What We Publish

Periodic impact cratering and extinction events over the last 260 million years

Michael R. Rampino and Ken Caldeira · Monthly Notices of the Royal Astronomical Society 454, 3480-3484 · 2015

Key finding. Circular spectral analysis of 37 revised impact-crater ages spanning 15 to 254 million years ago yields a significant cycle of 25.8 ± 0.6 million years, and the same method applied to the eight recognized marine extinction events of that interval yields a significant cycle of 27.0 ± 0.7 million years at a similar phase.

Two stacked line plots of the R-statistic from circular spectral analysis against trial period in millions of years, from 5 to 50. In each, a blue line is the analysed record and an orange line is the 95 per cent confidence level from 10,000 synthetic data sets. The upper panel, for 37 impact-crater ages, has blue peaks rising above the orange line near 18 and 26 million years. The lower panel, for eight mass-extinction ages, has a single tall blue peak crossing the orange line near 27 million years, reaching about 0.77 against a confidence level near 0.58.
Both records carry a signal near the same period. The extinction spectrum (lower) has one dominant peak at 27 million years and a much higher R-statistic than the crater spectrum (upper), whose peak sits at 25.8 million years — the periodic component explains more of the extinction record than of the cratering record. Figure 1 from Rampino and Caldeira (2015), Monthly Notices of the Royal Astronomical Society 454, 3480-3484. Reproduced under author reuse rights. Extracted from the published PDF and resized for web display.

What question did this research address?

For thirty years, studies have reported that terrestrial impact craters formed in the last 260 million years arrive periodically, with cycles variously estimated between 26 and 36 million years. Other analyses of the same record found no significant cycle at all. The disagreement is partly methodological and partly a matter of which craters are included.

Crater ages have since been substantially revised — the sixteen ages used in the founding 1984 study have all changed, some by as much as 60 million years. This paper asked whether periodicity survives in the updated record, and whether any cycle in cratering matches the cycle claimed for mass extinctions.

What did we find?

The crater record was restricted to the 37 craters older than 5 million years whose ages carry 1-sigma errors of 10 million years or less. Very young craters were excluded because recent impacts are over-represented in the record, which would bias the phase.

Circular spectral analysis — a method suited to unevenly spaced event times with no amplitude information — gives the crater record a peak at 25.8 ± 0.6 million years, above a 95 per cent confidence level built from 10,000 synthetic crater sets matched to the real one in number and in long-term trend. The most recent maximum falls at 16.0 ± 1.3 million years ago.

A second peak at 18.4 million years also clears the confidence level, but is interpreted as a false peak at two-thirds the period of the 25.8-million-year cycle rather than an independent signal.

The eight significant marine extinction events of the last 260 million years, redated to the 2012 geologic time scale, give a peak at 27.0 ± 0.7 million years with its most recent maximum at 11.8 ± 1.0 million years ago. The periodic component explains more of the variability in the extinction record than in the cratering record.

The two signals share a phase as well as a period. Fitting the extinctions at the craters' 25.8-million-year period puts the most recent extinction maximum at 14.7 million years ago, 1.3 million years from the crater estimate; fitting the craters at the extinctions' 27.0-million-year period puts the most recent crater maximum at 13.0 million years ago.

The agreement does not rest on the one impact-extinction pair already established. Removing the end-Cretaceous extinction leaves the extinction period at 27.0 million years and its phase unmoved; removing the Chicxulub crater leaves the crater period at 25.8 million years.

Smoothed crater ages show 11 apparent peaks over the interval, at least 5 of which fall close to significant extinction peaks — 6 if two additional possible extinction events are counted.

Why does it matter?

A shared period in two independently dated records is a stronger claim than a period in either alone, because the sources of error differ. Crater ages come from radiometric dating of impact melt; extinction dates come from the marine fossil record and the geologic time scale.

A roughly 26-to-27-million-year clock in Earth's impact history would point to something outside the solar system setting the pace, most likely periodic perturbation of the Oort cloud producing comet showers. The alternative — that the pattern is an artifact of a short, incomplete, and imprecisely dated record — remains live, and the paper claims viability for the hypothesis rather than proof of it.

Citation

Michael R. Rampino and Ken Caldeira (2015). Periodic impact cratering and extinction events over the last 260 million years. Monthly Notices of the Royal Astronomical Society 454, 3480-3484.

Related