where they combine and add to a water-only body.

Being a giant primary, late heavy bombardment will come this way, knocking more pieces off, repeadly for millions of years. If one parent body is too large, how about a belt of smaller separated bodies. They crunch together over time, not hard because they are all going the same way. The small pieces can get thrown out of the belt due to gravitational slingshots.

Or, a large body that fracionated and solidified can be cracked up (perhaps from volume changes due to phase change and cooling or rewarming on a very eccentric orbit) and then it gets close to a giant and tidal forces pull the rubble pile apart without heating it! On an approach, the outer layers' pieces get captured in one cohort, and inner pieces another. Or the weak ice was more cracked and easier pulled apart. Now we can't have it reform within Roche's limit (that tore it apart!) But that was just perigee of a singular approach to the giant. Their new orbit circularizes and re-combines, perhaps with the help of Lagrange points or resonances.

A variation of that: a close approach causes tidal forces to pull the liquid surface off a body, cleanly taking only the liquid and not the durable solid at that distance. This would form three lobes, as escape-velocity tides. You might end up with a dense planet having two water moons, or they might recombine. Having two eliinates the problem of combining releasing too much energy. Unless a vapor ring transfers material from the smaller to the larger over geologic time.

Enough? Food for thought.

  • 2
    $\begingroup$ I love your idea of a rocky core being separated from its ocean into a rocky world and two water moons, since the Roche limit for water is at a larger radius than the Roche limit for rock. $\endgroup$ – trichoplax Dec 6 '14 at 15:34

I'm going to try to answer part of the question. The question, "is it possible" has been mostly answered. I'll try to describe the possibility of one that is known to exist to help with the question.

NASA, etc. have discovered an amazing number of exoplanets in the last few years, and one GJ1214b in 2012 appears to be completely made of water (the atmosphere might not be 100% "water"); the surface appears to be liquid and the center is not "ice" but highly compressed water - there is a difference. So yes, it would be compressed H2O that is still not ice, but is not frozen water.

I stress that it "appears to be," and I agree that it is very likely, but the details are not as well confirmed as our closer neighbors.

  • $\begingroup$ The gallon of water would not last, but evaporate into space since it would not have enough gravity to hold on to the water vapour as an atmosphere. Whether liquid or solid (near to or far from its sun), either way it would be losing vapour until it disappeared completely. $\endgroup$ – trichoplax Dec 6 '14 at 15:38
  • $\begingroup$ Ah, I meant in your own artificial environment - a tiny little world you can create - that's why I threw it in at the end; agreed: fried or frozen, it'd be gone. $\endgroup$ – Mikey Dec 6 '14 at 17:48

Within the Solar System, Saturn's moon Titan is a fairly close analogue. According to http://en.wikipedia.org/wiki/Titan_%28moon%29

Based on its bulk density of 1.88 g/cm3, Titan's bulk composition is half water ice and half rocky material.

This is presumably by mass. From the following we can deduce that the rocky core is expected to be (2100/3200)^3 = 28% of its volume.

Titan is 5,150 kilometres (3,200 mi) in diameter

Titan is likely differentiated into several layers with a 3,400-kilometre (2,100 mi) rocky center surrounded by several layers composed of different crystal forms of ice.[27] Its interior may still be hot and there may be a liquid layer consisting of a "magma" composed of water and ammonia between the ice Ih crust and deeper ice layers made of high-pressure forms of ice. The presence of ammonia allows water to remain liquid even at temperatures as low as 176 K (−97 °C) (for eutectic mixture with water).

We can also deduce (making the gross overestimate that gravity is the same all the way to the core) that the pressure at the core is density x radius x gravitational acceleration = 1880 x 5150000m/2 x 1.352=6.5GPa and that the triple point of ices VI and VII with liquid water (355K, 2.216GPa) will be reached at a depth of 2216000000/1.352/1000=1693000m implying that with sufficient temperature increase, water might become liquid all the way to the rocky core. See http://en.wikipedia.org/wiki/Ice#mediaviewer/File:Phase_diagram_of_water.svg The triple point for liquid water with ices VI and V is at a temperature close to the "normal" freezing point of water that would be expected at the surface and is probably a better value for pressure reference. This has a pressure of 632MPa, giving an ocean depth of 467000m.

Currently, Titan has a solid ice surface and a largely nitrogen atmosphere, with a surface gravity of 0.14g. But in future, this will change as the sun expands. Saturn will also be strongly affected which isn't mentioned in the text below, and given the low gravity, the nitrogen atmosphere will be depleted by the increased temperature. I do wonder if sufficient atmospheric pressure could be maintained to keep the surface water in the liquid range without freezing, but Titan's current surface pressure (146kPa) is 45% higher than Earth's. Selective capture of the heaviest gases boiled off from Saturn's atmosphere (CO2) might help maintain atmospheric pressure.

Conditions on Titan could become far more habitable in the far future. Five billion years from now, as the Sun becomes a red giant, surface temperatures could rise enough for Titan to support liquid water on its surface making it habitable.[157] As the Sun's ultraviolet output decreases, the haze in Titan's upper atmosphere will be depleted, lessening the anti-greenhouse effect on the surface and enabling the greenhouse created by atmospheric methane to play a far greater role. These conditions together could create a habitable environment, and could persist for several hundred million years. This was sufficient time for simple life to evolve on Earth, although the presence of ammonia on Titan would cause chemical reactions to proceed more slowly.

So.. what about the rocky core? Obviously pure water is impossible, and any heavy impurities are bound to sink to the bottom. Titan's rocky core is believed to be already surrounded by ice, and in a concentrated ammonia solution. Possible mechanismas that could reduce the amount of rocky material concentrated in the core are tectonic activity and biological activity. BTW, I fail to see why ammonia should slow chemical reactions in extraterrestrial life, which would evolve for the prevailing conditions. It is also possible that our descendants will (intentionally or unintentionally) introduce terrestrial life to Titan.

Most rock is composed of SiO2 and Al2O3, either alone or combined with metal oxides to form silicates and aluminates. Ammonia is present on titan, and under certain conditions soluble ammonium silicates can be formed: http://pubs.acs.org/doi/abs/10.1021/i360034a025 though these would be unstable on a geologial timescale.

To transport silica out of the ice core and into the ocean a large degree of tectonic activity would be required, and I'm struggling to come up with a mechanism for that. The ice phase under the conditions is ice VI or VII which have a similar density to water, so mountain peaks if any could be very high. Geothermal activity is out I think, because this requires heavy elements (for radioactive heating), which are precisely what we want to avoid.

Although the temperature gradient in an ocean is small, the best way I can think of to have turnover of the core material and get the silica out into the ocean is to have melting at the equator and freezing at the poles, causing slow but steady deformation of the core. Photosynthesis and biological activity may also produce gradients of concentration between equator and poles which cause global mass transport from/to the poles.

Although molluscs build their shells from calcite, a group of microscopic terrestrial organisms called diatoms build their shells from silica. A colony of silica shelled organisms could remove silica from the ocean and concentrate it in particulate form in their shells, which could remain mobile due to swimming. This would tend to help the silica in the core to dissolve in the aqueous ammonia ocean.

TL;DR something similar but not exactly the same as what you are asking may be possible, even within our own solar system (albeit in the distant future.) A core of ice with an ocean interface of ice VI or VII is probable for a body with sufficient gravity to produce the necessary atmospheric pressure to have liquid water at its surface, but might be avoided if the conditions are just right. The core is likely to contain rocky impurities, which gravity will tend to pull toward the centre. Mechanisms for distributing and dissolving the rocky impurities are conceivable but limited.


Without at least a primordial magnetic field, the water can't be replaced and any planet of any reasonable size will still lose its atmosphere and its water. I'm not sure about something the size of Jupiter.

This takes place over a long period of time by direct impact of unfiltered radiation from the star which just effectively boils it all away.

  • $\begingroup$ If I may be blunt, I hated your last answer, but I like this one. It makes perfect sense. $\endgroup$ – HDE 226868 Dec 7 '14 at 22:24
  • $\begingroup$ The most recent weekly SETI seminar, on youtube, is a presentation on the latest study on just how the sun stipped the atmosphere from Mars. $\endgroup$ – JDługosz Dec 7 '14 at 22:52
  • $\begingroup$ Water solids can support electric charges and is a potential theory why a planet like neptune has a magnetic field. $\endgroup$ – Twelfth Dec 20 '14 at 1:25

Your water world must be born as an ice moon

Icy moons are moons that mainly consist of ice and water, it is possible that the core may be consisted of Ice II or some other polymorph of water ice. Now this is in no way what you want, but if we make this icy moon orbit a gas giant and then have this gas giant change its orbit from a Jupiter style orbit into a hot Jupiter orbit, then the icy moon is now technically in the habitable zone. And what does that mean? The ice will melt into water, creating a planet of mainly water.


Assuming you mean a liquid ocean of water which beings sufficiently adapted could potentially swim all the way through, it would have to be small because water when compressed enough becomes ice - or else - to have a hot core, which it might have soon after formation, or be tidally heated.

So, first, the easiest case, if you don’t need it to have enough gravity to hold an atmosphere, I don’t see why not. Basically you want a large comet, in an orbit which keeps it permanently liquid. We could create such a world artificially in our solar system with mega engineering by diverting a comet into just the right orbit around the Sun.

However, unless we add something extra to the picture, it wouldn’t last long. The problem is that water evaporates rapidly in a vacuum. And to have enough gravity to stop that happening

With surface temperature of 273.15 K and using the equation for mass loss of liquid water in a vacuum of

$\text{(pe/7.2)} \times\sqrt(M/T) kg / m2 / sec$ (equation 3.26 - compare calculation results here: Modern Vacuum Physics)

where M is the molar mass, T is the temperature in kelvin, pe is the vapour pressure, which for water at 0 C (273.15 K) is 611.3Pa, (Vapour pressure of water at 0 C), M = 0.018 kg, gives $$\text{(611.3/7.2)} \times \sqrt(0.018/295) = \text{0.663 kg / m2 / sec}$$.

So you lose about 57 meters a day thickness of liquid water exposed to a vacuum, or about 20.9 kilometers of water per year. The rate of loss goes up if the temperature increases and is 2.495 kg / m2 /sec at 295 k, or 22 C. That’s 215.6 meters per day and 78.6 km per year.

So, a liquid water comet would not last for long. That is unless you get a constant influx of other comets bringing more water to it.

What if the object is large enough to retain liquid water for long periods of time?That’s only possible if it has at least enough gravity to retain a significant amount of atmosphere, even if the atmosphere is just water vapour, or oxygen (after dissociation of the water by radiation).

~But then - it will surely have a solid ice core. In that case, if the water is also salty, it might well have a “club sandwich” type pattern of alternating layers of ice and water as suggested for Ganymede, of various types of ice, with some of them “snowing upwards” [Ganymede׳s internal structure including thermodynamics of magnesium sulfate oceans in contact with ice]

Ganymede׳s internal structure including thermodynamics of magnesium sulfate oceans in contact with ice

But even Ganymede is not large enough to retain an atmosphere to protect the surface layer of water. Its diameter is 5,268 km so if brought close enough to the Sun to have a permanently liquid surface layer, it would vanish completely in 67 years.

It could build up a temporary atmosphere though as the water evaporated. It’s gravity is similar to the Moon’s.

Can we terraform the Moon? If yes, how difficult is it and is it possible with the current technology?

So using a calculation from that answer, if you hit it with a comet 164 km in diameter you’d have enough material for an atmosphere which would last for 10,000 years. Since the volume goes up as the cube, that means with a similar pressure atmosphere, a moon the size of Ganymede could last for $\text{10,000}\times\text{(5,268/164)}^{3}$ = 331 million years before evaporating completely if it built up an Earth pressure atmosphere. And the atmosphere would consist of water vapour and oxygen, so might well be breathable too, especially if you can somehow introduce some nitrogen as a buffer gas.

But that’s still no good if you want the core to be liquid all the way through.

There is another solution though. If you are willing to do it artificially, you could cover the entire surface of a small comet with a low density liquid which also has a low evaporation pressure.

Indeed, comets are rich in organics anyway, so if you could bring a comet to just the right distance from the Sun, not too far, not too close, then as it melted, it would develop a layer of scum like that. And that might well be habitable too, with organics and an oxygen rich ocean too, due to similar processes to the ones that make Europa’s ocean oxygen rich.

Organics with a high evaporation rate would disappear leaving only those with a low evaporation rate, and perhaps solid layers as well.

So if you are okay with your planet being a tiny comet sized object, and your water can be a bit “dirty” with organics, which means it can also support life, I’d say yes, it does seem possible.

Europa’s ocean may be as much as 100 km thick, with a surface layer 10 - 30 km thick.

enter image description here Based on that, you could have a minor planet made of ice, 260 km in diameter, and consisting entirely of water, I think, with a surface layer of organic ionic fluids or a scum of organics in solid form floating on the surface. That could last for billions of years.

That makes it about the same size as 88 Thisbe enter image description here

Vesta’s double that diameter enter image description here

Vesta, Ceres and the Moon to scale at 20 km per px

I’m just using the figures for Europa and the depth of its subsurface ocean, which is kept liquid by tidal heating, and assuming the situation is similar - so this is just a rough estimate as it would depend on what you have by way of an energy source to keep your planet or moon warm. With just surface heating, surely the center would cool down eventually.

Tidal heating could be a way to keep your planet liquid just as for Europa, so if you make it so that it orbits a hot Jupiter - those are planets like Jupiter that end up in orbits close to their sun, and they may well have liquid water moons.

Another solution, without the layer of ionic liquids or similar, is to have a constant influx of comets to replenish the water. I can imagine some scenarios where that could work, e.g. soon after formation of a solar system. It also might work for a while later on in a white dwarf star with material brought into it through destruction of its Oort cloud and perturbing effects of an extra planet, see Our Solar System Could Lose One Or More Of Its Gas Giants Billions Of Years In The Future - and that would also help keep it hot. In a situation like that maybe even quite a large minor planet would stay hot enough to stay liquid all the way through. But the tidal heating + surface thin layer seems the easiest solution to me.

So, in short, I think this scenario could actually exist in nature, if you don’t mind having an ocean rich in organics, covered with a thin layer of organics, and make it a moon orbiting a gas giant rather than a planet on its own.

This is just a rough estimate. Would be interesting if someone was to do a paper on it - has anyone? Would a liquid water world the size of Vesta or even Ceres be possible, with tidal heating to keep it warm? Can a hot Jupiter have a moon of pure ice? (I don’t see why not if it formed far enough away from its host star originally, but would be interesting to know how likely that is).

This is a copy of my quora answer to Is a planet entirely made of only liquid possible?


Despite having found nebulouses of basically only water/oxygen and planets covered completely of water (Earth for example was very close), the question asks for only water planets.

A planet without a core has no mechanism to heat itself, so in this case the outer layer will be ice, due to outer space is close to absolute zero. And if it is close enough to a sun, so the crust will be liquid, water will be lost in evaporation in big scale.

In addition I doubt gravity will be stable enough to hold the "planet", since, again, there is no core, so there is no "gravity machine". Mostly it depends on how big a planet should be, and external conditions.

So a water only planet seems very hard, but the problem is not the water but the planet word. What is a planet? A rock in space is a planet? Any round object is a planet? To my understanding, to something be a planet needs several mechanism to be in place, the origin of some of them lays in the core. In addition I think the word planet should be subclassified.

Can be created a water planet naturally?

Well, probably "yes". Using the analogy how "first generation" stars are created from a field of hydrogen, change hydrogen with water. From here can happen two things it grows bigger and it becomes a star of water with a fusion core (so no more only water), or it stays small with no really water** layers, and any gravity field will disrupt it easily.

So you need really fine tuning to maintain it stable. For example, having an orbit seems to be one of the needs of being a planet, how are you going to accomplish it without electromagnetic energy in a natural manner?

** some people are calling it: no no sh*t water


if it's only water, wouldn't that turn into a star? I mean, if there is some energy from another star which splits the water into its pieces, 2 H2 O <> 2 H2 + 1 O2 (chemically, I don't know how to format this, Hydrogen and Oxygen) and then add energy, wouldn't result that into a fusion reactor like our sun? which combines the H2 into He and so on.... or am I completely wrong

  • $\begingroup$ The second point. Water electrolysis needs a lot of energy. With the same reason, our oceans should start to boil... $\endgroup$ – bilbo_pingouin Sep 29 '15 at 19:26
  • 1
    $\begingroup$ Welcome to the site Schwertspize, this doesn't really address the question being asked, once you have earned more reputation you will be able to leave comments on questions and answers (full list of privileges: worldbuilding.stackexchange.com/help/privileges) $\endgroup$ – James Sep 29 '15 at 19:33

Tidal heating and salt would make a suitable planet/moon of your choice. Imagine a large belt consists of many icy bodies and planetoids: the planetoids are made of rocks and ice, and if they goes near the sun they evaporates, leaving behind a rocky core, while the icy crust and water boils off and condenses on the outer bodies, distilling the water to the outer reaches of the solar system. Over time, you would end up with some rocks near the sun/star, and a ring of snow outside the freezing line. Snow like this is not stable, and tend to clump together, forming snowballs, or planetoids made of pure ice.

Given time, the snowballs grow large enough to become a planet/planets, then planetary migration can send them to the inner skirts of the solar system.

Let’s say that one gets captured by a gas giant: gas giants have a strong magnetic field, if the planet stays inside, it can keep an atmosphere, just like titan. When small bodies get captured they also tend to be in an ecliptical orbit, which dissipates energy by tidal heating. If left unchecked, tidal heat can make the inside of the planet a hot liquid of almost uniform density, ranging in 1.33 to 1.6/2.6, while blocking mist ices from forming. Interaction with the gas giant’s magnetosphere would ionize the planet’s water content, forming an atmosphere of oxygen, protected by the magnetic field of the gas giant within a plasma torus. While the planet itself is conductive with salts dissolved in the water, coupled by the internal heating this would effectively create a magnetic field of it’s own. This safeguards the atmosphere and therefore the liquid water on the surface of the planet even further, allowing habitation by humans(or story equivalent). The planet will likely have low gravity, however, as long as the escape velocity if on the order of 2000 to 3000 m/s, similar to Ganymede’s or Europa’s, the atmosphere won’t escape fast, and there is plenty of planetary material beneath to escape.

Such a planet can have a flux tube powered by it’s interaction with the gas giant’s magnetic field, forming some rather spectacular auroral display.

Alternatively, if the planet formed hot, it can stand on itself orbiting the star, powering internal convection and thus a magnetic field from crystallization of ices within the center of the planet, shedding less dense ammonia in the process to drive a dynamo in an electrolytic environment. This would also retain an atmosphere of pure oxygen, making it inhabitable by humans or story equivalent.

Water won’t be gone in space as quickly as many people thinks, after all.


Real answers to serious questions in Physics and Cosmology are a tad complicated and require at least a basic understanding of the principles of cause and effect, a rudimentary knowledge of Chemistry and just a bit about the Electromagnetic Field.

I am afraid that All Ten answers are speculatively interesting but essentially incorrect.

In all of Physics, there are only Four known forces in the universe; the strong nuclear force, the weak nuclear force, gravitational force, and the electromagnetic force. Since the first three forces can be, in this instance, entirely ruled out, and in space with the absence of planetary chemistry, with a very limited set of leggo blocks to work with, it is plainly evident that the only remaining operating principle available, that might provide a path to a solution, is the electromagnetic force.

In order to have water, one must have both Hydrogen and Oxygen OR their several ionic constituents as well as an ordered system with which to assemble the constituents into atoms and molecules.

All Stars provide the basic constituents for the production of Hydrogen on planetary bodies (free Electrons and Protons) as well as atomic Oxygen ... ions, all with an electric charge. They streak radially outward from the Sun and from all other stars on a continuous basis, creating (by definition) an electric current in space. Hydrogen and Oxygen are the only atoms required for the production of Water. The solar wind is IN FACT continuous, only varying in intensity over time.

Obviously all stars have Hydrogen but the problem is getting hydrogen from the star to the planet for the production of water. It isn't possible to transport Hydrogen from the star directly since Hydrogen and just about everything else is striped of electrons and turned into ions. The ions must be transported, captured and reassembled by some natural system into molecular Hydrogen and Oxygen at the planet.

Additionally, elemental compounds containing Oxygen are also produced in every supernova. This old "Star Stuff" is the stuff that most planets are made of, which includes a large family of elemental oxides.

The primary requirement for the ongoing production of Planetary Water (aside from a planet) is a Magnetic field, either primordial or global. Planetary Water cannot be produced or maintained in the universe without one.

Generally speaking, the production of a Planetary Magnetic Field REQUIRES an Iron bearing core with SPIN. Fortunately, nearly all rocky planets have them. Planetary magnetic fields are only produced:

  1. When the planetary body has enough Iron AND is massive enough to melt the Iron/Nickel in order to form a laminated central core consisting of a solid center of heavier metals, with the molten Iron/Nickel Alloy "floating" between IT and the lighter materials above.

    ( This is a "Basic" but accurate general statement of fact; a simple version of a reality that only becomes really complicated when considering the nearly unending string of variables like the percentages of the various elements in the body of a "typical " planetary body, the type of star contributing to the makeup of the nebula in which the new star and it's planets formed ... etc., etc. BUT it is sufficiently accurate for our purposes. )

The idea that the heavier metals are locked in compounds in the lighter materials is incorrect. Under the heat and pressure of planetary formation and particularly with the presence of stellar carbon, most of the heavier metallic Oxides, Sulphides and similar compounds are reduced to relatively pure metals which form a somewhat layered structure in the vicinity of the core. Heat, pressure and the lack of reactive elements like Oxygen, Chlorine, etc., keep them in that state but do not necessarily prevent them from alloying to some degree at their respective interfaces. The idea of separation of the metals in the core by crystallization is just a theory and, if it exists at all may only apply to some metals or may only be applicable at their interfaces, thereby isolating one layer from another and preventing them from forming a large, complex single-alloy core. Whether the "State" of the core itself is solid, liquid or some other exotic state is unknown.

What IS known is that an electrically conductive, mobile, liquid layer exists above and isolated from the core and that the electric circuit of which it is a part, forms the dynamo resulting in the global magnetic field. We deduce that said layer will be metallic and Iron-bearing.

  1. When the planet has a rate of spin sufficient to cause a convective rotation of the molten Iron Alloy layer ... resulting in a circular electric current. The planet's orbit through the remnant magnetic field of the Star itself can be sufficient to trigger the development of a primordial field, due to a polar electric current, with further heating of the core, but the primary magnetic field source is the aforementioned convection of the conductive Iron/Nickel layer present in most "Rocky" planets. Once established, the circular current establishes the permanent planetary magnetic field ... and if the field is of sufficient strength, the production of Planetary Water can continue. Contrary to popular belief and within limits, The stronger the field, the greater the relative rate of production of water.

    (Should a water bearing planet lose it's global magnetic field, the production of water will stop and it will lose it's water and it's atmosphere to space through several separate processes. Mars is indicative and a convenient case in point; We know now that it had water and a global field in its early life. We also know now that it lost its water and its atmosphere after the failure of the global field.)

  2. When the planet has, during it's formation, acquired even the minutest trace of water ice from asteroid and cometary impacts. This trace moisture is boiled out of the material of the hot planet and becomes a constituent of the forming primordial atmosphere, combining through a ladder of reactions with various lighter crustal elements to produce oxygen bearing compounds that litter the planetary surface. These compounds are important because under bombardment of ionized particles from the star, they release their Oxygen, which contributes to the acquisition of a great deal of water quite early in the planet's life.

    (Once the Global Magnetic Field is fully established, the ongoing rate of water production slows down considerably, ideally reaching an equilibrium. As the field strength waxes and wanes over the life of the planet, so does the amount of water on the surface.)

    For these reasons, it is NOT generally possible to have a water planet without at least a small Iron Core

However, ANY rotating planetary body can also acquire water provided it moves in a circular orbit through a strong enough magnetic field in a stellar environment, has a molten or otherwise conductive convective core or layer or becomes, through induction, part of an electric current path in association with the object around which it orbits, giving it an induced field.

Both Saturn and Jupiter have very strong global magnetic fields, within which are trapped the ionic constituents required for the production of water ... and each has a moon ( Enceladus and Europa ) comprised largely of water acquired as a result of it's orbit through those trapped constituents. Europa will be found to have a small Iron core and Enceladus a subsurface conductive layer, probably saltwater ... both with induction fields and both subject to runaway water acquisition since their oceans are frozen and cannot easily vaporize into space.

On a smaller scale, with rocky planets in the Goldilocks zone of their respective stars and an expectation of liquid water, the magnetic field is important because:

  1. Charged particles from the star are largely deflected around the planet, preventing a catastrophic accumulation of water like what happened on Enceladus, Europa and the Gas Giants.

  2. There are several processes and leakage pathways, particularly near the polar regions, that allow Electrons, Protons and atomic Oxygen from the star to spiral in along the magnetic field lines, colliding with each other to form Hydrogen atoms and Oxygen doublets which along with any free Oxygen in the atmosphere, combine to produce water ice on a slow but continuous basis that varies with the sunspot and CME activity on the star.

    The reason for the magnetic field is the need to capture the ions (Electrons, Protons & Atomic oxygen) and then REASSEMBLE them in the upper atmosphere. The reassembly takes energy and is driven by the electromagnetic force.

Any moving charge is, by definition, an electric current. When the charged particles meet the Earth's magnetic field, they are attracted to it or provisionally "captured" by it. Most of them follow the surface of the "teardrop" shape and exit the system at high velocity into space behind the planet. At certain angles, however, the particles are truly captured and spiral in toward the surface along the field lines. They spiral in because they are charge carriers moving in a magnetic field which causes them to rotate in opposite directions around the same field lines. Their opposite charges AND their extra impact energy (when they collide) provides the required energy for molecular reassembly. There are a couple of other process that help in the acquisition of ions but they are too complex to address here.

Granted, there will be exceptions for stars off the main sequence like Magnetars and Neutron Stars, BUT this means that, with few other exceptions, virtually ALL rocky planets, about our size and bigger, in the Goldilocks zones of their respective stars, and particularly Class-G stars ... HAVE LIQUID WATER. The probability of ubiquitous Life in the universe ... is ubiquitous.

We are actually watching the planetary production of water, every time we see an Aurora Borealis ... and even when we don't see it.

Even though the oceans appear to contain an incredible amount of water, in relation to the 8,000 mile diameter of the planet, at about 1.5 miles deep, it is only the finest trace of surface moisture ... hardly anything at all really.

  • 2
    $\begingroup$ Please don't call all other answers essentially incorrect, if you end up giving the same result ("small iron core needed") than some of them using different reasoning. Also you do not give an actual reason why a magnetic field is needed for water. Just few examples of how it regulates water accumulation for Earth. Or any reason why there would be only one mechanism to generate a magnetic field. Explaining the current theory on how magnetic field is generated for Earth doesn't really exclude any other mechanisms. $\endgroup$ – Ville Niemi Dec 7 '14 at 9:57
  • 1
    $\begingroup$ You didn't read the entire post ... or you didn't understand it. I said they were incorrect because they are IN FACT, given the actual laws of Physics, Incorrect. $\endgroup$ – TemplarKnight Dec 7 '14 at 10:12
  • 1
    $\begingroup$ I have a ot of problems with many of the specific statements in your A. Can you give a citation and state the quantity of solar wind material pulled onto the planet due to the magnetic field? How much oxygen is present in the outgassing of a main sequence star? A quasar is not a star. The solid core contains Fe, Ni, Cobalt, and expells heavier elements as it crystallizes. The liquid phase has more different elements, not fewer. Most of the superheavy elements are not settled toward the core but are chemically bound in the mantle. Europa etc. Got all the water this way?? No.. and no need. $\endgroup$ – JDługosz Dec 7 '14 at 10:28
  • 3
    $\begingroup$ Not to heap on, but there are a bunch of inaccuracies here (and so stop trying to lecture people about what "knowledge" is required when you get things wrong!). 1) Stars have hydrogen from birth; they do not "produce" it. 2) The solar wind is composed of charged particles, but it is not continuous, and is not an electric current. 3) The primary requirement for liquid water is that the temperature be between 0 and 100 degrees C (32 and 212 degrees F). Magnetic fields have nothing to do with it. 4) How the heck is water "produced" on a planet, and why does it require a magnetic field? $\endgroup$ – HDE 226868 Dec 7 '14 at 20:14
  • 2
    $\begingroup$ I don't beleive you said just a theory! We know from skepics' forums that this is code for I don't know what words mean. You dismiss the accepted scientific consensus as being unproven, while touting your own highly speculative fringe ideas as "FACT"? Just a note to those who find this thread later: pseudoscience. It might make a good story device, as is the main reason for this thread, but ignore the delusions of competence. $\endgroup$ – JDługosz Dec 7 '14 at 22:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.