Asteroid shower on the Earth-Moon system immediately before the Cryogenian period revealed by KAGUYA

Understanding the timing, magnitude, and potential Earth-system impacts of ancient meteoroid bombardment is challenging because terrestrial records are obscured by erosion and resurfacing. While mass extinctions since the Cambrian have sometimes been linked to impacts, direct evidence for older bombardment episodes is scarce. Building on the well-documented Ordovician L-chondrite parent body disruption and associated meteoroid shower on Earth (~470–480 Ma), this study asks whether the lunar cratering record preserves evidence for other ancient asteroid showers and whether such events temporally coincide with major environmental changes on Earth. Using high-resolution lunar data from the KAGUYA (SELENE) mission, the authors aim to determine formation ages for a global set of fresh lunar craters and to test for statistically significant clustering indicative of a system-wide meteoroid shower around the late Proterozoic, immediately preceding the Cryogenian.

Multiple lines of evidence indicate that the Ordovician (~470–480 Ma) experienced a major meteoroid shower: widespread L-chondritic debris and chromite grains in sediments, Ir enrichment, several large terrestrial craters dated to ~430–470 Ma, and 39Ar–40Ar ages of L‑chondrites near 470 Ma. This episode has been linked to catastrophic disruption of the L-chondrite parent body and potentially to climatic effects via dust-induced cooling. Beyond this event, ancient impacts and their environmental roles remain poorly constrained due to the incomplete terrestrial record. Recent lunar and asteroid-family studies provide context: thermal observations of ejecta suggest a higher lunar impact flux in the last ~290 Myr, while numerical models of asteroid family dynamics (e.g., Agnia, Hansa, and especially Eulalia) offer potential sources and timing for impact spikes in the inner solar system. Geochemical signals in Apollo impact glasses indicate an ~800 Ma spike, suggesting a transient global increase in lunar impact flux beyond Copernicus alone.

- Target selection and imaging: 59 lunar craters with fresh morphologies and diameters ≳20 km were analyzed using Terrain Camera (TC) images (10 m/pixel) from the KAGUYA (SELENE) orbiter. Map-projected (transverse Mercator) images were used. Counting regions excluded obvious impact melt ponds to avoid target property effects. - Crater size–frequency distribution (CSFD): Small (0.1–1 km) craters on continuous ejecta blankets (out to one crater radius from rim) were counted. Obvious secondaries (chains, ellipticals, clusters) were excluded. Self-secondary contamination was assessed as negligible: for the youngest crater (Giordano Bruno), crater density ≥1 km is 4.9×10^-6 km^-2 vs ~5.5×10^-4 km^-2 at 660 Ma, implying <1% contribution. Statistical uncertainties of data points are mostly <30% (1σ). - Age conversion: The standard lunar CSFD and the Neukum chronology were applied to convert CSFD to absolute model ages. Ages were first derived under a conventional constant flux over the last 3 Ga. A Monte Carlo simulation (100,000 iterations) assessed the chance probability of observed age clustering, assuming uniform crater formation probability from 0–3.0 Ga. - Spike flux model: Motivated by radiometric constraints (Copernicus ejecta and Apollo impact spherules) indicating an ~800 Ma event, a new crater-production model was proposed: constant flux reduced to 75% of the conventional level, with a 30 Myr spike (830–800 Ma) at 23× the background, adjusted so total crater counts over 3 Ga match the constant-flux model. Ages were recomputed under this 800 Ma spike model. - Impactor mass estimates: Using crater scaling (D = 1.37·ρ_i^(1/3)·ρ_c^(−1/3)·v^(4/3)·m^(1/3)·g^(−2/3)), impactor masses and sizes were estimated for craters coincident with Copernicus. Assumptions: impact speed v = 20 km s^-1; lunar gravity; crustal density; impactor densities representative of near-Earth asteroids (C-type 1.29 g cm^-3, S-type 1.9–2.7 g cm^-3). Sensitivity tests indicate conclusions are robust to reasonable variations in density (2.5–3.5 g cm^-3) and velocity (10–20 km s^-1). - Earth flux estimate: Using the collision probability ratio Earth:Moon = 23:1, the total mass delivered to Earth was inferred from the lunar bombardment mass.

- Age clustering: Of 59 fresh craters, eight show coeval formation near 660 Ma under the constant-flux model; weighted mean 658 ± 16 Ma. The spatial distribution shows no significant near- vs far-side bias. - Statistical significance: Monte Carlo simulations yield a 0.69% chance that seven of 59 craters fall within a 50 Myr window (630–680 Ma) by chance (excluding one outlier with large uncertainty). Including the outlier, the chance of eight within 100 Myr is 7%. - Absolute timing anchored at ~800 Ma: Considering radiometric ages (Copernicus ejecta soils; Apollo impact spherule 40Ar/39Ar data) indicating a transient global lunar impact increase at ~800 Ma, the clustered craters are interpreted as reflecting an ~800 Ma spike rather than 660 Ma. - 800 Ma spike model results: Recalculation with a 30 Myr spike (830–800 Ma) yields 16 of 59 craters coincident with Copernicus within uncertainties; Table 2 lists 17 craters coincident under the spike model. The constant flux outside the spike is 75% of the conventional model; the spike flux is 23× background. - Impactor mass on the Moon: For the eight clustered craters (constant model), total impactor mass is (1.3–1.6) × 10^15 kg, dominated by Copernicus (D ≈ 93 km; impactor ~10 km). Under the spike model, the total is (1.8–2.3) × 10^15 kg. - Earth delivery: Using the 23:1 collision probability ratio, at least (4–5) × 10^16 kg of meteoroids impacted Earth at ~800 Ma, corresponding to an equivalent single body ~30–40 km in diameter and 30–60× the Chicxulub impactor mass. - Source of the spike: Thermal ejecta analyses show no distinct 660 Ma peak and a generally increased flux over the last ~290 Myr; asteroid family dynamics suggest Eulalia (age 830 [+370, −100] Ma) is a plausible source, capable of injecting debris into the 3:1 resonance. Agnia and Hansa are less favorable sources for producing a ~10 km impactor at lunar low inclinations. - Volatiles and geochemical implications: A C-type asteroid shower at ~800 Ma could have supplied ~10^14 kg of C and H2O to the lunar surface, consistent with observations of H2O at Copernicus and persistent global C+ emissions. For Earth, ~10^14 kg of extraterrestrial phosphorus (assuming CI composition, P ≈ 0.1 wt%) would have been delivered—exceeding the total modern oceanic P inventory by an order of magnitude. - Flux trend: Age–rank plots indicate gentler slopes <300 Ma in both models, consistent with independent evidence that the production rate of lunar craters (D ≥ 10 km) was 2–3× higher over the last ~290 Myr.

The clustered ages of fresh lunar craters, together with independent radiometric ages of Copernicus ejecta and Apollo impact glasses, support a transient increase in lunar impact flux at ~800 Ma. This spike implies a large asteroid shower across the Earth–Moon system immediately preceding the Cryogenian, potentially sourced by disruption of the Eulalia family feeding resonances that deliver debris to the inner solar system. The mass estimates indicate a lunar bombardment of order 10^15 kg and an Earth bombardment of (4–5) × 10^16 kg—far exceeding the Chicxulub event—suggesting capacity for significant environmental perturbations. On the Moon, such a shower could account for observed volatiles (H2O, C+) retained in near-surface materials. On Earth, the influx of bioavailable elements (notably phosphorus) and dust could have influenced marine biogeochemical cycles, redox conditions, and climate during the late Proterozoic, aligning temporally with major environmental transitions prior to and during the Cryogenian. While the terrestrial record lacks clear impact markers (e.g., Ir anomalies) for this interval, subsequent glaciations and erosional processes may have obscured evidence. Overall, the lunar cratering record provides a robust proxy for constraining ancient meteoroid fluxes and their potential Earth-system impacts.

This study identifies a statistically significant cluster of lunar crater formation ages and, when anchored by radiometric constraints, proposes an ~800 Ma spike in the lunar impact flux. A new constant-with-spike model reconciles crater-count ages with radiometric data, implicating a major asteroid shower likely linked to disruption of a low-inclination carbonaceous family (e.g., Eulalia). The inferred bombardment delivered ~10^15 kg to the Moon and (4–5) × 10^16 kg to Earth, with implications for lunar volatile inventories and for Earth’s late Proterozoic climate and biogeochemistry (e.g., phosphorus delivery). Future work should refine the spike’s duration and source dynamics, characterize the size–frequency and dust properties of delivered materials, and seek terrestrial correlates in Neoproterozoic stratigraphy despite preservation challenges.

- Terrestrial corroboration: Direct geological evidence for large Neoproterozoic impacts is scarce; subsequent Snowball Earth glaciations and erosion may have erased impact markers (e.g., Ir, PGE anomalies). - Model dependence: Crater-count ages depend on the assumed production function and chronology (Neukum) and on the adopted flux model; the spike’s duration (assumed 30 Myr) and amplitude (23×) are inferred to reconcile datasets and are not uniquely constrained. - Counting uncertainties: Although obvious secondaries were excluded and self-secondary contributions estimated to be <1%, statistical uncertainties in CSFD measurements (~<30% per point) and potential local target property effects can affect individual ages. Area selection near Copernicus can bias ages via secondary contamination. - Impactor properties: Mass estimates rely on assumed impactor densities and velocities; while sensitivity tests suggest robustness, exact values remain uncertain. - External constraints: Thermal ejecta observations lack resolution near 800 Ma and asteroid family ages have ranges; multiple families could contribute, and delivery efficiencies to Earth–Moon can vary.