Supporters of the Younger Dryas impact hypothesis hold that about 12,900 years ago a comet or its fragments struck or exploded over North America, set off abrupt cooling, contributed to the extinction of mammoths and other large animals, and disrupted the Clovis people; many popular retellings add that the oval Carolina bays of the Atlantic coastal plain are the scars. Critics in archaeology and geology answer that the claimed impact layer is not reliably dated to a single moment and that the bays are older, long-lived lakes. This is a live dispute in peer-reviewed journals, and both sides publish.
The hypothesis entered the literature in 2007. In the Proceedings of the National Academy of Sciences (https://pmc.ncbi.nlm.nih.gov/articles/PMC1994902/), R. B. Firestone, Allen West, James Kennett and colleagues described a carbon-rich black layer at about 50 Clovis-age sites, with megafauna bones and Clovis tools below it and not within or above it. They reported peaks of magnetic grains with iridium, microspherules, charcoal, soot, carbon spherules, glass-like carbon containing nanodiamonds, and fullerenes with extraterrestrial helium, and proposed that one or more large, low-density objects exploded over northern North America. They wrote that the shock wave, heat and related effects contributed to megafaunal extinctions and to adaptive shifts among PaleoAmericans, wording that stops short of a Clovis extinction. Their abstract states that the layer "also extends throughout at least 15 Carolina Bays." That sentence is where the bays enter the scientific version of the story.
Supporters have since built on that start. In 2015, again in PNAS (https://pmc.ncbi.nlm.nih.gov/articles/PMC4538614/), James Kennett and colleagues applied Bayesian statistics to 354 dates from 23 sedimentary sequences in 12 countries on four continents and reported a modeled age for the boundary layer of 12,835 to 12,735 calibrated years before present at 95 percent probability. They argued that the dates are consistent with synchronous deposition within about a century, that the range overlaps a platinum peak in the Greenland ice sheet, and that it matches the onset of Younger Dryas cooling in six key records. Firestone and West are among its co-authors, so the 2015 paper is the original team, joined by others, defending its own timeline.
Critics answer on the same ground of dates. In PNAS in 2014 (https://pmc.ncbi.nlm.nih.gov/articles/PMC4040610/), David Meltzer, Vance Holliday, Michael Cannon and D. Shane Miller examined the age basis of the boundary layer at the 29 sites and regions where proponents reported it. They found that several had no age control, others had irrelevant or statistically flawed ages, and only three fell inside the window that proponents themselves defined. They concluded that the hypothesis fails the test of an isochronous event. The two PNAS papers, a year apart, test the same claim with different handling of the same kind of evidence, and neither side has withdrawn. A reader can set the 354 dates against the 29 sites and see where the disagreement lies: in which dates count, and how their uncertainty is treated.
The Carolina bays are a separate question with its own record. In 2007, at the Southeastern Section of the Geological Society of America, Andrew Ivester, Mark Brooks and Barbara Taylor presented optically stimulated luminescence dates from bay sand rims in Georgia and South Carolina (https://gsa.confex.com/gsa/2007SE/webprogram/Paper119179.html). From a cumulative 45 dates they reported phases of shoreline activity at about 12 to 50 thousand years ago, 60 to 80 thousand, and 120 to 140 thousand, with some activity in the Holocene. Where bays have several concentric rims, the rims grow younger toward the centre, which the authors read as confirming that the bays "are not single-event features" but evolved through episodic processes over a long period.
A later study followed one bay through time. In 2014, Christopher Moore, Brooks and colleagues described Herndon Bay in Robeson County, North Carolina (https://gsa.confex.com/gsa/2014SE/webprogram/Paper237385.html), where LiDAR shows a regressive sequence of sand rims and a basin that migrated more than 600 metres to the northwest. Single-grain luminescence placed the formation of successive rims at about 31.8, 29.6 and 27.2 thousand years ago. The authors concluded that the bays are oriented lakes shaped by lacustrine processes, able to scour and migrate through their own basins while keeping their oval shape. Moore is also a co-author of the 2015 Bayesian paper in support of the impact layer, which shows that accepting a boundary layer and reading the bays as old lakes are separable positions.
Some impact supporters have moved the bays to a different event altogether. Michael Davias, whose LiDAR survey reports about 50,000 measured bays arranged in a broad annulus focused on Michigan, presented a slide deck to the Geological Society of America in 2019 (https://gsa.confex.com/gsa/2019AM/webprogram/Handout/Paper332326/81-1_Davias_Slide_Deck.pdf) proposing that the bays are features within ejecta from a highly oblique impact about 788,000 years ago, centred on Saginaw Bay, Michigan. He describes the bays as emplaced in sheets of mass flow, not as primary craters, and names a test that would falsify the idea: burial dating of terraces under the bays. In his model the bays are far older than the Younger Dryas, so they cannot be scars of a Clovis-era comet.
The Younger Dryas impact hypothesis is a live scientific argument. Its supporters report a dated boundary layer with impact proxies on four continents; its critics report that most of those dates fail an isochronous test. The Carolina bays are a separate matter: the 2007 paper reported its black layer, dated to approximately 12.9 ka and contemporaneous with the abrupt onset of Younger Dryas cooling, in at least 15 bays, but luminescence dating places their rims across tens of thousands of years, and even an impact proponent who maps the bays dates them to a far older event.
What Stays Contested
Whether a cosmic event occurred near 12,800 years ago, and whether it contributed to cooling or extinctions, remains disputed in the journals. What the dating does not support is the popular link that makes the Carolina bays the comet's craters. None of the cited papers reports that Clovis people were wiped out by an impact.