The first few billion years of Earth’s history are largely inaccessible to direct geological investigation. The surface that recorded the events of the early Hadean and Archean has been almost entirely destroyed, recycled through plate tectonics, eroded by water and wind, buried beneath younger sediment, or melted and reformed in successive cycles of mountain-building and crustal renewal. The handful of rocks on Earth that have survived from before approximately 3 billion years ago are exceptional outliers, and most of what they contain has been substantially altered by the long passage of geological time.
The result is a strange situation. The period in which life first emerged on this planet, between approximately 4 billion and 3.5 billion years ago, is one of the most consequential intervals in the history of the solar system, and it is also one of the periods for which Earth itself preserves the least direct evidence.
To understand what was happening here when life was beginning, scientists have to look somewhere else.
Why the Moon preserves what the Earth has destroyed
The Moon, on the available evidence, has none of the geological processes that have erased Earth’s earliest record. There is no plate tectonics on the Moon, no flowing water, no atmosphere capable of weathering rock, no biosphere capable of breaking it down. The lunar surface is a passive recording medium that preserves whatever falls on it for as long as the rock itself survives. The same impact events that have left no visible trace on Earth, because the Earth has moved on, are still readable on the Moon, because the Moon has not.
The Moon also shares Earth’s neighbourhood and its impact history. The two bodies have been close companions for approximately 4.5 billion years, sweeping the same volume of space and being struck by debris from the same population of asteroids and comets. What hit the Moon, on the available evidence, was broadly representative of what was hitting the Earth at the same time. The lunar record is, in this sense, a recoverable substitute for the terrestrial record that no longer exists.
Lunar samples reach Earth through two routes. The first is direct collection by spacecraft, including the Apollo missions, the Soviet Luna missions, and the Chinese Chang’e missions, all of which returned material from specific locations on the lunar surface. The second is more accidental. Occasionally an asteroid strikes the Moon hard enough to eject fragments of lunar rock at velocities exceeding the lunar escape velocity. Some of these fragments drift through the Earth-Moon system for years or millennia before falling to the surface of the Earth as meteorites. Approximately 600 lunar meteorites have been catalogued so far, and each contains a record of the part of the lunar surface from which it was ejected.
What the Africa meteorite contains
In May 2026, a team led by Carolyn Crow at the University of Colorado Boulder published the results of a detailed analysis of a particular lunar meteorite, catalogued as NWA 12593, in the journal Geology. The meteorite had been found in northwest Africa, in a region where meteorite hunting has become a substantial commercial activity, and had been recovered for scientific analysis. The Crow team applied a combination of radiometric dating techniques, mineralogical analysis, and electron backscatter diffraction imaging to reconstruct the events the rock had recorded.
NWA 12593 turned out to contain evidence of three separate impact events on the Moon, each of which had left distinguishable mineralogical signatures in the same small fragment of rock.
The oldest and most significant event, on the radiometric dating evidence, occurred approximately 3.486 billion years ago. The energy released was sufficient to melt the surface of the surrounding lunar region into a flowing sheet of liquid rock. The temperatures reached during the event were high enough to produce cubic zirconia, a mineral form of zirconium dioxide that is also synthesised commercially for use in jewellery. Cubic zirconia forms only at temperatures above approximately 2,370 degrees Celsius and does not normally survive in nature, because the mineral undergoes structural transitions to lower-temperature forms as it cools. What the Crow team detected in NWA 12593 was not intact cubic zirconia but the characteristic structural ghost left behind in the crystal lattice, called cubic zirconia phase heritage, which is diagnostic of the original high-temperature formation.
The second impact event was a smaller collision that occurred some time after the first. It shattered the solidified melt sheet produced by the first event into fragments and welded them together under impact-generated heat and pressure into a type of rock called a breccia. NWA 12593 is itself that breccia, a fused composite of broken material from the original melt sheet and surrounding rocks, the mineralogical equivalent of crushed concrete reformed under enormous pressure.
The third event was the more recent collision that knocked the breccia off the lunar surface entirely and launched it on a trajectory that eventually delivered it to Earth. The Crow team has not yet been able to date this third event precisely, but it was recent enough in geological terms that the rock survived the journey without further significant alteration.
What 3.486 billion years ago means
The oldest impact recorded in NWA 12593 is interesting on its own, as evidence of a major event in the lunar bombardment history. It becomes substantially more interesting when compared to the impact records preserved on other bodies in the inner Solar System.
On Earth, the same approximate period is recorded in a series of geological formations called spherule beds, layers of glass droplets and shattered rock created by the deposition of debris from large impact events. The oldest well-dated spherule beds on Earth, found in the Barberton Greenstone Belt in South Africa and in the Pilbara Craton in Western Australia, date to approximately 3.47 billion years ago. The matches between the lunar impact age and the terrestrial spherule beds, on the Crow team’s analysis, are close enough to suggest a shared bombardment event rather than independent coincidence.
The third match is on 4 Vesta, the fourth-largest body in the asteroid belt and the source of a substantial family of meteorites called eucrites that have fallen to Earth. The eucrites carry their own radiometric record of impact events on their parent body, and the oldest large impact events recorded in the eucrite record cluster around the same 3.5 billion year window. The Crow team’s interpretation is that the convergence of impact ages on the Moon, on Earth, and on 4 Vesta, three separate bodies in different parts of the inner Solar System, points to a common cause rather than a series of independent coincidences.
The most parsimonious common cause, on the available evidence, is the catastrophic breakup of a large asteroid somewhere in the inner Solar System at approximately that time. The resulting debris would have spread across the inner Solar System over a period of approximately 500 million years, producing a wave of impacts on every body it encountered. The bombardment window the Crow team identifies in the paper title, 3.7 billion to 3.2 billion years ago, is consistent with the expected duration of such a debris wave.
The connection to early life
The earliest well-accepted fossil evidence of life on Earth, documented in a peer-reviewed 2006 study by Abigail Allwood and colleagues in the journal Nature, was found in stromatolite formations in the Pilbara Craton in Western Australia and dates to approximately 3.43 billion years ago. The Pilbara stromatolites are layered sedimentary structures produced by communities of ancient microorganisms living in shallow marine environments, and the Allwood team’s analysis established their biogenic origin against competing abiotic hypotheses that had been actively defended in the peer-reviewed literature for several decades. The Apex Chert microfossils, also from the Pilbara region, are dated to approximately the same period and represent some of the earliest candidate evidence of microbial life on the planet. Life on Earth, on the strongest current reading of the available evidence, was emerging and beginning to colonise the planetary surface at exactly the same moment that the bombardment wave the Crow team has identified was striking the inner Solar System.
The relationship between the impacts and the emergence of life is genuinely contested. One view, supported by some peer-reviewed analyses, is that large impact events would have been catastrophically destructive to any nascent biosphere, sterilising the surface and forcing life either to retreat into deep subsurface refugia or to begin again after the bombardment subsided. A second view, supported by other analyses, is that the impacts may have been crucial to the emergence of life rather than hostile to it. Major impacts can produce sustained hydrothermal systems, deliver organic molecules and water from the impactors themselves, and create chemically diverse environments of the kind that prebiotic chemistry models have suggested as plausible sites for the assembly of the first biological molecules.
The Crow team’s findings do not directly resolve this dispute. What they establish is the cadence of the bombardment, the fact that major impacts were occurring at the moment life was emerging, and the fact that the same bombardment was hitting multiple bodies in the inner Solar System simultaneously. Whether the impacts helped life or hindered it, on the available evidence, is a question for further peer-reviewed work.
The honest limitations
Several methodological caveats apply to the literature described above.
The radiometric dating of impact events depends on isotopic systems that can be partially reset by subsequent thermal events. The 3.486 billion year date for the first impact in NWA 12593 is robust, but the assumption that the date reflects a single discrete impact rather than a cluster of closely spaced events cannot be made definitively on the basis of one rock. The broader bombardment window of 3.7 to 3.2 billion years ago that the Crow team identifies is more securely established than any single impact age within it.
The interpretation that the impact ages on the Moon, Earth, and 4 Vesta reflect a shared cause is the most parsimonious explanation but is not the only available one. The convergence of impact ages could, in principle, be the result of three independent processes that happened to produce similar timing, although the prior probability of such independent convergence is low. The shared cause interpretation is the best current reading of the evidence but is not definitively established.
The connection between bombardment and the emergence of life is a correlation rather than a demonstrated causal relationship. The timing matches, but the timing also matches a great many other geological and chemical events that were happening on the early Earth at the same period. Establishing that the impacts caused or contributed to the emergence of life, rather than merely coinciding with it, would require evidence that the current peer-reviewed literature does not yet have.
What it means
Several things follow from the Crow team’s evidence that are worth saying clearly.
The first is that the early history of the inner Solar System, in the period between approximately 4 and 3 billion years ago, was substantially more chaotic than the terrestrial geological record alone can convey. Earth has erased most of the evidence of its own bombardment history. The Moon and the asteroid belt have not. The lunar and asteroidal records, on the strongest current reading of the available evidence, indicate that major impact events were continuing across the inner Solar System for hundreds of millions of years after the conventional end of the so-called Late Heavy Bombardment around 3.9 billion years ago.
The second is that the bombardment, whatever its cause, was happening at exactly the moment life on Earth was leaving its earliest detectable traces. The 3.5 billion year window includes the Pilbara stromatolites, the Apex Chert microfossils, and the geochemical isotope evidence of early biological activity. It also includes the impact event recorded in NWA 12593, the corresponding spherule beds on Earth, and the corresponding impact ages on 4 Vesta. The two stories, of bombardment and biogenesis, are operating in the same period and on the same planet.
The third is that the methodological approach of reconstructing Earth’s earliest history through lunar and meteoritic samples is now genuinely productive. The Crow team’s analysis of a single small rock from northwest Africa has produced evidence of events that occurred 3.486 billion years ago on the lunar surface, has tied those events to independently recorded events on Earth and in the asteroid belt, and has placed all of them in the context of the emergence of terrestrial life. The geological record that Earth has lost is, on the available evidence, partially recoverable from rocks that fell from elsewhere.
The fourth, on the strongest current reading of the peer-reviewed evidence, is that the first 1.5 billion years of life on Earth were lived under a sky that was substantially more dangerous than the modern sky, on a planet that was being struck repeatedly by debris from events the surviving terrestrial geology can no longer fully describe.
What survived that period went on to become every living thing.
The rest of the story is now being recovered, in fragments, from the rocks that fell here from elsewhere.
