ALEC BRENNER/HARVARD UNIVERSITY
Across the ages, sea levels have risen and fallen with temperatures—but Earth’s total surface water was always assumed to be constant. Now, evidence is mounting that some 3 billion to 4 billion years ago, the planet’s oceans held nearly twice as much water—enough to submerge today’s continents above the peak of Mount Everest. The flood could have primed the engine of plate tectonics and made it more difficult for life to start on land.
Rocks in today’s mantle, the thick layer of rock beneath the crust, are thought to sequester an ocean’s worth of water or more in their mineral structures. But early in Earth’s history, the mantle, warmed by radioactivity, was four times hotter. Recent work using hydraulic presses has shown that many minerals would be unable to hold as much hydrogen and oxygen at mantle temperatures and pressures. “That suggests the water must have been somewhere else,” says Junjie Dong, a graduate student in mineral physics at Harvard University who led a model, based on those lab experiments, that was published today in AGU Advances. “And the most likely reservoir is the surface.”
The paper makes intuitive sense, says Michael Walter, an experimental petrologist at the Carnegie Institution for Science. “It’s a simple idea that could have important implications.”
Two minerals found deep in the mantle store much of its water today: wadsleyite and ringwoodite, high-pressure variants of the volcanic mineral olivine. Rocks rich in those minerals make up 7% of the planet’s mass, and although only 2% of their weight is water today, “a little bit adds up to a lot,” says Steven Jacobsen, an experimental mineralogist at Northwestern University.
Jacobsen and others have created these mantle minerals by squeezing rock powders to tens of thousands of atmospheres and heating them to 1600°C or more. Dong’s team stitched together the experiments to show wadsleyite and ringwoodite hold fractionally less water at higher temperatures. Moreover, the team predicts, as the mantle cooled, these minerals themselves would become more abundant, adding to their ability to soak up water as Earth aged.
The experiments aren’t alone in suggesting a water-bound planet. “There’s pretty clear geological evidence,” too, says Benjamin Johnson, a geochemist at Iowa State University. Titanium concentrations in 4-billion-year-old zircon crystals from Western Australia suggest they formed underwater. And some of the oldest known rocks on Earth, 3-billion-year-old formations in Australia and Greenland, are pillow basalts, bulbous rocks that only form as magma cools underwater.
Work by Johnson and Boswell Wing, a geobiologist at the University of Colorado, Boulder, offers more evidence. Samples from a 3.24-billion-year-old chunk of oceanic crust left on Australia’s mainland were far richer in a heavy oxygen isotope than the present-day oceans. Because water loses this heavy oxygen when rain reacts with the continental crust to form clays, its abundance in the ancient ocean suggests the continents had barely emerged by that point, Johnson and Wing concluded in a 2020 Nature Geoscience study. The finding doesn’t necessarily mean the oceans were larger, Johnson notes, but, “It is easier to have submerged continents if the oceans are bigger.”
Although the larger ocean would have made it harder for the continents to stick their necks out, it could explain why they appear to have been on the move early in Earth’s history, says Rebecca Fischer, an experimental petrologist at Harvard and co-author on the AGU Advances study. Larger oceans could have helped kick off plate tectonics as water penetrated fractures and weakened the crust, creating subduction zones where one slab of crust slipped below another. And once a subducting slab began its dive, the dryer, inherently stronger mantle would have helped bend the slab, ensuring its plunge would continue, says Jun Korenaga, a geophysicist at Yale University. “If you cannot bend plates, you cannot have plate tectonics.”
The evidence for larger oceans challenges scenarios for how life began on Earth, says Thomas Carell, a biochemist at Ludwig Maximilian University of Munich. Some researchers believe it began at nutrient-rich hydrothermal vents in the ocean, whereas others favor shallow ponds on dry land, which would have frequently evaporated, creating a concentrated bath of chemicals.
A larger ocean exacerbates the biggest strike against the underwater scenario: that the ocean itself would have diluted any nascent biomolecules to insignificance. But by drowning most land, it also complicates the thin pond scenario. Carell, a pond advocate, says in light of the new paper, he is now considering a different birthplace for life: sheltered, watery pockets within oceanic rocks that broke the surface in volcanic seamounts. “Maybe we had little caves in which it all happened,” he says.
The ancient water world is also a reminder of how conditional Earth’s evolution is. The planet was likely parched until water-rich asteroids bombarded it shortly after its birth. If the asteroids had deposited twice as much water or the present day mantle had less appetite for water, then the continents, so essential for the planet’s life and climate, would never have emerged. “It’s a very delicate system, the Earth,” Dong says. “Too much water, or too little, and it wouldn’t work.”