The early inner Solar System was struck far more often than it is today. The disputed point is not whether large impacts occurred, but how their frequency changed. Did bombardment decline gradually as leftover planet-forming debris was removed, or did a delayed dynamical disturbance produce a brief cataclysm around 3.9 billion years ago? The phrase “Late Heavy Bombardment” is often used as if it names a settled interval. It actually describes one influential interpretation of an incomplete record.
Apollo Samples Created an Apparent Spike
The hypothesis took shape after lunar samples returned by Apollo missions were radiometrically dated. Many impact-melt rocks appeared to record ages near 3.9 billion years. Impact melt forms when a collision heats target rock enough to liquefy it; once the melt cools, isotopic clocks can record that event. A cluster of similar ages from multiple landing sites seemed to indicate that several enormous lunar basins formed within a comparatively narrow window.
That pattern was surprising. In the simplest model, impacts should have been most common immediately after planets formed, then become progressively rarer as they swept up or ejected remaining debris. A late spike required a mechanism that destabilized a new population of impactors. Models involving migration of the giant planets showed how changing resonances could scatter asteroids and other small bodies toward the inner planets.
The Near Side May Repeat One Event
Apollo did not sample the Moon evenly. All six landing sites occupy the near side, and much of that hemisphere was showered with ejecta from the enormous Imbrium basin. Rocks collected far apart can therefore contain fragments or thermal signatures ultimately related to the same basin-forming collision. If Imbrium formed near 3.9 billion years ago, its wide influence could create an apparent global concentration of ages without recording numerous independent basins.
An impact age can also be altered. Later heating may partly or completely reset an isotope system, and a breccia can combine pieces formed during different events. Different minerals close their isotopic clocks at different temperatures. Assigning a date to a basin consequently requires more than measuring an age: the sample’s origin, mineral structure, chemical history, and regional geology must also support the connection.
The oldest lunar record is especially difficult to preserve. A new impact can melt earlier deposits or bury them beneath ejecta. The absence of very old ages among sampled melt rocks may reflect destruction and resetting rather than an absence of impacts before 4 billion years ago.
A Cataclysm and a Decline Predict Different Histories
In a classic cataclysm model, impact activity falls after planet formation and then surges hundreds of millions of years later. In a monotonic-decline model, the rate starts extremely high and generally diminishes, though individual large basins still punctuate the trend. Intermediate proposals allow several pulses or a prolonged tail rather than one sharp episode.
Each scenario affects more than lunar chronology. Because Earth’s crust is continually recycled, the Moon is used as a proxy for impacts that also reached Earth. A short cataclysm would imply an abrupt environmental crisis near the end of the Hadean. A long decline would distribute impact heating, volatile delivery, crustal disruption, and possible habitable intervals differently through early history. It would also change how crater counts are converted into ages for unsampled surfaces on Mars and Mercury.
Far-Side Samples Weaken the Narrow-Spike Model
The first samples returned from the Moon’s far side provide a crucial test independent of Apollo’s nearside geography. Chang’e-6 collected material in 2024 from the Apollo basin inside the much larger South Pole–Aitken basin. Analyses published in 2026 dated individual impact-melt clasts with the argon-40/argon-39 method.
The measured events ranged from about 4.33 to 1.13 billion years ago. Three clasts placed formation of the Apollo basin near 4.16 billion years ago, and only two reported ages fell between 4.0 and 3.7 billion years ago. The presence of pre-4.0-billion-year melt and the lack of a dominant group near 3.9 billion years support a smoother long-term decline punctuated by later impacts rather than a single overwhelming lunar cataclysm.
One sample set cannot reconstruct every collision, and the precise age of the South Pole–Aitken basin remains under examination. Argon systems can be reset by heating, source regions can be mixed, and local geology affects which events survive in a returned collection. The new evidence is powerful because it samples a previously missing hemisphere, not because it eliminates all uncertainty.
The Debate Is Becoming More Precise
“No Late Heavy Bombardment” would be misleading if it implied a quiet early Earth and Moon. Immense basins formed, and the declining background rate was still heavy by modern standards. What is losing support is the narrowest version of the hypothesis: a Solar System-wide, sharply delayed spike inferred mainly from the 3.9-billion-year clustering of nearside lunar samples.
Further returned samples must be tied to well-mapped basins with independent geological context. Combining multiple isotope systems, microscopic examination, crater statistics, and orbital models can distinguish a source event from later resetting. The question is no longer simply whether the Moon was battered. It is whether a biased archive made a long, uneven decline look like one late catastrophe.