We assumed water had to be delivered to planets by comets or asteroids — but a 2025 Nature study found that sub-Neptunes, the most common type of planet in the galaxy, can forge their own oceans by reacting hydrogen atmospheres with molten rock deep inside their cores

For a long time the story of how a planet gets its water had one main plot. Water forms easily in the cold outer reaches of a young star system, freezes into the ice that coats comets and distant asteroids, and is then delivered inward when some of those bodies crash into a rocky planet closer to the star. On that account, a world with water was a world that got lucky with its deliveries.

A 2025 study in the journal Nature offers a different route, at least for one very common kind of planet. It finds that sub-Neptunes can make their own water, deep inside themselves, by reacting their thick hydrogen atmospheres with molten rock.

The delivery model, and why it mattered

The older picture is built around the snow line, the distance from a star beyond which it is cold enough for water to freeze. Inside that line, in the warm zone where rocky planets form, water struggles to condense, so a planet built there starts out fairly dry. To end up wet, it has to be topped up from outside, by icy bodies wandering in from beyond the snow line.

This is the framework often used to explain Earth’s oceans, and it carries an implication worth noticing: water on any given planet is something of an accident of delivery, dependent on where the icy debris happened to go. If that were the only way, water worlds would be a matter of luck.

What the new study found

The work, led by Dan Shim at Arizona State University and Alona Vazan at the Open University of Israel, comes at the question from the inside of the planet rather than the outside.

Sub-Neptunes are bodies larger than Earth but smaller than Neptune, and many of them are thought to hold a deep magma ocean, a layer of molten rock, beneath a thick, hydrogen-rich atmosphere. At the base of that atmosphere the pressure and heat are enormous. The researchers recreated those conditions in the laboratory and watched what hydrogen does when it is pressed against molten silicate rock under that kind of squeeze.

It reacts. The hydrogen pulls oxygen out of the silicate melt, and the freed oxygen bonds with hydrogen to make H2O. The amount produced was not a trace. In their results the reaction could turn a meaningful fraction of the planet’s material into water, up to a few tens of per cent by weight, far more than earlier estimates had suggested. The planet, in other words, manufactures water out of ingredients it already has, using its own atmosphere as one of the reactants.

Why it matters

The significance comes from how common these planets are. Sub-Neptunes, along with their close cousins the super-Earths, are the most abundant type of planet found so far in the galaxy, far outnumbering worlds like our own. If making water internally is a normal part of how such planets form, then water is not a rare prize handed out by chance collisions. It may be a routine by-product of building a very ordinary planet.

That reframes a question that sits underneath the search for life. Instead of asking which planets were lucky enough to receive water, the finding suggests water could be widespread, produced in place across a huge population of worlds without any help from comets at all.

Where the water actually ends up

Here the picture needs care, because more water inside a planet is not the same as an ocean on top of one.

The water in this process is forged deep down, mixed into the magma and the dense atmosphere under crushing pressure, not poured onto a surface. Whether it stays locked in the interior or works its way upward to form a distinct watery layer is a separate question, and it depends on details the experiments can only partly capture. Related work has argued that much of this internally made water may remain dissolved in the deep interior, which would mean a planet could be genuinely water-rich by mass and still not have a habitable surface sea.

That tension runs straight into the current argument over planets like K2-18b, where researchers disagree about whether a sub-Neptune is a true ocean world or a gas-dominated one with its water hidden away. The new result does not settle that. It changes the starting assumption, by showing that the water need not have been imported, while leaving open the harder question of what form it takes once it is there.

What comes next

The study is experimental and theoretical, a recreation of interior conditions rather than a direct look inside a real planet, so it carries the usual caution that comes with extrapolating laboratory results to whole worlds. The next step is to fold this chemistry into models of how sub-Neptunes form and evolve, and to test those models against what telescopes actually measure in the atmospheres of these planets.

If it holds up, the shift is a quiet but real one. The question stops being where a planet found its water and becomes how much water a planet inevitably makes, and where that water goes. For the most common planets in the galaxy, the ocean may turn out to be less a matter of delivery than of chemistry.