Giant meteorites could be responsible for Earth’s continents

The Pilbara craton, in the Pilbara region of Western Australia, is the best-preserved remnant of Earth’s ancient continental crust (see figure 1a). It formed during the Archean eon 4 billion to 2.5 billion years ago. At that time, Earth is believed to have had a water-covered primordial crust that was more or less a single, continuous shell around the mantle. Eventually, the primordial crust transformed into continental crust, which was later joined by its thinner and denser cousin, oceanic crust.

Figure 1.

Most present-day oceanic crust is at most 200 million years old, and its formation is ongoing and well understood: When a gap opens between tectonic plates, magma bubbles up into the gap, cools, and spreads out from the initial ridge. Continental crust is much older; three-quarters of the present-day landmass worldwide formed in the Archean as did the Pilbara craton. Because the continents have been around so long, figuring out their genesis is a challenge. Earth’s geological evolution gradually and often violently transforms and mixes continental crust until evidence of the distant past is erased.

Since the mid 1960s, one theory has argued that continents formed at the sites of collisions with asteroids tens or a couple hundred kilometers wide. ¹ The resulting heating and excavation of the primordial crust would have triggered a sequence of events in which the melted mantle swelled up and overfilled the crater to create a plateau. That seed would then have grown into a continent.

The idea makes chronological sense; the solar system saw a barrage of asteroids around 3.9 billion years ago, a period known as the Late Heavy Bombardment. The Moon’s craters are evidence of that period (see the article by Brett Denevi, Physics Today, June 2017, page 38 ). Earth surely underwent a more ferocious onslaught, as it has more surface area and a larger mass to draw in objects, but the evidence has been limited. And no concrete proof has connected the Late Heavy Bombardment to the formation of continents.

Now Tim Johnson of Curtin University in Australia and his colleagues have found support for the theory that continents formed at the sites of giant impacts. ² They show that the isotopic composition of Pilbara craton rocks matches what’s expected if an asteroid hit the region. Their description of how the craton emerged may explain not only the Pilbara’s formation but the making of Earth’s continents.

Rock of ages

Johnson, who hails from the UK, arrived in Western Australia in early 2014, just after he had begun thinking deeply about early Earth—that is, the first billion or so years of its four-and-a-half-billion-year existence. He soon met Hugh Smithies of the Geological Survey of Western Australia (GSWA), a division of the state’s Department of Mines, Industry Regulation, and Safety. The GSWA was founded over a century ago to gather geological information.

Smithies had long been researching ancient rocks from the Pilbara craton and elsewhere in Western Australia. Those rocks offered an ideal way to study early Earth. Over the course of nearly a decade, Smithies, Johnson, and their collaborators teamed up on a series of projects that used data from Pilbara craton samples. The new study, a culmination of that work, started when Smithies and Yongjun Lu, also from the GSWA, went to Curtin University to discuss some intriguing recent data with Johnson and his colleague Chris Kirkland, an expert on isotope geology.

The team used secondary-ion mass spectroscopy on 26 Pilbara craton rocks that were 2.9–3.6 billion years old. The measurements focused on grains of the magmatic mineral zircon (shown in figure 1b), which is common in igneous rocks. Such grains are useful because they can be dated through the decay of the material’s radioactive uranium. What’s more, zircon’s compositional ratio of oxygen-18 to oxygen-16 isotopes provides a snapshot of its environment—source materials, temperature, and so on—at the time it formed. For the Pilbara samples, the ratios showed three distinct ranges of values that depended on the age of the zircon, as shown in figure 2.

Figure 2.

The team wondered whether a giant impact could explain the observation. A strong sign would be that the oldest zircon grains show evidence of forming not from mantle-derived magma originating deeper in Earth but from surface materials melted together by the exceptional heat and compression of an asteroid impact. Such a collision would also produce a whole series of events with potentially discernible influences on the oxygen isotope ratios.

Watered down

The researchers investigated the ¹⁸O/¹⁶O isotope ratio of the zircon samples relative to Vienna Standard Mean Ocean Water, a pure-water benchmark distilled from ocean water gathered from around the globe (see Physics Today online, “Setting standards with old rocks and ocean water ,” 6 December 2018). The zircon’s ratio is then expressed as a relative difference δ¹⁸O, typically in parts per thousand (‰).

A useful point of comparison is Earth’s mantle: Its δ¹⁸O, shown as a gray horizontal stripe in figure 2, is reasonably homogenous and has been stable over time. The mantle serves as an oxygen isotope reservoir, and minerals formed there will more or less share the same δ¹⁸O. For minerals formed outside the mantle, δ¹⁸O generally depends on the environmental temperatures. ³ The low-temperature hydrothermal interactions on Earth’s surface lead to materials with higher δ¹⁸O than the mantle. Those just below the surface have higher-temperature rock–water interactions and lower δ¹⁸O.

Johnson and his colleagues found that for zircon grains older than 3.4 billion years, which they named stage 1, a third have ratios less than that of the mantle. Zircon ages 3.0–3.4 billion years (stage 2) have a higher median δ¹⁸O than that in stage 1, with more than half of the ratios falling in the range expected for the mantle and nearly all the rest higher. For the youngest zircon, with ages less than 3.0 billion years (stage 3), more than three-fourths have δ¹⁸O higher than the mantle.

Johnson and his colleagues argue that the data fit with what would be expected from a giant impact. The low δ¹⁸O values in stage 1 indicate a material source near Earth’s surface that was hydrothermally altered at high temperatures, as expected if a burning meteorite crashed into, cracked, and melted a primordial crust covered with ocean water.

The heating and compression from such an impact would spread down to the mantle, which would melt and eventually swell up to fill the crater until it becomes a plateau. That continental nucleus would have crystallizing magmas at its base, which would melt to produce granites with the mantle-like isotope ratios seen in stage 2. Finally, in stage 3, dense near-surface rocks, with their higher δ¹⁸O, would sink and end up in a process of intracrustal recycling sometimes called sagduction, akin to subduction but without tectonic plates. From there that continental nucleus would grow.

Additional evidence supports the idea that giant impacts occurred in the region. The Pilbara craton has what’s known as spherule beds, a layer of spherical sediment droplets. Spherules are commonly believed to arise from ejected liquid or vaporized rock from a giant impact. The oldest spherule beds in the Pilbara are about 3.4 billion years old, which reassuringly is the same age as a large cluster of low δ¹⁸O zircon grains.

Johnson says that some researchers have been reluctant to accept that giant impacts shaped our planet in any appreciable way, so he suspects that the new work will face opposition in the community. “I’m hoping that the research will stimulate vigorous activity among several groups that will be keen to prove us wrong,” Johnson says. “The early Earth is a very controversial scientific endeavor. We’re talking about so long ago, and the rocks are so old and have undergone such an uncommonly violent subsequent history.”

Any alternative explanation would need to fit the data as well as a giant impact does, though. Two popular alternative theories—that hot matter bubbled up from the mantle to form a continental seed or that plate tectonics were already at play in the Archean—both predict the oldest zircons would have origins from deep below the surface and thus have mantle-like δ¹⁸O values.

Johnson and his colleagues’ next step is to investigate other well-preserved ancient rocks in northwest Canada, southern Greenland, and a few other territories to see whether giant impacts could explain all continental masses, not just the Pilbara region. Their initial look at data from the Slave craton in Canada is promising.