The Pilbara craton in Western Australia is made up of some rocks that are 3.5 billion years old
Elizabeth Czitronyi / Alamy
Rocks in Australia preserve evidence that plates in Earth’s crust were moving 3.5 billion years ago, a finding that pushes back the beginnings of plate tectonics by hundreds of millions of years.
Today, around eight vast, rigid plates of rock at the surface of the planet, plus some smaller plates, are pulled or pushed along a softer layer of rock beneath. When the edges of these plates slip or slide past one another, sudden geological events can occur, like earthquakes, as well as more gradual processes, such as the formation of mountain ranges.
But geologists disagree over how many plates there once were, when they started moving and how they used to move. Some researchers claim they have found evidence from as far back as 4 billion years ago, when the planet was significantly hotter, while others say the strongest evidence is more recent, from 3.2 billion years ago.
Most of this evidence consists of hints from the chemical composition of rocks, which geologists can use to infer how those rocks moved in the past. However, there is little record of how early plates may have moved relative to each other, which is seen as the strongest evidence of tectonic plate movements.
Now, Alec Brenner at Yale University and his colleagues say they have found unambiguous evidence of relative plate motions around 3.5 billion years ago in the eastern Pilbara craton in Western Australia. The researchers tracked how the magnetic field of the rocks, which was aligned with Earth’s magnetic field, moved over time, similar to how a compass buried in the rock would change its needle direction as the ground moved.
Brenner and his team first dated the rocks by analysing the radioactive isotopes they contain, then proved the rocks’ magnetisation hadn’t been reset at some point. By tracking how this magnetisation had moved, they could show that the entire rock region had migrated over time, at a rate of tens of centimetres a year. Then, they compared this with rocks that had been dated and tracked using the same technique in the Barberton Greenstone Belt in South Africa, which showed no movement.
“It means that there had to have been some kind of plate boundary in between these two [regions] to accommodate that relative motion. That’s plate motion, definitionally,” Brenner told the Goldschmidt geochemistry conference in Prague, Czech Republic, on 9 July.
“The Pilbara, around 3.8 billion years ago, moves from mid-to-high latitudes to very high latitudes, actually within the area of the geomagnetic pole, and probably close to around where Svalbard’s latitude is today, in just a few million years. While the Barberton is just sitting there, doing nothing much at all on the equator,” said Brenner.
“If two plates are moving relative to each other, there has to be an awful lot of stuff going on between as well,” says Robert Hazen at the Carnegie Institution for Science in Washington DC. “It can’t just be an entirely local thing.”
But there is scope for different interpretations of what was causing that movement, says Hazen. This is partly because there is widespread uncertainty on how fast the plate was moving, and the data could fit several different theories of what Earth’s interior looked like at that time.
At the very least, the finding implies the existence of a tectonic boundary, says Michael Brown at the University of Maryland. However, he says that the motion of the rocks appears markedly different from what we understand as plate tectonics today. “Essentially, the Pilbara [plate] goes steaming up to higher latitudes and stops dead, which is unusual in any plate tectonic context.”
Brown argues that this fits with a theory that Earth’s crust at that time was made up of many smaller plates that were pushed around by columns of hot rock, called plumes, surging up from the more molten mantle. The surviving remnants of these smaller plates, which in this view Brenner and his team would have sampled from, are useful to indicate that there was motion, but because they are only a small proportion of the crust, they might not be representative of how Earth was moving, says Brown.
Brenner and his team also found evidence that Earth’s magnetic field direction flipped 3.46 billion years ago, which is 200 million years before the next-most-recent flip. Unlike today’s magnetic field, which reverses roughly every 1 million years, the magnetic field back then appeared to flip less frequently, at a rate of tens of millions of years. This might imply “quite different underlying driving energetics and mechanisms”, said Brenner.
What Earth’s magnetic field looked like at that point in its development is also hotly debated, says Hazen, in part due to the lack of magnetic data. “I think this moves the bar,” he says. “It’s a really significant finding of a reversal that early. It tells you something about the geodynamics of the core that wasn’t nailed down.”
Topics: