There might be remnants of an ancient planet buried inside Earth

Meanwhile, on Earth

Today, Earth’s mantle isn’t completely uniform. About 8 percent of it is a little different from the rest, and forms two big piles near the core-mantle boundary. These two piles are called Large Low-Shear-Velocity Provinces (LLSVPs), so named because seismic waves called shear waves move about 1 or 2 percent slower when passing through them. And they’re huge: One is beneath the African continent, and the other under the Pacific Ocean.

Some researchers think the LLSVPs slow down the shear waves because they’re a higher temperature than the rest of the mantle. Others, like Yuan and his colleagues, think they’re denser and compositionally different in addition to being hotter.

Yuan says he was sitting in a planetary geochemistry class when the idea struck that the LLSVPs might be related to Theia. As he tells it, he was in ASU professor Micha Zolotov’s class, learning about the giant impact hypothesis for the formation of the Moon. Zolotov mentioned that the weakest part of the theory was the hypothetical planet Theia — no one had ever found any direct evidence to support its existence. It’s totally gone. There’s no evidence of it in meteorites, the asteroid belt, anywhere. When Zolotov said this, Yuan recalls, “It struck me so hard. I thought, after the impact, [Theia] would’ve gone into Earth. Is it possible it went into the Earth and formed the LLSVPs?”

Searching for Theia

Yuan’s first move was to do some simple calculations, first comparing the size of the two LLSVPs to the size of Mars’ mantle — a rough estimate for Theia’s. He found the two LLSVPs were 80 or 90 percent of the size of the Mars mantle. When he added the Moon? “Almost a perfect match,” he says. “So then I thought, it’s not that crazy.”

He pulled up a 2012 Nature paper by geochemist Sujoy Mukhopadhyay at the University of California, Davis, which examined noble gas isotopes from volcanic basalts in Iceland. Mukhopadhyay had shown that Earth’s mantle is heterogenous, with at least two separate sources, and that those sources are at least 4.5 billion years old. That is, older than the Moon. “That was consistent with our hypothesis,” says Yuan. One of the sources could be Theia’s mantle, preserved in Earth’s mantle after the impact.

Next Yuan turned to ASU astrophysicist Steven Desch, who in 2019 had published new estimates for the composition of Theia itself. Desch, along with Katharine Robinson at the Lunar and Planetary Institute in Houston, used the composition of lunar samples from the Apollo missions to model a likely Theia, concluding it was much bigger than expected — about the size of 1 proto-Earth, or 4 Mars planets. Even more important for Yuan, Desch and Robinson estimated that Theia’s mantle had a higher abundance of iron oxide than Earth’s. This means it was denser, so when the two planets collided, Theia’s mantle would sink.

Yuan and Desch joined forces to figure out what Theia’s mantle composition would have needed to look like in order for it to resemble today’s LLSVPs after 4.5 billion years of mantle convection. They found that if Theia was any denser than Desch’s earlier estimate, its mantle would’ve sank too much, forming a global layer instead of two piles. Instead, their calculations revealed the estimates for Theia’s size and density were just right.