Constraining the Lunar Cratering Chronology via Numerical Modeling

CONSTRAINING THE LUNAR CRATERING CHRONOLOGY VIA NUMERICAL MODELING. A. M. Blevins1, D. A. Minton1, Y. H. Huang2, J. Du1, M. M. Tremblay1, C. I. Fassett3, 1Department of Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, Indiana 47907, USA ([email protected]), 2Research Center for Soil & Water Resources and Natural Disaster Prevention, National Yunlin University of Science and Technology, Douliou, Yunlin 640301, Taiwan, 3Applied Physics Laboratory, Johns Hopkins University, Laurel, MD 20723, USA. Introduction: Crater chronology functions describe the amount of cratering experienced in a given location over time. The lunar chronology function is anchored by calibration points derived from radiometric ages of Apollo samples that have been correlated to surfaces with known crater densities measured by crater counting [1,2]. The youngest calibration points are associated with small craters located near Apollo landing sites, while the oldest calibration points are associated with large basins. The remaining calibration points are associated with mare basalts that originated from volcanism ca. 2-3.8 Ga. The chronology of Neukum et al. [1] is widely used to describe lunar cratering rates. Robbins [2] performed his own counts of lunar mare regions and found them to be significantly different than Neukum’s. He then constructed his own chronology function based on his counts and found significant differences in the region where the chronology function should be most well constrained (fig. 1). In particular, he found evidence for a rapid decline in impact bombardment during the Imbrian period (3.23.9 Ga), which is when most of the mare formed. The chronology of Xie and Xiao [3], which incorporates new calibration points from the Chang’E missions, also includes a more rapid decline than the chronology of Neukum et al. Figure 1: Chronology of Neukum et al. [1], and Robbins [2], with calibration points shown. Also shown are regions where the function is poorly constrained and unconstrained, respectively [17]. At least 74 basins (craters >200 km in diameter) impacted the Moon in its early history [4]. However, unlike mare basalts, it is extremely difficult to correlate samples associated with impacts to their craters of origin [5]. This is due to the impact process transporting large amounts of material throughout the Moon, making the origin of each particular sample unclear. Imbrium is the only basin with a welldefined age that is universally agreed upon [6,7]. As a result, the chronology function is unconstrained for ages older than Imbrium (3.9 Ga) [2]. Using a Numerical Model to Simulate Lunar Impact Hustory: The goal of our work is to utilize numerical modeling techniques to provide constraints on the timing of lunar bombardment as described by chronology functions. We focus on modeling impact melt, since the radiometric ages of impact melt samples should also date the impact that formed them, as the melt was produced at essentially the same time as crater formation [8]. We use the Cratered Terrain Evolution Model (CTEM) [9,10] to simulate impact bombardment throughout lunar geologic time. CTEM is a Monte Carlo model at its core, emplacing random craters in random locations on the model grid with the timing constrained by the input chronology function. However, we have modified the code to accept manuallyemplaced craters. We then emplaced the lunar basins catalogued by [4] in a sequence based on [11]. The result is a model grid that resembles the real Moon, including model Apollo landing site locations based on distance from bearside basins like Imbrium and Serenitatis (fig. 2). Figure 2: Comparison of a pure Monte Carlo simulation in CTEM (left) and our modifed simulation (right), which includes the basins. The model Apollo 14-17 landing sites are also shown. Each crater consists of a melt zone and an ejecta zone, which overlap. Melt located in the ejecta zone Imbrium A14 A16 A17 A15 1000 km 1000 km Crisium Serenitatis Nectaris Orientale SPA SPA Asperitatis is ejected to a point on the surface based on the Maxwell Z-model [12-14]. The remainder of the melt coalesces in the interior of the crater, forming a melt sheet. If the ejecta zone of a subsequent crater falls within the melt sheet of the first crater, that melt will be ejected. CTEM tracks the crater of origin and age of all impact melts. At the end of the simulation, the age and source crater for melt in the upper meter of regolith at the Apollo 14-17 landing sites is examined. Results shown are for the aggregation of 85 separate simulations. Rapid Decline in Impact Bombardment Dring the Imbrian Period: We modeled the impact bombardment during the Imbrian period under both the Neukum and Robbins chronologies. Under the Neukum chronology, there is more overall melt in the Imbrian period and its age distribution is more widespread. Under the Robbins chronology, melt is concentrated at older ages. The Apollo 16 landing site is the best site for this comparison, as it has the most returned impact melts that have been radiometrically dated [15]. These melts have been associated with multiple impacts [16]. We scaled our numerical model results to the number of samples and found the Robbins function is a better match to the age distribution of impact melts (fig. 3). Therefore, our model supports a rapid decline in impact flux during the Imbrian [17]. Figure 3: Scaled model results for the Neukum (left) and Robbins (right) chronologies. Model melt ages (blue) are compared to observed sample ages (gold). Scenarios for the Timing of Basins: For this analysis, our numerical model results for the basinforming period were inputted into an algorithm based on Bayesian statistics that was developed to find plausible correlations between the timing of basins and the age distribution of basin-era Apollo impact melts catalogued by [15]. We find that Serenitatis has a high likelihood of being correlated to 4.2 Ga-old Apollo 16 samples. This matches its correlation to non-melt Apollo 17 samples with the same age [18]. If we assume this age for Serenitatis, as well as Imbrium’s age of ~3.9 Ga, different scenarios arise depending on which sample groups are correlated with Nectaris and Crisium, respectively: Scenario ALate Spike in Impact Flux: In this scenario, Nectaris and Crisium are correlated to sample groups with ages between 3.9 and 4 Ga. Thus, basin formation spiked during this period, not unlike the cataclysm or “late heavy bombardment” theories [19]. Scenario BGradual Decline in Impact Flux: In this scenario, Crisium is correlated to a group at 4 Ga and Nectaris is correlated to a group at 4.1 Ga, which matches a peak in non-melt breccias [15]. This scenario is more in line with a gradual decline in impacts or “accretion tail” [20]. Scenario CHybrid of A and B: In this scenario, Crisium formed not long before Imbrium (like in Scenario A), and Nectaris formed at ~4.1 Ga (like in Scenario B). The impact flux could thus resemble the “sawtooth” shape proposed by [21]. Each of these scenarios are possible based on our analysis. Depending on the scenario, the behavior of impactors could be anything from a “late heavy bombardment” to a smooth decline in impact flux. Each scenario relies on Serenitatis being 4.2 Ga, which matches its model age derived from extrapolating the Neukum function past 3.9 Ga [22]. Thus, an age of 4.2 Ga for Serenitatis [18] is not evidence that the Neukum chronology is valid that far back in time. This problem can only be solved by the collection and dating of more samples from the Moon, preferably undisputed basin samples from melt sheets [5]. References: [1] Neukum, G. et al. (2001) SSR 96, 55–86. [2] Robbins, S. (2014) EPSL 403, 188198. [3] Xie, M. & Xiao, Z. (2023) EPSL 602, 117963. [4] Neumann, G. A. et al. (2015) Science Advances 1.9, e1500852. [5] Stoffler, D. & Ryder, G. (2001) SSR 96, 9–54. [6] Haskin, L. et al. (1998) MPS 33, 959-975. [7] Nemchin, A. et al. (2021) Geochemistry 81.1, 125683. [8] Osinski, G. R. et al. (2023) Reviews in Mineralogy and Geochemistry 89.1, 339–371. [9] Richardson, J. (2009) Icarus 204, 697–715. [10] Minton, D. A. et al. (2015) Icarus 247, 172–190. [11] Byrne, C.J. (2016) Springer, Cham. [12] Maxwell, D. E. & Seifert, K. (1974) Defense Nuclear Agency, Washington. [13] Croft, S. (1980) Proc. LPSC 11, 2347-2378. [14] Huang, Y. H. et al. (2017) JGR 122, 1158–1180. [15] Michael, G. et al. (2018) Icarus 302, 80-103. [16] Norman, M. et al. (2006) GCA 70, 6032-6049. [17] Blevins, A. et al. (2025) JGR 130, e2024JE008722. [18] Cernok, A. et al. (2021) Communications: Earth and Environment 2, 1–9. [19] Bottke, W. & Norman, M. (2017) Ann. Rev. Earth Planet. Sci 45, 619-647. [20] Morbidelli, A. et al. (2018) Icarus 305, 262-276. [21] Morbidelli, A. et al. (2012) EPSL 355, 144-151. [22] Orgel, C. et al. (2018) JGR 123, 748–762.