When an impact crater forms, it ejects numerous fragments at high speeds, some of which fall back on the surrounding continuous ejecta deposits and form self-secondary craters (i.e., self-secondaries). Since crater statistics on continuous ejecta deposits of impact craters is important in determining model ages for individual craters, the presence of self-secondaries may complicate age interpretations. While abundant self-secondaries have been observed on the Moon and Mercury, their spatial distribution on Mars and their potential effect on martian crater chronology remain uncharacterized.

Key Findings of the Study

We examined six young martian complex craters with layered ejecta deposits. According to the number of visible ejecta layers, martian layered ejecta deposits are classified as single-layered ejecta (SLE), double-layered ejecta (DLE), and multiple-layered ejecta (MLE). DLE deposits are further classified into two types.: Type 1, which has more uniform inner ejecta boundaries, and Type 2, which exhibits more sinuous flow patterns similar to those of SLE and MLE. Among the selected craters, there are one SLE crater, two Type 1 DLE craters, and three MLE craters.

High-resolution geological investigations around the selected craters reveal that self-secondaries are generally rare in SLE or MLE deposits and are mostly confined to crater rims (Fig. 1). Crater densities across different azimuths and radial distances in SLE or MLE deposits are comparable, suggesting that self-secondaries are a negligible population in SLE or MLE deposits. In contrast, a small number of self-secondaries are observed in Type 1 DLE deposits, where crater densities show a recognizable nonuniform distribution. Among martian craters with diameters of larger than 3 km, only six have Type 1 DLE deposits and formed in the Late Amazonian. Given that the proportion of self-secondaries in visible crater populations should decrease with time, our findings suggest that self-secondaries are not a major concern in age estimation for most martian craters, except for the few young craters with Type 1 DLE deposits.

To further explore the origins of these distribution patterns, we conducted numerical simulations to model the theoretical landing positions of fragments forming self-secondaries. When we assume near-vertical ejection angles, which were previously thought to be common for self-secondaries, the model predicted a preferential deposition in the western ejecta of the parent craters and an absence in the east (Fig. 2a). But this was not what we observed. In reality, self-secondaries do not avoid occurring on the eastern rims or ejecta deposits of the investigated craters, suggesting that fragments forming self-secondaries did not uniformly have near-vertical ejection angles.

We also simulated fragment ejection at angles that are slightly larger than those of normal excavation flows, resulting in a widespread distribution of self-secondaries on continuous ejecta deposits (Fig. 2b). Yet again, our observations showed that self-secondaries were absent in SLE and MLE deposits and preferentially occurred on crater rims. This suggests that most self-secondaries did form during the emplacement of layered ejecta deposits and were subsequently erased by the advancing ejecta layers. This interpretation further implies that the duration of emplacement of SLE and MLE deposits should be comparable to or slightly longer than flight times of fragments forming self-secondaries.

Unlike the scarcity of self-secondaries in SLE or MLE deposits, the prominent existence of self-secondaries in Type 1 DLE deposits indicates that DLE was emplaced slightly earlier than SLE and MLE. This timing likely allowed some self-secondaries to survive. The pre-impact target of Type 1 DLE craters contained larger amounts of volatiles, such as water ice, which contributed to the higher emplacement speeds and, consequently, earlier emplacement than SLE and MLE deposits. Based on the observations of well-preserved chains of secondaries in Type 1 DLE deposits and numerical simulation of fragments forming self-secondaries, we provide estimates for the emplacement speeds of martian layered ejecta deposits. For Type 1 DLE deposits, their emplacement speeds may approach or even exceed the speed of sound in current martian atmosphere.

Implications of the Findings

Our study first systematically investigates the distribution pattern of self-secondaries on Mars. By demonstrating that self-secondaries have minimal effect on crater statistics derived from martian continuous ejecta deposits, we reinforce the reliability of age determination technique using crater statistics on Mars. Additionally, we suggest that the formation of self-secondaries may not require tightly constrained conditions, and we offer estimates for the emplacement speeds for martian layered ejecta deposits, with Type 1 DLE deposits possibly reaching speeds close to or even exceeding the speed of sound in current martian atmosphere. These findings not only refine our understanding of formation mechanism of self-secondaries but also offer valuable insights for future studies and impact cratering simulations on Mars.