Crater Count Dating: Self-Secondaries Reduce Age Estimates

A standard method for inferring the ages of planetary surfaces continues to be plagued by bad assumptions.

We’ve reported before about the problem of “secondary craters” in crater-count dating.* Planetary scientists have long used crater counts to estimate the age of a surface. In short, the more craters, the older the surface. That seemed reasonable until secondary cratering smudged the theory.

Earth’s moon from Cassini, 1999 (NASA)

A single large impact (the primary) could launch up to a million pieces of ejecta (the secondaries) that fall back down, forming additional craters – all from that single event. Some large pieces could orbit for centuries before falling, and some could even travel between moons, messing up the ages of different bodies entirely. These realizations have tended to vastly reduce estimates of surface ages – not increase them. The assumption of ‘one impact = one crater’ is not necessarily true.

Planetary scientists went to work trying to identify secondary craters, and distinguish them from primaries, in order to improve the accuracy of their estimates. The project has not worked well. Many times they appear indistinguishable unless other assumptions are made. Now, a new paper in Icarus by Plescia and Robinson adds another complication: the problem of “self-secondaries.” These can mislead scientists into thinking that multiple phenomena from a single impact were long ages apart, when they in fact were related by the same impact. The Abstract explains:

Giordano Bruno is a lunar farside Copernican-age crater. The relatively few superposed impact craters on its floor and ejecta blanket and extensive bright rays indicate its youth. High-resolution Lunar Reconnaissance Orbiter Camera images reveal that the frequency and the characteristics of the cumulative size-frequency distributions of the small-diameter impact craters and their morphology vary across the clastic ejecta and impact melt covered surfaces. Crater frequencies (N(10) defined as ≥10 m km⁻²) vary by a factor of 10s to 100s (N(10) 2 -700) across the ejecta blanket, and between the ejecta and the melt deposits. Numerous craters on the ejecta blanket are degraded and buried by debris and impact melt, indicating that these partially buried craters formed during the deposition of the ejecta and prior to or during the emplacement of the impact melt. From geologic relations and crater statistics we conclude that a significant fraction of the craters observed on the ejecta blanket and the melt were formed during the cratering process itself and represent “self-secondaries.” Further, we conclude that these craters do not represent an extra-lunar primary impact production population. Self-secondary craters are formed by material launched into near-vertical trajectories and having velocities such that their flight time is sufficiently long that the bulk of the clastic ejecta and impact melt are deposited before that material impacts the surface. The presence of a significant number of self-secondary craters on the ejecta makes the determination of relative and absolute age dates problematic. Crater counts would indicate an inappropriately old age. Using data for craters on the melt surfaces and for small-diameter bright ejecta craters, an absolute model age of 1 Ma is estimated. This age is considerably younger than that estimated by other studies and probably represents a maximum age.

How old is this crater? Giordano Bruno from Lunar Reconnaissance Orbiter, 2017 (NASA)

What they’re saying is that ejecta from the main impact can fly vertically above the surface, and take long enough to fall down that geological processes can degrade the ejecta blanket and harden the melted material before the “self-secondaries” land on it. A scientist would be tempted to conclude that the ejecta blanket is old, and the secondary impact is young, with a long time having passed between. Not necessarily; the ‘appearance of age’ could be fake. The series of events could have been relatively rapid.

Another assumption, that of incoming impactor rate, is also non-empirical. Nobody has lived long enough to watch how fast new impactors arrive on a surface. And even if scientists had good data on how many new craters are currently forming on the moon or Mars, there’s no way to know if the rate has been constant for long periods of time. Scientists assume that ‘space weathering’ darkens material over time, but the rate is hard to constrain, resting on other assumptions (such as steady states).

Scientists can deduce the order in which some impacts occurred, by looking at craters on top of other craters. But as we reported in a previous article on this, it’s doubtful that crater populations can reveal much of anything about long passages of time.

J. B. Plescia and Mark S. Robinson, “Giordano Bruno: Small Crater Populations – Implications for Self-Secondary Cratering.” Icarus 19 Oct 2018, https://doi.org/10.1016/j.icarus.2018.09.029.

*Previous articles on crater count dating:

18 July 2018: “Time to Revisit the Lunar Dust Problem?”

12 October 2016: “Moon Just Got 100-fold Younger”

2 March 2014: “Record Impact on Moon Ups Cratering Rate Estimates”

22 May 2012: “Crater Count Dating Still Unreliable”

3 April 2011: “Assuming Reality: Can Crater Dating Be Tested?”

25 July 2010: “Dating of Impacts and Impacts of Dating”

25 September 2007: “More Impacts on Crater Count Dating”

A million years is still a long age estimate, but notice they are calling it a maximum age – it could be far less. If this dating method has been so plagued up till now by unexamined assumptions, we need to ask, what other “unknown unknowns” remain? Let this be a lesson on all dating methods. You can’t watch a million years, or a billion years, without making assumptions. Assumption is the mother of all failures, a maxim goes: why? Because to ASSUME makes an ASS* of U and ME.  (*Ass: A long-eared, slow mammal related to the horse; an onager, figuratively, a stupid or stubborn person.)