How Did Earth’s Oceans Form and Where Did the Water Come From?

Earth’s oceans were not delivered in a single dramatic event but accumulated through overlapping processes that began before the planet finished forming. Most of the water arrived locked inside the rocky materials that built Earth in the first place, with smaller contributions from the primordial gas cloud surrounding the young Sun and from icy bodies in the outer solar system. The story is more complicated than the old textbook image of comets raining water onto a bone-dry world, and researchers are still working out exactly how much each source contributed.

The Rocks That Built Earth Were Already Wet

For decades, scientists assumed that because Earth formed close to the Sun, its building blocks must have been too hot to hold onto water. Any water on the surface, by this logic, had to arrive later from wetter regions of the solar system. That picture has shifted dramatically. Measurements of enstatite chondrite meteorites, which have an isotopic makeup closely matching Earth’s rocks across multiple elements, show they contain far more hydrogen than anyone expected. A 2020 study in Science found that enstatite chondrites carry enough hydrogen to have supplied at least three times the mass of water currently in Earth’s oceans, and their hydrogen and nitrogen isotopic signatures match Earth’s mantle.¹ A 2025 follow-up confirmed that despite their dry-looking mineral makeup, enstatite chondrites hold enough hydrogen to readily explain Earth’s present-day water abundance.²

Carbonaceous chondrites, a different class of meteorite from the outer solar system, have long been the default suspect for delivering Earth’s water. Isotopic measurements of the asteroid Vesta, which samples one of the oldest hydrogen reservoirs in the solar system, show a hydrogen isotopic composition matching carbonaceous chondrites. Importantly, those ratios also match Earth’s, and the old ages of the meteorites involved suggest volatiles could have been incorporated into Earth early in its assembly rather than arriving in a late splash.³

More recent modeling that tracks how nitrogen, hydrogen, sulfur, and carbon were distributed during Earth’s formation finds that the earliest enstatite chondrite-like impactors supplied roughly 45% of Earth’s hydrogen, while later, more oxidized impactors contributed another 46%. The small remainder came from carbonaceous chondrite-like material.⁴ The picture that emerges is not one dominant water source but a mix, with the proportions changing as Earth grew.

A Dose of Hydrogen from the Solar Nebula

The Sun formed inside a cloud of gas and dust, and that cloud lingered for several million years before dispersing. If Earth had grown large enough while the nebula was still around, it would have gravitationally attracted a thick hydrogen-rich atmosphere. Some of that hydrogen could have dissolved into the molten rock of the early Earth’s surface and interior, a process called nebular ingassing. One modeling study estimated that Earth needed roughly seven to eight oceans’ worth of water from chondritic sources supplemented by up to about half an ocean’s worth of nebular hydrogen.⁵

Why does the nebular contribution matter if it was relatively small? Because nebular hydrogen has a distinctive isotopic fingerprint: a very low ratio of deuterium (heavy hydrogen) to ordinary hydrogen. Even a small amount of this light hydrogen, mixed in with the heavier signature from chondritic and cometary sources, helps explain the specific isotopic ratio observed in Earth’s water today. Without some nebular input, the numbers do not quite add up.⁶

How Isotopes Fingerprint the Water’s Origin

Scientists distinguish between water sources using the ratio of deuterium to hydrogen in water molecules. The Sun’s original gas cloud had a very low ratio. Comets from the outer solar system tend to have a high ratio. Earth’s water falls somewhere in between, which immediately tells researchers that no single source dominates completely.

Quantum chemical calculations show that when hydrogen-bearing gas condenses onto mineral surfaces and forms hydrated minerals, deuterium preferentially ends up in the solid phase. The enrichment increases at lower temperatures. At temperatures around 175 to 300 degrees Kelvin, the calculated enrichment matches what is actually seen in terrestrial water, carbonaceous chondrites, Mars, and asteroid Vesta.⁷ This means much of the deuterium enrichment happened naturally as the solar system cooled, not necessarily because water was delivered from some exotic far-flung source.

Noble gases offer another line of evidence. Helium, neon, and argon isotopes in Earth’s mantle point toward solar wind implantation into the tiny grains that eventually clumped together to form our planet, during the first few million years of the solar system.⁸ The atmosphere’s noble gas composition, meanwhile, appears to be a blend of about two-thirds chondritic material and one-third of that solar-wind-implanted signature.

What About Comets?

The romantic image of icy comets filling Earth’s oceans persists in popular science, but the evidence increasingly pushes comets into a minor supporting role. The Rosetta spacecraft, which studied comet 67P/Churyumov-Gerasimenko up close, measured the isotopic composition of krypton and xenon in the comet’s gas. Those measurements showed that cometary noble gases were mixed with chondritic sources, not solar ones, to produce Earth’s atmospheric composition. The cometary contribution to atmospheric krypton came out to roughly 21%, with xenon showing a similar fraction of about 22%.⁹

That means comets did deliver some volatiles to Earth, but they were not the dominant supplier. Their most distinctive contribution may have been to the atmosphere’s noble gas inventory rather than to the oceans themselves. The heavy lifting, in terms of hydrogen and water, was done by the rocky materials that built Earth’s bulk.

When Did the Oceans First Appear?

One of the most surprising findings of the past two decades is just how early liquid water existed on Earth’s surface. The planet finished forming around 4.5 billion years ago, and for a long time scientists assumed it took hundreds of millions of years for the surface to cool enough for water to condense. Tiny crystals of the mineral zircon from the Jack Hills region in Western Australia overturned that assumption.

The oldest known terrestrial zircon, dated at about 4.4 billion years old, contains oxygen isotope signatures and micro-inclusions consistent with formation from a granitic melt that had interacted with liquid water at or near the surface.¹⁰ Other Jack Hills zircons, aged around 4.3 billion years, tell the same story: their oxygen isotope values point to a liquid hydrosphere interacting with the crust by that time.¹¹ Lithium isotope data from the same zircons add further support, suggesting not just that water was present but that it was weathering rock extensively, much as modern oceans do.¹²

So within about 150 to 250 million years of Earth’s formation, something recognizable as an ocean was already there. That is remarkably fast, and it implies that water was not a latecomer that dribbled in over eons. It was available almost from the start, locked in the planet’s building materials and released as those materials melted and degassed.

Keeping Water Liquid Under a Faint Sun

Here is the puzzle: the young Sun was roughly 25% dimmer than it is today. If Earth’s atmosphere had been the same as the modern one, models predict the entire surface should have been frozen for the first two billion years.¹³ Yet the zircon evidence confirms liquid water within a few hundred million years, and there is ample geological evidence for oceans throughout the Archean eon. Something kept things warm.

The leading explanation is a much thicker blanket of greenhouse gases, particularly carbon dioxide and possibly methane. With no plants to draw down CO₂ and limited weathering early on, atmospheric carbon dioxide levels could have been vastly higher than today. Modeling work supports this: a high-CO₂ greenhouse atmosphere efficiently counterbalanced the faint young Sun before about three billion years ago.¹⁴ The solution is elegant. Volcanoes belched out CO₂ relentlessly, and without efficient weathering to scrub it from the air, the greenhouse effect could compensate for the weaker sunshine.

Water Hidden Deep Inside the Planet

Earth’s oceans are only part of the planet’s water budget. A vast reservoir sits hundreds of kilometers below the surface, locked inside the crystal structures of high-pressure minerals in the mantle. The transition zone, between about 410 and 660 kilometers deep, is especially important because the minerals wadsleyite and ringwoodite can incorporate water into their structure at up to around 2.5% of their weight.¹⁵ That does not sound like much per grain, but the transition zone is enormous. If it were saturated, it could hold multiple oceans’ worth of water.

The clinching evidence came from a diamond found in Brazil that had a tiny inclusion of ringwoodite inside it, brought up from transition-zone depths. That ringwoodite contained water, confirming that the transition zone is not just theoretically wet but actually contains hydrous minerals.¹⁶

This deep water is not sitting as a subterranean ocean. It exists as individual hydrogen and oxygen atoms bonded into the mineral lattice, invisible and inaccessible to us but chemically real. The exchange between this deep reservoir and the surface oceans happens through volcanism and subduction over geological time. Mid-ocean ridges and volcanoes above subduction zones return water to the surface, while subducting tectonic plates carry ocean water back down into the mantle.¹⁷ Most of the water dissolved in the primordial magma ocean was entrapped in the solidifying mantle and then massively released by volcanism during the Hadean and Archean eons, forming the early oceans.¹⁸

This deep water cycle means sea level is not fixed by the total amount of water on the surface. Over hundreds of millions of years, changes in tectonic rates can shift water between the surface and the interior, causing sea levels to rise or fall independent of ice sheets.¹⁹

Surviving the Moon-Forming Impact

About 4.5 billion years ago, a Mars-sized body, often called Theia, slammed into the proto-Earth and produced enough debris to form the Moon. An event that violent should have boiled away any surface water and blasted much of the atmosphere into space. So did it reset the clock on Earth’s water?

Analysis of lunar highland rocks that crystallized about 4.35 billion years ago shows that both Theia and the proto-Earth were already strongly depleted in volatile elements compared to primitive meteorites before the collision happened. The volatile depletion seen in the Moon did not arise from the giant impact itself; it was inherited from the precursors.²⁰ In other words, both bodies had already lost their most easily escaping volatiles before they met. The water that remained was the more tightly bound fraction, stored in minerals deep inside the planet rather than sitting on the surface where it could be stripped away.

The bombardment that followed the Moon’s formation, a prolonged period of asteroid and comet impacts, may have actually replenished some of what was lost. Modeling of atmospheric erosion and replenishment during heavy bombardment suggests that depending on the composition of the impactors, Earth could have accumulated a more massive atmosphere of several bars of surface pressure, mostly as carbon monoxide and carbon dioxide.²¹ The net effect of the late bombardment was not purely destructive. Impacts simultaneously stripped volatiles and delivered new ones, and for Earth, the delivery side seems to have won.

Why Venus and Mars Ended Up Dry

Earth is not the only rocky planet that started out with water. Venus likely began with a water endowment similar to Earth’s, and early in its history it may even have had oceans. But Venus sits closer to the Sun, and that made all the difference. Water vapor rose high into the atmosphere, where ultraviolet light split it apart, and the freed hydrogen escaped to space. This runaway process drained Venus of its water on a geologically short timescale.²² Modeling suggests that even under the most favorable cloud conditions, Venus’s surface temperature would have been at least 80 to 100 degrees Celsius, and water vapor would have been a major atmospheric component at all altitudes, guaranteeing rapid hydrogen loss.²³

Mars had the opposite problem. Its small size meant a weaker gravitational pull and no sustained magnetic field to shield its atmosphere from the solar wind. Over billions of years, Mars lost most of its atmosphere and, with it, the surface pressure needed to keep water liquid. Some water is locked in polar ice and subsurface deposits, but the thick oceans that may have covered parts of early Mars are long gone. Earth threaded the needle: large enough to hold its atmosphere, far enough from the Sun to avoid a runaway greenhouse, and geologically active enough to recycle volatiles through volcanism and plate tectonics.

Watching Water Move in Other Star Systems

For the first time, astronomers can now observe the water delivery process as it happens around other young stars. The James Webb Space Telescope has examined protoplanetary disks, the swirling disks of gas and dust where planets form, and detected water vapor in the inner regions where rocky planets are expected to assemble. In compact disks, JWST found excess emission from cool water vapor near the snow line, the boundary where water transitions between ice and gas. The observations are consistent with icy pebbles drifting inward from the outer disk and releasing their water as they cross the snow line.²⁴

This process, icy material migrating inward and delivering water to the warm inner zone where rocky planets form, is essentially what scientists have long proposed happened in our own solar system. Seeing it play out around other stars validates the general framework and suggests that water-rich rocky planets may not be unusual. JWST’s improved spectral resolution can distinguish differences in water abundance by a factor of five, meaning it can tell the difference between a very wet inner disk and a dry one.²⁵ As more disks are surveyed, the data should reveal whether our solar system’s water delivery was typical or whether Earth got lucky.

Water Worlds Beyond Earth

The question of where water comes from extends well beyond the rocky inner planets. Several moons orbiting Jupiter and Saturn, including Europa and Enceladus, harbor liquid water oceans beneath thick shells of ice. These oceans are not kept liquid by sunlight. Instead, tidal forces generated by the gravitational pull of the giant planets flex the moons during their orbits, generating internal heat. Research has shown that tidal dissipation occurs not just in the solid rock and ice of these moons but also directly in their liquid oceans, where a previously overlooked tidal force related to the moons’ axial tilt can resonantly excite large-scale waves and contribute to ocean heating.²⁶

These icy moons formed in the outer solar system where water ice was abundant, so their water origin story is simpler than Earth’s: they formed from ice-rich material in the first place. The fact that liquid water persists beneath their surfaces billions of years later, kept warm by a completely different energy source than the one that sustains Earth’s oceans, broadens the range of environments where water can exist in liquid form. It is a reminder that while Earth’s ocean formation was a complex, multi-source story involving rocky meteorites, nebular gas, volcanism, and plate tectonics, water itself is one of the most common molecules in the universe. Getting it to a planet’s surface and keeping it there as a liquid is the hard part.