Unexpectedly, many chondrites, of every type, contain secondary minerals that formed as the primary mineral reacted with liquid water. The reactions took place soon after the formation of the chondrules and often during accretion of the incorporating bodies (Doyle et al. 2015). In some cases the minerals hydrated even before accretion. Water molecules suspended in space were wetting the grain surfaces and making them stickier, accelerating the process by which the grains coalesced. Carbonaceous chondrites tend to be particularly water-rich, possibly because they formed beyond the ‘snow line’ within which the infant Sun’s heat would have prevented the volatile from condensing.
Hydrated minerals have also been detected in in comets. The presence of liquid water disproves the idea that comets, which contain copious amounts of frozen water, never got warm enough to melt their icy interiors (Berger et al. 2011). In the case of the asteroid Ryugu the hydration was dated to within the first 1.8 Ma of solar system history (McCain et al. 2023). Ceres, the only asteroid large enough to be rounded by its own gravity, is essentially a stationary comet, with 50% of its volume estimated to be ice (27% by mass). Occasionally cryovolcanoes erupt salty water from its interior.
Interplanetary space was wet. Evidence for this doesn’t just come from asteroids and comets. All the terrestrial planets show signs of having once been drenched by water – even Mercury. One of the most astonishing findings of the Messenger mission to Mercury was that under the beshadowed walls of its high-latitude craters water abounds. Deposits 50 m thick are thought to lie beneath the rims.
No less surprising is the evidence from Venus. It is so hostile to life one might almost think it exists purposely to demonstrate that an Earth-like planet is not an inevitability. But although Venus today is dry, 465° C at the surface day and night, and shrouded under clouds of carbon dioxide and sulphuric acid, the high ratio of deuterium to hydrogen in its atmosphere suggests that it once hosted a substantial ocean, subsequently evaporated or blasted away. Deuterium is an isotope of hydrogen, heavier than the ordinary form because the nucleus includes a neutron. It combines with oxygen to produce a heavy form of water. The idea is that when ultraviolet radiation from the Sun split the evaporated water into hydrogen, deuterium and oxygen, the lightest gas, hydrogen, and most of the deuterium escaped into space. Some of the deuterium remained in the atmosphere, while the heavier oxygen oxidised the crust. Another mechanism of hydrogen loss would be through its dissociation from atmospheric HCO.
Until quite recently, the Moon was believed to be devoid of water. Then in October 2009 the LCROSS mission (Lunar Crater Observation and Sensing Satellite) discovered significant quantities of water on crashing part of the spacecraft into a crater close to the permanently shadowed south pole. Five months after that, it was announced that millions of tons of ice lay hidden deep within craters around the north pole. However, because of the difficulty of understanding where it might have come from, scientists continued to doubt that water was present. Hence it was still newsworthy when in 2018 a paper analysing data gathered by India’s Chandrayaan-1 probe gave the first definitive proof of surface ice. The presence of water can be doubted no longer. Some of the water is likely to be a product of the solar wind (Xu et al. 2022); that at the poles is undoubtedly ancient.
Significant amounts have even been found in fragments of rock from beneath the surface, for example, in olivine crystals that grew while the containing volcanic melt was as yet unerupted (Hauri et al. 2011) and in apatite crystals that formed 4.4 Ga ago when the crust was pounded by asteroids (Tartèse et al. 2014). Analysis of the younger basalts recovered by China’s Chang’e-5 mission suggests that by 2.0 Ga the interior was essentially dry (Hu et al. 2021).
Water covers most of the Earth’s surface, to an average depth of almost 4 km. According to the nebula hypothesis, Earth should not have had oceans to start with, since it orbited within the snow line. Rather, it should have been covered by a global ‘magma ocean’, driving off whatever water might have been delivered by early-forming comets. Yet water has been abundant on or in the Earth from as far back as datable minerals can take us, in geological time as early as 4.4 Ga ago. At the beginning of the Archaean 4.0 Ga ago, the entire planet is thought to have been under water.
Mars’s early history is also puzzling. Today its surface is cold, but like the other rocky planets it has an igneous crust, entailing that the surface once was molten. Evidence from a Martian meteorite suggests the crust formed surprisingly early: ‘no later than 4547 Ma’ (Bouvier et al. 2018) or 15 Ma after the formation of the planet itself, based on modelling. Not much more than 70 Ma after that, Mars was struck by asteroids. As the evidence comes from only one meteorite, we cannot categorically say that this was for the first time, but it is unlikely to have been an isolated occurrence. Mars’s surface is peppered with ancient craters.
Wherever one looks there is evidence of former water. The depression in its northern hemisphere once contained an ocean over 400 metres deep, covering a third of Mars’s surface. Formerly water-conducting deltas and valley networks fringe the basin. Within it smaller craters are visible, their walls eroded by the ocean. In other regions, splashes of sediment around the craters show that the ground already had a mud-like consistency: the surface was wet when asteroids bombarded the planet. Thereafter, clouds continued to rain on the lowlands, repeating their cycles of evaporation, re-precipitation and runoff as the water gradually seeped into the ground.
Approximately 1 bar of CO2 atmosphere is needed to maintain liquid water on the surface (Palumbo et al. 2018). High levels of CO2 could have built up as a result of high rates of volcanic outgassing. Today the planet is colder, has only a 7 mbar atmosphere (nearly all CO2), and what remains of the water is locked up, liquid as well as frozen, beneath the surface and in large permanent ice caps around the poles.
Jupiter mostly consists of hydrogen and helium. The helium cannot be primordial and is best understood as a product of in-situ nuclear fusion, though the process is certainly not going on now. The atmosphere also contains small amounts of oxygen, combining with hydrogen to form H2O. At the equator, water makes up just 0.25% of the molecules in Jupiter’s atmosphere. As with the other giant planets, the atmosphere also contains substantial amounts of ammonia (NH3) and methane.
Europa, one of the four moons that Galileo saw with his telescope, has a rocky interior with an icy shell; below the ice, an ocean of salty water is inferred. Charged particles streaming from Jupiter are continually splitting the frozen water into oxygen and hydrogen gas. Ganymede, the largest moon, is 50% water. Like Vesta, its rocky interior is differentiated, with a metallic core. Callisto’s low density suggests that it too is 50% water. Io, the innermost and densest of the four moons, is partly molten and has almost no water.
Saturn is best known for its rings, which mostly consist of ice particles. One of the rings is fed by vapour ejected from the moon Enceladus, which hides an ocean of water beneath its icy surface. Because they are undarkened by interplanetary dust, astronomers conclude that they must be relatively young. How they originated is not known. A collision between two icy moons is one possibility. Jupiter, Uranus and Neptune also have rings, but faint, unspectacular ones that evoke no wonder.
An ocean of water has also been discovered 20-30 km beneath the surface of Mimas, the smallest of Saturn’s regular moons. The surface is heavily cratered. Despite that, the ocean is inferred to be remarkably young, less than 15 million years (Lainey et al 2024). Titan is the second largest moon in the solar system after Ganymede and has a water-ice crust with lakes and rivers of liquid methane and ethane (C2H6). Liquid water lies beneath its surface, beneath that probably a mantle of frozen water and a rocky core.
Uranus is thought to be primarily a mixture of water and rock, with the water fraction anything from 33 ± 11% to 70 ± 17% (Morf et al. 2024). Uranus’s moons are predominantly rocky. The four largest may have oceans beneath their icy crusts, with dissolved ammonia and salts acting as antifreeze.
Because of its pale greenish-blue colour (not ultramarine as in some images) Neptune was named after the Roman god of the Ocean, and its moons after minor water deities. The colour comes from the absorption of red and infrared light by methane in the atmosphere, which makes up about 10% of the bulk and is primarily hydrogen and helium. Though similar in size, Neptune is 30% denser than Uranus, so probably contains more rock. Little is known about the moons’ composition, except for Triton.
the fragmentary nature of the TNOs – more than 100,000 objects over 50 km in size and trillions of objects 10–100 metres in size (Cooray 2006);- the low total mass of the Kuiper Belt objects (more easily determined than that of the Scattered Disc);
- the ‘surprisingly high level of dynamical excitation’ of the objects – they have highly elliptical orbits at various angles to the ecliptic;
- the largest objects appear to be rocky.
The largest Kuiper Belt Objects (KBOs) are Pluto, Haumea and Makemake, all classified as dwarf planets. They are around 70% rock (Eris more like 90%), with a substantial ice component. Some of the smaller KBOs may be mainly ice. As with the asteroid belt, the vast number of these objects is thought to reflect collisions between larger bodies. Consequently the present state of the belt does not reflect its primeval state, and its more recent history may be one of disaggregation rather than aggregation.
The composition of KBOs large enough to be analysed is inferred from their reflected light. Interaction with polymer-producing cosmic rays and the Sun’s ultraviolet radiation has complicated their chemical history. In simple terms, the surfaces of the largest bodies are mainly water-ice, nitrogen and various carbon compounds. Smaller bodies are not cold or massive enough to have retained significant amounts of volatile ices. Pluto, about one third the size of Earth’s moon by volume, is one third water and two-thirds rock.
The ‘dynamical excitation’ of objects formerly in Pluto’s vicinity is evident from the thousands of impact craters dotting its surface. At least some of the smaller KBOs are fragments of larger ones and therefore younger than them. Some of the water ice is crystalline and must have formed in temperatures higher than those prevailing now. This may not have been long ago, since cosmic rays will reduce crystalline ice to an amorphous state within 0.1–1.0 million years.
What is true for water is also true for CO2: carbon dioxide ice is ubiquitous through the solar system, from the Kuiper Belt to Mercury’s beshadowed craters (De Prá et al. 2024). An explanation for this is not obvious.
Several pre-scientific peoples believed that a celestial ocean existed above the terrestrial one. The Egyptians visualised the Sun travelling through the heaven in a boat. Babylon’s creation myth, Enuma Elish, related that how Tiamat, a personification of the primordial waters, was split into an upper ocean and a lower ocean. Hebrew tradition maintained that the space containing the sun, moon and planets was the result of God separating the waters.
Can ancient tradition and modern astronomical knowledge be reconciled? The non-mythological Hebrew account suggests a protective, nebulous, circumambient shell not unlike the shape of the postulated Oort cloud. Because of the radiation from the Milky Way’s once luminous nucleus, interstellar space was warmer than today. Over time, the Sun’s gravity caused much of the vapour to diffuse inwards. By 4.57 Ga ago in geological time, interplanetary space may have hosted a substantial volume of water. Water requires surfaces known as a ‘cloud condensation nucleii’ to make the transition from vapour to liquid, and the most plentiful source of such nuclei would have been dust particles from the explosion of the largest rocky planets. Dust would also have come from the extant rocky planets when they were struck by the debris. Beyond Neptune, cooling and electrostatic sticking would have caused the droplets to consolidate into ice particles. The Kuiper Belt and Scattered Disc would be the mixed remains of water from the circumambient shell and rock from the outermost planets.
According to Genesis, the original Earth was destroyed in a terrible watery cataclysm. There were two agents of destruction. One was the springs of the great subterranean deep, which all at once exploded and flooded the planet. The other was the rain that for 40 days fell through ‘apertures of the heaven’. As a consequence of the rain, all life was blotted out.
The rain is problematic. It is difficult to model an atmosphere that might have held that amount of water, and also difficult to see how 40 days of rain, however torrential, could have drowned all the mountains and resulted in life’s being obliterated. Genesis implies that up to that point rain was unknown: the Earth’s surface was watered from below, not above. While some water must have evaporated from lakes and inland seas and from the ground itself, rain seems not to have been a major component of the hydrological cycle. Possibly some water reached the Earth from interplanetary space and lingered in the upper atmosphere. But then, what triggered its collapse?
On the other hand, water might not have been what primarily rained on the planet. In the story about Sodom and Gomorrah the text says that Yahweh rained sulphur and fire on the cities; on another occasion he rained down coals (Ps 11); in the book of Joshua he brings down a hail of ‘large stones’, i.e. meteorites, on Israel’s enemies. Thus the rain might have been a shower of asteroids, accompanied by water as impact-generated caused vapour in the atmosphere to condense and precipitate. Just such a bombardment is known to have pummeled the Earth early in its history, as attested by the giant impact craters on the Moon, just a stone’s throw away in astronomical terms. Asteroids would certainly have obliterated terrestrial life.
