Read our COVID-19 research and news.

Earth’s magnetic field produces an aurora when disturbed by the solar wind.

tiziano valeno/iStock

New signs of a shielding magnetic field found in Earth’s oldest rock crystals

Earth’s earliest lands, hot and hellish, were sheltered from above. Researchers have found more evidence that our planet had a strong magnetic field 4.2 billion years ago, three-quarters of a billion years earlier than previously thought and just 350 million years after Earth formed. The field would have shielded Earth, protecting its atmosphere from being stripped away by high-energy particles from the Sun—and perhaps helping life gain a foothold.

With few surviving rocks to study, geologists struggle to reconstruct the time known as the Hadean, which ran from 4.55 billion years ago to 4 billion years ago. But fragmentary—and controversial—clues can be found in younger, 3-billion-year-old rocks from the Jack Hills of Western Australia. These rocks contain tiny crystals of a hardy mineral called zircon, which are chips off an even older block: 4.2-billion-year-old Hadean rocks that formed from cooling magma.

The crystals also preserve evidence of an ancient magnetic field, according to an international team led by John Tarduno, a geophysicist at the University of Rochester. Not all researchers are convinced by the result, because it would push back the accepted birth date of Earth’s magnetic field by 750 million years. “There’s been a large research group trying to prove our results wrong,” Tarduno says. “That’s part of science.”

Tarduno and his colleagues first reported the fields in an analysis published in 2015 in Science. Trapped inside about two dozen zircons were even tinier grains of an iron-containing mineral, magnetite, that effectively turned each crystal into a miniscule bar magnet. The researchers found that the magnetic fields held by the grains were all aligned, which occurs only if magnetite is exposed to a magnetic field as it cools.

A zircon crystal fits comfortably inside the “o” on a U.S. dime’s lettering.

Julie Mansy; Rory Cottrell/University of Rochester; John Tarduno/University of Rochester

Tarduno’s team says the magnetization was imprinted 4.2 billion years ago, when the original zircon-containing rock first cooled. However, if the magnetite grains, at any point in their history, became hot enough—above about 600°C—they would have lost the magnetic alignment and gained a new one as they cooled down again. And just such a heating event may have occurred about 2.6 billion years ago, says Benjamin Weiss, a geologist at the Massachusetts Institute of Technology (MIT) and leader of a group that has challenged Tarduno’s claims. “These zircons have mind-bogglingly long and largely unknown histories,” he says.

In the Proceedings of the National Academy of Sciences today, Tarduno’s team marshals a new line of evidence to show that the magnetization happened in the Hadean instead of much later. Zircon crystals contain zones rich in lithium ions that, when heated, can bleed over time into adjacent areas through a process called chemical diffusion. Tarduno and his colleagues measured the lithium concentrations across the boundaries of these zones in three of their zircon crystals. In two, they found limited evidence of diffusion. Tarduno says it’s a sign they were never heated up above 600°C in their 4.2-billion-year history—and that their magnetic signature is original. “I think that’s a remarkable finding.”

What’s more, the single zircon crystal that did show signs of heating is an outlier in another way: It preserves an unusually weak magnetic signal. In 2015, Tarduno and his colleagues concluded this might be evidence of a strangely variable magnetic field 4.2 billion years ago. But now, he says, it seems that this particular zircon gained its magnetism later, and its weak signal can be ignored in discussions of Earth’s earliest magnetic field. Doing so, says Tarduno, suggests Hadean Earth actually had a magnetic field, and a churning dynamo in its core, as strong as the one operating today.

Mark Harrison, a geologist at the University of California, Los Angeles, isn’t convinced by the new analysis, because it takes both high temperatures and long periods of time for the lithium ions to diffuse across a boundary. If a heating incident brought the zircons above 600°C but persisted for only 10,000 years—a blink of an eye for a geologist—it would be enough time to reset the zircons’ magnetism but not to diffuse lithium. “I don’t think this is even remotely smoking gun evidence,” he says.

MIT geologist Claire Nichols hasn’t been involved in any of the Jack Hills studies but she did recently argue that Earth’s magnetic field was in place at least 3.7 billion years ago after analyzing rocks in Greenland. “It’s great to see different research groups pushing each other to use more and more advanced techniques,” she says of the new study. “It gives us the best chance of understanding the earliest record of the geodynamo.”

Posted in:


This story, on our recent paper in the Proceedings of the National Academy of Sciences (PNAS) documenting Earth’s most ancient magnetic field, omits key information. Comments attributed to Mark Harrison are inconsistent with a paper that Harrison co-authored on this exact issue. He and his colleagues wrote [bold highlight added] “… a hypothetical heating event at 600°C would generate this slope in only ~10 kyr. Because this duration is implausibly short for a geologic (tectonic) event, it would be reasonable to conclude that the zircon under consideration did not approach 600°C after formation of the Li profile.” [Trail, Cherniak, Watson, Harrison, Weiss, Szumila, Contributions to Mineralogy and Petrology, 2016] This statement indicates that these authors agree that Li data can test scenarios that are plausible for resetting the magnetization of Jack Hills zircons. Our select Hadean zircons pass such tests. The story also does not report that our PNAS article cites multiple lines of new evidence that support a strong magnetic field more than 4 billion years ago. Of central importance are the paleomagnetic analyses that test and exclude remagnetization events. Nor does the story mention any agreement between results from our group and others. Specifically, the high temperature magnetite carriers of the ancient magnetization that we first reported in 2015—whose presence was challenged—have been confirmed. We invite readers to download our peer-reviewed open access PNAS article to learn more about how we have developed and applied methods to recover records of the geomagnetic field using zircons and how our new data elucidate processes in Earth’s core.
John Tarduno, University of Rochester

More from News