The Younger Dryas Impact Hypothesis: A Guide For The Perplexed - Graham Hanco*ck Official Website (2024)

Part I

Preface

This article was written over a period of ~19 months, during which time there have been several developments in the YDIH. In the most significant development, the central group of YDIH critics, corralled by Mark Boslough, chief enforcer for the impact mafia, published a behemoth of a “gish gallop” masquerading as a “comprehensive refutation” to the YDIH. Despite its exhaustive length, (~96,000 words) it suffers from the same issues that all of this group’s previous contributions do: They don’t offer credible alternative explanations for the physical and geochemical evidence for cosmic impact found at the Younger Dryas boundary. Their only relevant arguments (there are plenty of irrelevant ones including dozens of fallacies) are to call into question the YDIH team’s interpretation of the evidence, saying that it may be all caused by other independent processes, but refusing to elaborate. Luckily, this article does not need much of an update, and Dr. Martin Sweatman has written a comprehensive point-by-point rebuttal on his blog Prehistory Decoded that sufficiently addresses some of the more specific points made in their failed refutation. While this article addresses a few of those points, I would appeal to people who enjoy this article to also read Martin’s responses. For those who have not followed the debate closely over the years, and find the 300+ entries of the comprehensive YDIH bibliography intimidating, this article aims to serve as a comparatively concise guide to the YDIH. It provides commentary and analysis of many arguments made against the YDIH over its 16+ year history and demonstrates that, contrary to repeated claims of its demise, it is actually in a stronger position than ever.

Introduction

Since it was first officially proposed in 2007, the Younger Dryas Impact Hypothesis (YDIH) has attracted significant controversy. The hypothesis claims that a major extraterrestrial impact affected much of the world ~12,800 years ago, initiating a ~1,200-year period of cooling called the Younger Dryas that had many environmental consequences. Around the time of the Younger Dryas onset, more than 50 genera of megafauna (animals heavier than 44 kg) disappeared from the fossil record in a geological instant. At the same time, Clovis fluted projectile points also appear to have disappeared, though it is important to note that ‘Clovis’ refers to a specific technology or method of manufacturing stone tools, and not to a culture or a people. The people who manufactured Clovis points were not wiped out. Many different anomalous objects were found in the layer of sediment that dates to the onset of the Younger Dryas in many parts of the world, including:

• Magnetic grains and microspherules

• Strange carbon spherules

• Fragments of melted carbon that resembled glass

• Exotic charcoal and soot

• Multiple species of nanodiamonds and exotic graphene polymorphs

• Fused quartz grains, shocked quartz, and melt glass (limited sites)

Many of these proxies have associated with cosmic impacts and intense wildfires in the past, but some have not, which is one of the reasons for the strong pushback by critics. All this means is that we are dealing with cutting edge science here; airbursts of the type proposed for the YDIH are poorly understood, and all of the diagnostic proxies -can- be produced by airburst events; it is just that they haven’t yet been shown to within the scientific community and the field of planetary science.

The YDIH suffered significant setbacks in its infancy when a few independent groups were unable to replicate key results from the original sites, leading to its premature rejection. Fortunately, over the next decade these early replicability issues were shown to be due to either misunderstandings of the original claims, or methodological flaws, and were largely resolved. Many of these early issues are covered extensively in this article. Over the years we have gained significantly more details and understanding of the proposed impact proxy evidence, and this has gradually convinced a considerable number of scientists of the YDIH’s merit, at least those who aren’t ideologically opposed and who keep up with the literature. For every vocal critic of the YDIH, there are at least a dozen silent supporters within the scientific community, scores of which have reached out to proponents to offer supporting words. It has also received significant attention from the general public, who are primarily introduced to it in popular culture by figures such as Graham Hanco*ck and Joe Rogan, rather than scientific literature. This tends to draw the ire of authoritarian “science communicators” who see themselves as the anointed ones, see people like Graham Hanco*ck as existential threats to their carefully woven narratives. The reality is, many of these “science communicators”, particularly Mark Boslough, are merely science deniers who have devoted their lives to worshipping and enforcing the prevailing paradigm on social media and discrediting novel hypotheses. As explained and cited in my previous article on this website, they see the public as inherently inferior rabble, incapable of properly understanding scientific data, and consider it their duty preach the gospel of scientism. They fancy themselves the priests of scientism, disseminators of the word of science to the unwashed masses. However, not everyone is convinced; pushback mainly comes from a small group of persistent, yet vocal and influential critics insist that the YDIH has no merit and has been conclusively debunked as pseudoscience. This could not be further from the truth; this review examines the evidence for the YDIH in depth, providing commentary on some of the early issues and arguments against the hypothesis, and demonstrating that the YDIH is far from debunked.

Many journalists and non-academic speculators have gone beyond published scientific literature in exploring the consequences of the YDIH, and there is a long history of vague narratives of catastrophe at the end of the ice age. For more than 500 years, well educated people (for their day) such as Edmond Halley, William Whiston, Isaac Newton and Charles Darwin have all agreed that some sort of major catastrophe occurred at the terminal Pleistocene. Contrary to critics’ like John Hoopes’ interpretation of this fact as evidence against the YDIH, this should actually strengthen the argument. All of those people could see the result of the catastrophe as plain as day with their own eyes, and now hundreds of years later, we finally have the instruments and scientific knowledge to examine the direct physical evidence for it.

The formal proposal of the YDIH in the scientific literature, which despite a thematic link is distinct from those vague narratives of catastrophe, marked the development of a specific, coherent, plausible and testable hypothesis based on abductive reasoning. Thus, these narratives are irrelevant to the YDIH. In a similar vein, promotion of the YDIH by controversial figures like Graham Hanco*ck has nothing whatsoever to do with the data and evidence for it. A common tactic of YDIH critics is to incorrectly frame the hypothesis as being primarily driven by people like Graham and various alternative history influencers, rather than the cold hard scientific evidence published in peer reviewed journals. Nothing could be further from the truth, as the debate has generated more than 200 peer reviewed scientific papers and approximately 100 conference presentations. Most people, especially the public, are unconvinced by these tactics.

Thus, analysis throughout this article includes detailed dissections of arguments, evidence and data, historical and chronological context for the YDIH debate, explanations of why many issues pertinent to the YDIH are still unresolved in science or based on shaky paradigmatic foundations. Occasionally, original arguments are made in conjunction with rampant speculation to parsimoniously explain or resolve long-running debates. Such speculation should not be seen as a positive claim, but rather a potential alternative explanation and/or thought experiment. It provides a broad context of peer reviewed literature on issues pertaining to, but distinct from, the YDIH to provide a counter narrative to the critics who seek to claim it has been debunked. It explores the history of the YDIH, the claims, arguments, and evidence for and against its various aspects, and most importantly, details why many of the claims of critics are simply nonsensical quibbling. Many of their claims are unsupportable because they assume a solid understanding of the simplest concepts; the most up to date data has shattered more than 50 years of paradigmatic development concerning the distinction between comets and asteroids. Areas of scientific inquiry study suffering from a significant lack of consensus or uncertainties that the YDIH can potentially help resolve are also periodically examined. Some of the points made will likely fit the definition of various logical fallacies, but there is another type of logical fallacy called the fallacy fallacy, where just because an argument is fallacious, that does not make it wrong. Regardless, the fallacy to page ratio of this article is far lower than Holliday et al.’s ‘comprehensive refutation’. Martin Sweatman is leading a parallel effort to directly respond to that paper in the peer reviewed literature, which I and several other authors are also involved in; this article has a much broader focus than just their paper.

The Younger Dryas

~12,800 years ago, a prolonged cooling event known as the Younger Dryas (YD) occurred throughout the Northern hemisphere and much of the Southern hemisphere, marking a return to late glacial conditions for ~1200 years (Broecker et al. 1988). Since it was first discovered by Scandinavian paleobotanists in the early 1900’s and named for Dryas octopetala, a cold-adapted alpine flower that flourished in the glacial conditions of the YD, it has captured the imagination of countless scientists. Following several thousand years of warming throughout the Bølling-Allerød, temperatures in areas of the Northern hemisphere rapidly plummeted by ~8°C in as little as a few decades, up to a century (Alley 2000). However, while these temperature changes seem to have occurred over a few decades, atmospheric changes occurred more rapidly, and in some places hydrological changes occurred in as little as 1 year or less, as recorded in lake varves (Neugebauer et al. 2012). After this ~1200-year epoch of glacial conditions, temperatures skyrocketed by up to ~10°C in around a decade, a change even more rapid than at the YD onset (Dansgaard et al. 1989; Alley 2000). Similar changes, though less severe, are a normal part of ice age cycles, but they typically occur gradually over hundreds or thousands of years and are poorly understood; such sudden and extreme changes wreak havoc on the ecosystems they affect.

Vegetation species are usually suited for a specific climate; such rapid and significant changes in temperature and rainfall would lead to plant life in a particular region being no longer adapted for that region, meaning it would die off and be replaced with species more suited to the new conditions. The same is true of animals who rely on that vegetation for food or habitat; they would suddenly find themselves without any food, and seasonal behaviours like migration and hibernation would have experienced major interruptions. People would also have been forced to change their lifeways; as the plants and animals they relied on for food were killed off or forced out by the environmental changes, people would have needed to either adapt to the rapid changes or move to new areas where they could survive.

Ice core records show that ice ages, known as glacial periods, are often interspersed with major swings, or oscillations called Dansgaard-Oeschger (D-O) events (Figure 1), where temperatures rise or fall very quickly in a relatively short time. D-O events are characterised by a rapid rise in temperature, typically around ~5°C within a few decades, followed by prolonged cooling over several hundred to several thousand years. While aspects of the YD somewhat resemble previously recognised Dansgaard-Oeschger (D-O) cooling events, it also has some significant differences. Atmospheric methane concentrations fell rapidly from ~680 ppb down to ~460 ppb at the YD onset (Broecker 2006), where it remained throughout the YD before rapidly spiking to ~750 ppb at its termination (Figure 1). This leads some to believe it was likely initiated by an outsized “freak event” (Broecker 2006) unique among past deglacial processes. Furthermore, the YD onset occurred as Earth was receiving more energy from the sun than it had at any time in the previous ~65,000 years; a 1,200-year return to glacial conditions should be difficult to sustain in this scenario, adding to its intrigue.

Early on, the effects of the YD were thought to be global (Schneider et al. 1987; Peteet 1995; Clapperton et al. 1997; Andres et al. 2003), but later work caused this notion to fall out of favour (Lowell & Kelly 2008; Tibby 2012; Mendelová et al. 2020). However, recent data from South Africa and South America has shown it did significantly affect some areas of the Southern hemisphere (Stansell et al. 2010; Truc et al. 2013; Pino et al. 2019). Abrupt warming in the Northern hemisphere is often contradicted by abrupt cooling in the Southern, and vice-versa, an effect known as the ‘bipolar seesaw’ (Pedro et al. 2016; Renssen et al. 2018; Pino et al. 2019; Svensson et al. 2020). The warming of the Bølling-Allerød in the Northern hemisphere was countered in the Southern hemisphere by the cooling of the Antarctic Cold Reversal (ACR), which transitioned to warming just as the YD cooling was beginning in the north (Pedro et al. 2016). A recent pre-print pioneering a new paleoclimate proxy appeared to support the idea of YD cooling in New Zealand (Holdaway 2021), but the paper was rejected during peer review. Upon reading the peer review records, it appears to have been rejected for ostensibly no reason other than its findings conflict with the prevailing paradigm (McGlone 2021; Jara 2021). In summary, there is still quite a lot of unanswered questions about the nature of the YD period in the Southern hemisphere. For example, when exploring possible research projects for postgraduate studies, it became clear that many climate records for Australia show significant perturbations at the YD onset. In some records, hiatuses, or time skips in the record, larger than the peak of the glacial period are present (Falster et al. 2018), suggesting significant environmental changes were taking place. The result of these hiatuses, which can be caused by erosional events or periods of low deposition, is that we can’t claim to know what occurred in the time the tape recording was paused with any certainty.

As glacial ice melts as a normal part of the cycle, water pools up behind obstructions like mountains, moraines, ridges, or ice dams, eventually forming glacial lakes, of which there were many in North America throughout the ice age. Glacial Lake Missoula in Montana measured around ~7,800 km² and glacial Lake Agassiz, which covered up to 440,000 km² between 5 provinces and states (Manitoba, Ontario, Saskatchewan, Minnesota and North Dakota). Around the YD onset, significant volumes of water drained catastrophically from Lake Agassiz into the northern oceans, and this has long been thought to have caused the YD cooling by preventing the exchange of warmth between the ocean and atmosphere. In this scenario, a “freshwater cap” would have formed on the surface, acting as a barrier against the warmer, saltier ocean water, like how oil sits on top of water (Condron & Winsor 2012). This “freshwater cap” would have prevented warm water from the tropics reaching the surface, and cold water from the Labrador & Greenland seas would have sunk to the deep ocean (Broecker et al. 1988; Broecker 2006; Condron & Winsor 2012). However, to date there is no consensus on the route the meltwater took from Lake Agassiz around the YD onset, or even which ocean it drained into, though several potential routes have been explored (Murton et al. 2010; Condron et al. 2012; Rayburn et al. 2017; Leydet et al. 2018, Figure 2). Murton et al. (2010) find that Lake Agassiz was able to discharge north through the Mackenzie River into the Arctic Ocean (Figure 2), and Condron & Winsor (2012) agree. Conversely, Rayburn et al. (2017) find that Lake Agassiz likely discharged eastwards through the St. Lawrence River into the North Atlantic Ocean (Figure 2), and Leydet et al. (2018) agree. In total, at least 3 drainage routes into 3 different oceans are proposed (Figure 2).

These three alternative drainage routes have been intensely studied over the decades and are usually seen as mutually exclusive; the meltwater must have either drained through one or the other, and there is still no consensus as to which route is correct, because there is evidence supporting all. To further confuse things, a recent study compared a newly obtained suite of radiocarbon dates to a database of older ones (with “outliers” pruned out) in a statistical analysis; they concluded that the proposed drainage of Lake Agassiz occurred several hundred years after the YD onset (Norris et al. 2021; Young et al. 2021). Their conclusions are disputed by experts (Teller 2021) for various reasons, another example of the disagreement over almost every aspect of the draining of Lake Agassiz and its relation to the YD cooling. Teller (2021) cautioned that several of the dates pruned by Young et al. (2021) are only seen as anomalous because of limited dating resolution from around that time, not because they are invalid. In other words, they may have been improperly excluded, which would skew the statistical analysis. Notably, the academic supervisor who no doubt helped design the study and prune the dates is a published critic of the YDIH; make of this what you will.

In order to resolve the long-running dispute over the chronology and drainage route of Lake Agassiz around the YD onset, perhaps an alternative explanation should be sought. Consider this: What if meltwater was flowing out of all outlets simultaneously? But where did that much water come from you might ask? Perhaps it could have come from some sort of catastrophic event affecting the ice sheet? This scenario on a shallow level unifies the divergent hypotheses held by different groups of Agassiz experts as to the drainage route out of Lake Agassiz at the YD; if all three routes were flowing simultaneously, there is no need to argue over which one flowed when. The idea, as proposed by Young et al. (2021), that meltwater drained from Lake Agassiz during the Younger Dryas defies logic; the onset of rapid cooling in North America caused a well-documented glacial advance; if glaciers began to advance again, would this not reduce the volume of available meltwater, or cut off the supply? If glaciers were advancing and providing a steady supply of ice to reinforce ice dams holding back the meltwater, what would cause them to fail? After decades of disagreement and debate over hypotheses with insufficient explanatory power, alternative explanations must be considered. When examined in the context of a catastrophic impact event that may have affected the stability of the ice sheets, the idea that all routes drained simultaneously becomes more reasonable. If there was no overland meltwater pulse from Lake Agassiz at the YD onset, as Young et al. (2021) suggest, perhaps supraglacial or subglacial flows into the ocean in other locations were responsible. A large impact into the Laurentide ice sheet could certainly have triggered subglacial and supraglacial meltwater flows from the ice sheet in sufficient volumes, as originally proposed by Firestone et al. (2007). As it happens, a large impact crater of sufficient size was found near the margin of the Greenland ice, which will be discussed later in this review.

Firestone et al. 2007

In 2007, a multidisciplinary team of scientists published a suite of geochemical evidence supporting a major cosmic impact event at the YD onset, which they call the Younger Dryas Boundary (YDB). Collectively, the authors published works from before and after this paper has been cited by other scientists more than 150,000 times, meaning that the YDB team are influential and well-respected in their fields. Biographic information for some prominent authors of Firestone et al. (2007) is shown below (Figure 3). In 2016, many of these scientists became the core of the Comet Research Group (CRG) and were joined by other proponents of the YDIH.

The evidence presented in Firestone et al. (2007) to support their claims of a cosmic impact event at the YDB included elevated concentrations of the following geochemical signatures:

• A peak in magnetic grains at the YD onset, irregularly shaped, often subrounded, more abundant in northern sites than southern ones, enriched in titanomagnetite.

• A peak in spherical and sub-spherical magnetic grains, termed magnetic microspherules (Figure 4), of between 10 and 250 microns in diameter were reported in concentrations ranging from 97 per kg at Topper, South Carolina, up to 2144 per kg at Gainey, Michigan.

• An iridium peak ranging from 2 ppb, ± 90%, to 117 ppb, ± 10% in magnetic grains at the YDB, with the largest peak being >5000 times the crustal abundance. At 9 of the 14 sites they tested, the only Iridium peaks were at the YD onset and inside the black mat.

• Bulk sediments at the YD (not the magnetic fraction) were modestly enriched in nickel and contained detectable levels of iridium.

• In 14 of 15 sites, the largest charcoal peak occurs at the YD onset, with peaks ranging from 0.06 to 11.63 g/kg between sites.

• Aciniform soot, a geochemical marker found at the K-Pg boundary, peaks at 21 ± 7 ppm at the YD onset at Murray Springs, and at Carolina Bay T13 soot peaks at 1969 ± 167 ppm at the same time.

• Another K-Pg boundary marker, polycyclic aromatic hydrocarbons (PAHs) were found in the YDB layer, but nowhere else at Daisy Cave, Murray Springs, and Blackwater Draw.

• Black, highly vesicular, subspherical to spherical carbon between 150 microns and 2.5 mm were found only at the YD boundary at 6 of 9 archaeological sites, and 13 of 15 Carolina bays, reporting that work is ongoing to confirm early reports that they contain nanodiamonds.

• Fullerenes, which are nanoscopic ‘spheres’ of carbon lattice, containing extraterrestrial helium are associated with many ET impacts, including the K-T (now K-Pg) boundary, and were reported from 3 of the 4 archaeological sites analysed.

• Fragments of glass-like carbon up to several cm in diameter, examples of which have since been shown to contain nanodiamonds, recovered from all sites they examined in concentrations ranging from 0.01 to 16 grams per kg at the YD boundary.

• Geochemical signatures from multiple ice cores around the YD onset; a peak of iridium in the GRIP (Greenland Ice Core Project) ice core, and large spikes of ammonium and nitrate, geochemical markers of biomass burning, in the GISP2 (Greenland Ice Sheet Project 2) ice core reported by prior studies.

• A widespread sedimentary layer called the “Black Mat” (Figure 4), which is present at >60 sites across North America at the YD onset, directly above the layer containing impact proxies.

In addition to the above geochemical evidence, they also connected the well-accepted extinction of more than 30 genera of megafauna and the disappearance of the Clovis technocomplex to the impact event; both occurred abruptly at, or very close to, the YD onset. The discovery of Big Eloise, a fully articulated mammoth (except for her back legs) at Murray Springs closely associated with Clovis points is interpreted as representing the day the impact event occurred, in the same way as the Tanis fossil site with the Chicxulub impact (DePalma et al. 2021; During et al. 2022). On that fateful day ~12,800 BP, a Clovis hunter dragged a mammoth haunch to a nearby campfire to cook it but was interrupted by a major cosmic event unfolding in the sky above. Eloise’s leg remained next to that hearth until it was uncovered by a team of archaeologists at the Murray Springs Clovis site in 1966. Eloise’s black-stained bones were found draped in the organic-rich black mat, untouched by scavengers and in direct contact with the layer containing the impact evidence. While Firestone et al. (2007) does not reference Eloise by name, she is mentioned briefly in the final paragraph before the conclusion.

In their paper, Firestone et al. (2007) propose two hypothetical impact scenarios that could potentially explain the evidence they were finding; because no crater has been linked to the YDIH, all they could do was provide potential scenarios. One scenario proposes that a large (>4 km diameter), low-density comet struck the Laurentide ice sheet, which cushioned the blow and minimized or prevented crater formation. The other is an interaction with a particularly dense and violent region of one of the many annual meteor showers that still occur today; fragments of cometary debris of various sizes orbiting within meteor streams rained down all over the planet causing smaller-scale local impacts and airbursts over vast areas, a so-called comet swarm. This rain of cometary debris remains the preferred impact scenario by most proponents of the YDIH to this day. These two scenarios, and arguments around them, will be examined in more detail later in this review. It is worth noting that Firestone et al. (2007) was not written in a tone that states “this is the evidence, this is what happened, everyone else is wrong, we’re right, come and get us”, but rather more like “this is the evidence, we think it is interesting and worthy of further investigation, here are some potential scenarios that might explain it”.

An important factor that is often overlooked, ignored, or significantly downplayed, is the utility of YDB impact proxies as a geochemical datum that can be correlated over vast distances as a very precise method of absolute dating; currently, significant uncertainties between hundreds and thousands of years are built into radiocarbon dating, and the proposed geochemical datum could be used for absolute, very high-resolution stratigraphic calibration. These layers, known as “spherule layers”, are used to correlate many other impact layers over vast distances (Simonson & Glass 2004; Glass & Simonson 2012). The uncertainties inherent to radiocarbon dating are especially pronounced at the YD boundary, where a major perturbation in atmospheric radiocarbon resulted in an anomaly of ~400 radiocarbon years in the space of ~100 calendar years (Fiedel 2011). This is another feature of the YD that is unique amongst previous D-O climate oscillations. A synchronous deposition of impact material across North America & other continents should allow direct stratigraphic correlation over vast distances, overcoming the need for reliance on radiocarbon dating alone for establishing chronostratigraphic relationships.

It is fair to say that various aspects of the Younger Dryas impact hypothesis (YDIH), just as with any scientific hypothesis, have evolved substantially since the original 2007 paper as new evidence has come to light. Additional lines of evidence have been added, existing evidence has been expanded and reinforced, and some original lines of evidence have fallen out of favour. One major development in the form of the discovery and extensive replication of a global platinum anomaly at the YDB has become one of the strongest lines of evidence supporting the YDIH (Petaev et al. 2013; Moore et al. 2017, 2019; Thackeray et al. 2019; Pino et al. 2019; Moore et al. 2020). Another significant development was the discovery of a major Y-chromosome genetic bottleneck in humans around the YDB (Karmin et al. 2015; Sepulveda et al. 2022); the ratio of males to females was reduced to as low as 1:17, meaning some sort of event reduced the male population significantly at the YDB. The geographic extent of much of the initial evidence has been expanded, such as microspherules, nanodiamonds, wildfire evidence, and megafaunal extinctions. In addition to geographic expansion, much work has been done on resolving questions pertaining to how this evidence relates to an impact event. As mentioned, some evidence from the original paper has been put on the backburner, with no attempts made to replicate it, namely the radiation anomalies and polycyclic aromatic hydrocarbons (PAHs). Unfortunately, despite being mentioned in Firestone et al. (2007) and presented at the AGU conference where the YDIH was announced, the fullerenes containing ET helium have never been published, and no attempt has been made to replicate them. This evidence was brought to the table by Dr. Luann Becker, an early member of the YDB team, who has never published it, and has since disappeared entirely from the scientific community. The reasons for this strange occurrence are unknown, but it does cast a shadow of doubt on the YDIH and has left room for critics of the YDIH to spread claims of malfeasance. Thus, the fullerene & ET helium evidence should not be considered as evidence supporting the YDIH and will not be discussed in this review. Because the hypothesis has evolved over time, this has enabled critics to pick and choose from a variety of claims made by various people since 2007 and manufacture the least charitable Frankenstein of a strawman to attack. Consequently, the use of arguments against a hypothesis or individual claims that are unrelated to the YDIH as it stands today is a major component of attacks on the YDIH (Boslough 2012; Holliday et al. 2023). Contrary to their claims that the evolution of the YDIH since 2007 is a major weakness, progression and refinement of a hypothesis is an entirely expected, if not mandatory, component of good science. Unfortunately, people like Mark Boslough don’t live in the real world, instead substituting reality for the models they have spent their career producing, like pathological liars who begin to believe their own lies are absolute truth.

Debate Summary

Since 2007, almost 200 papers in peer-reviewed journals and ~100 conference presentations have directly addressed the Younger Dryas impact hypothesis (Young & Howard 2018). Many of these papers are critical of the YDIH or attempt to refute or debunk it. Since a “requiem” was declared for the YDIH in 2011 based on several failed replications (Pinter et al. 2011a), more than 100 papers, including extensive replication and additional evidence, have been published. At the very least, this demonstrates that Pinter et al. (2011a) sought to bury a hypothesis that was far from dead. In response to the firestorm of criticism, YDIH proponents have had become increasingly rigorous with their science, often exceeding the standards of evidence that would be required for their claims within the practice of Kuhnian ‘normal science’ (Kuhn 1962) within the prevailing paradigm. Much of the latest work in the last 5 years has yet to receive a single negative response from critics, even after the “comprehensive refutation”. The standards of evidence for these recent studies has garnered a resurgence of support throughout the academic community for the YDIH. A recent review of the impact evidence by Sweatman (2021a) in Earth-Science Reviews, the same journal that declared the requiem for the YDIH just 10 years earlier, opines that:

“Probably, with the YD impact event essentially confirmed, the YD impact hypothesis should now be called a ‘theory’.”

James Lawrence Powell is an eminent geoscientist renowned for his expansive reviews of published literature on important geoscience issues like anthropogenic climate change and the K-Pg (Cretaceous-Paleogene) boundary impact hypothesis. In a recent peer-reviewed summary, Powell (2022a) chronicles the controversial history of the YDIH, with a particular focus on early criticisms and failed replications. He concurs with Sweatman (2021) that the YDIH was prematurely rejected early in its history and should be re-examined and taken seriously as a legitimate hypothesis. Nevertheless, the most vehement critics of the hypothesis persist in their claims that the YDIH has been debunked, and that continued debate over it is a waste of time and effort (Holliday et al. 2023).

Following the recent publication by the Comet Research Group (CRG) of evidence that an unrelated airburst destroyed the Bronze Age city of Tall el-Hammam around ~3650 BP (Bunch et al. 2021), attacks on the YDIH have switched gears. Due to the air-tight nature of the impact proxy evidence, critics have pivoted to attacking the affiliations and integrity of some CRG members, such as Drs. Allen West and Malcolm LeCompte (Boslough 2022a). This abhorrent behaviour has been criticised by Powell (2022b), who resigned in protest from the Committee for Skeptical Inquiry (CSI) following Mark Boslough’s non-peer-reviewed attack on the CRG in their tabloid magazine, Skeptical Inquirer (Powell 2022c). According to Powell, Boslough’s ad-hominemladen smear piece violated almost every tenet of proper skepticism, and ultimately amounts to an ethics violation (Powell 2022b). Funnily enough, Powell’s article in Research Ethics denouncing Boslough’s conduct only served to further enrage him, changing his relationship with the YDIH from one of mild yet persistent frustration to an all-consuming hatred. He has since devoted his entire existence to doing as much damage as possible the YDIH by any means, which is why the recent comprehensive refutation is so extensive and bitter; they threw every argument possible, including the kitchen sink, hoping one or a few might stick. Further adding fuel to the fire, the tentative association between the TeH airburst and the destruction of Biblical Sodom has added a whole new dimension to the attacks, and significantly amplified the level of hatred and vitriol directed at the study; perhaps the most prolific and normalized form of hatred today is anti-religious bigotry. The way that Boslough in particular talks about this tentative association, it would not be surprising to learn that he has a jumbo-sized signed portrait of Richard Dawkins in his living room. The ‘PubPeer’ page for the TeH paper has been vandalized by Boslough and various anonymous tier 2 critics frequently over the past year with a myriad of garbage arguments attacking Christianity rather than the actual impact evidence.

Boslough is the most dedicated and vehement critic of the YDIH; in his mind, it is entirely without merit and even downright fraudulent. Though he contributed to multiple critical peer-reviewed papers between 2008 and 2015, his criticisms have devolved to such an extent that they are unfit for peer-reviewed literature (Powell 2022b). He was in the crowd for the first press conference teasing the 2007 publication and participated in several panel discussions over the next few years with proponents, and he has always maintained a negative stance. Boslough is an expert on computer simulations of airbursts, and opposes the idea that airbursts were responsible for the impact at the YDB. However, as any scientist knows, models and simulations are not reflective of reality, and are only as good as the data that is fed into them (Wit et al. 2012; Saltelli & Funtowicz 2014; Welsing 2015). The only major airburst for which sufficient data has been collected is the 2014 Chelyabinsk event; any input parameters for simulating other types of airbursts can only be considered useful for predictive purposes, not for refuting a hypothesis. Outside of an ever-shrinking community of Kuhnian ‘paradigm holdouts’ (Kuhn 1962), the evidence for YDIH is steadily changing the minds of legitimate scientists and commentators all over the world. As the nomological network of cumulative evidence from a wide range of disciplines, not just the geosciences, steadily grows in strength and breadth, the early damage inflicted by biased and fundamentally flawed studies is slowly being undone. Michael Shermer, editor-in-chief of Skeptic Magazine found a recent publication on the YDIH at Abu Hureyra (Moore et al. 2020) so compelling that he publicly apologised to Graham Hanco*ck and “adjusted his priors” concerning Graham’s theories (Shermer 2020). Likewise, volcanologist & Skeptic Magazine contributor Marc Defant initially was very critical of the YDIH, but made a spectacular about-face after taking the time to understand the claims being made about it (Defant 2020):

“… I read a recent book by James Lawrence Powell entitled Deadly Voyager: The Ancient Comet Strike that Changed Earth and Human History (2020). It is a superb book and has absolutely convinced me there were comet airbursts at the Younger Dryas. And the airbursts probably killed the megafauna which in turn, caused the Clovis culture to cease existence (partly by diminishing human numbers but also because there was no need to have Clovis spearheads that could kill nonexistent megafauna). I have not been keeping up with the debate since 2017, and so I was thrilled to see the new evidence that has come to light and the lack of scientific merit in the studies that attempted to dismiss the hypothesis.”

Dr. Annelies van Hoesel, at the time a Ph.D. candidate, published multiple scathing criticisms of the YDIH containing many problems and inconsistencies, perhaps due to pressure from her academic supervisors (van Hoesel et al. 2012, 2013, 2014, 2015). However, after subsequent collaboration with third-party independent researchers (Andronikov et al. 2016a), she coauthored a paper presenting evidence in support of the YDIH and has not addressed it since. She deserves major kudos for this; like Shermer & Defant, she rightly approached the YDIH from a position of skepticism but remained open-minded and was willing to adjust her prior viewpoint in the face of new evidence. Unfortunately, this is not the case for many YDIH skeptics. In a personal communication with the author, she claimed she never had a strong stance on the YDIH and has not kept up with the literature since she finished her Ph.D. in 2014 (Annelies van Hoesel, Pers. Comm. 2019). Clearly this suggests that she was not particularly interested in the subject but was assigned the project by her academic supervisor. At least one coauthor of her papers has published other work denouncing the YDIH, with fatally flawed methods, explored in detail later. The capacity of a scientist to change their mind when presented with new evidence is of vital importance; ideologues like Mark Boslough have demonstrated they are incapable or unwilling to do so.

Today, the YDB ejecta layer extends throughout much of the western hemisphere (Figure 5); there has been no supporting evidence found in east Asia or Oceania, mostly because it has not been sought there using appropriate methods. Very few, if any, peer-reviewed arguments against the YDIH have not been sufficiently dismantled or refuted. However, due to the premature rejection stemming from damage inflicted by early studies and associated media reports, most scientists do not keep up-to-date with the latest work (Annelies van Hoesel, Pers. Comm. 2019; Defant 2020; Powell 2022a).

In the author’s view, had Firestone et al. (2007) taken a different approach, their evidence may have been easier to swallow by the scientific community. While the megafaunal extinctions and human population effects are to be expected in the event of such a catastrophe, these were already intensely debated issues within the sciences. Many scientists participating in debates over these issues may have perceived the extraordinary claims of Firestone et al. (2007) as particularly disruptive; while this did make some groups pay attention, much of the attention was negative, and a few small groups of critics have never gotten over what they see as an affront to the paradigms they have devoted their careers to forging. Had the YDB team focused more on establishing that the impact event occurred, without venturing into its environmental and ecological consequences, it would have been better received by critics; encroachment into dogmatic paradigms like overkill and Clovis-first only served to steepen the slope of their uphill battle. Even though these consequences are indeed strong evidence for a catastrophe, in the author’s opinion, debate over the implications of the impact may have been more palatable once the impact itself had come to be widely accepted. Incorporating all of these events from day one conjures a mental image in the minds of critics of an outsized, ‘one-and-done’ impact event that destroyed the whole world in a few short seconds; this is far from what the YDB team were claiming and does not reflect the reality of the YD impact event, wherein a ‘rain’ of smaller impacts which were dispersed throughout time and space caused local or regional destruction.

Results obtained during the practice of ‘normal science’ that fail to conform to the dominant paradigm are treated as outliers, or even mistakes made by the researcher, rather than as refuting the prevailing paradigm (Kuhn 1962). However, as these ‘anomalous’ results accrue to the point that they can no longer be ignored, a ‘crisis’ occurs, wherein the old, discredited paradigm is replaced by a new one that incorporates the anomalies into its framework. This process is called ‘revolutionary science’ and it is playing out in real time with the YDIH, just as with the K-Pg impact ~3 decades ago, and countless hypotheses throughout history. Far from being ‘debunked’, cracks in the paradigmatic dam are spreading and widening as increasing numbers of scientists become proponents of the YDIH, discovering its merits after reexamining the evidence and taking the time to understand the claims.

So, did the YD impact event (YDIE) really happen? Can it really be invoked as a parsimonious explanation for the vast evidence of catastrophe at the terminal Pleistocene that has been gradually compartmentalized over decades by disparate, uniformitarian explanations? The following sections examine the arguments and evidence for the YDIH by incorporating a diverse breadth of literature from many disciplines to determine whether the YDIH has truly been debunked, or whether it has merit.

Part II

Magnetic Microspherules

It has long been established that impact events produce a variety of spherical to sub-spherical objects called spherules, or microspherules, depending on their size (Glass & Simonson 2012). These objects form as impacts or airbursts interact with the Earth’s surface, melting sediment and rock and throwing the molten droplets in all directions, with distances depending on the energy and characteristics of the impact. As the molten material travels through the atmosphere, often at supersonic speeds, it cools rapidly, resulting in a variety of forms ranging from perfect spheres, through teardrops, and even dumbbells, depending on how they spin through the air. Impact spherules have a range of compositions from glass to metal, with the more metallic ones exhibiting surface features that indicate that they cooled rapidly. Importantly, these features distinguish them from other naturally occurring magnetic spherules, which can be formed by a variety of non-impact processes. Magnetic microspherules were the very first line of evidence that led scientists to believe some sort of major catastrophic event had occurred at the YD onset; William Topping discovered them in the Clovis layer at archaeological sites and took them to Richard Firestone, a nuclear chemist at Lawrence Berkeley National Labs. Firestone approached the problem according to his expertise, running tests on the layer to determine how they formed, and discovered evidence of a nuclear catastrophe, possibly a supernova, based on radioactive isotopes (Firestone & Topping 2001). Over the next few years, they assembled a multidisciplinary team of experts and a significant body of evidence from the Clovis layer supporting an impact event, culminating in the 2007 paper (Firestone et al. 2007).

Soon after Firestone et al. (2007) was published, an independent group of archaeologists started work on trying to replicate the magnetic microspherule results but were unable to do so (Surovell et al. 2009). The authors who performed the analyses had no prior experience, and they did not research how other groups had done this type of work previously; instead, they relied exclusively on the methods reported by Firestone et al. (2007), without even contacting them for advice before they started (Holliday et al. 2014). This study was particularly devastating to the YDIH; the first attempt to replicate key evidence being unsuccessful cast a long shadow over it just 2 years after its proposal. However, by this time, another independent analysis was underway by C. Vance Haynes, who was able to successfully replicate the results, but nonetheless offered an alternative, non-impact explanation for the spherules; by way of some mysterious process or phenomenon, the normal influx of cosmic dust and micrometeorites had become concentrated into this discreet YD boundary layer (Haynes 2010). However, once again new evidence has advanced our understanding of the physical characteristics of micrometeorites, which based on ‘fresh’ urban micrometeorites, vary significantly from all types of YDB impact spherules. The next attempt to replicate the presence of magnetic microspherules in the black mat was by Pigati et al. (2012). This study also successfully replicated their presence, but they found additional peaks of magnetic spherules at the base of other “black mats” above and below the YD mat, and ultimately conclude that they formed by normal wetland processes.

Notably, none of these three early studies examined any spherules they found using electron microscopy, and so they were unable to determine anything about how they formed. In addition to magnetic microspherules formed during impact events, there are many other types such as naturally occurring authigenic (formed in place) spherules that form over extended periods of time, including framboids; these are known to form by normal chemical processes within wetlands and saltmarshes (Sawlowicz 1993; Duverger et al. 2021). By contrast, spherules formed during an impact event typically form within a few seconds, often within the impact plume, from melted and/or vapourised mixtures of terrestrial and cosmic material. Without performing analyses that allow determination of how a microspherule formed, and/or their chemical composition, no conclusions about a given hypothesis can be made. Had Pigati et al. used these methods, they may have discovered that the black mat dated to the YDB contained both terrestrial and impact spherules, while the non-YDB mats may have only contained terrestrial ones. To be fair, Firestone et al. 2007 also did not publish electron microscopy results for their spherules, meaning groups attempting to replicate them or test the impact hypothesis would need to conduct some basic level of research into microspherule formation. As admitted in Holliday et al. (2014), Surovell et al. (2009) did not do so; instead, they claim to have relied solely on the methods reported in the supplemental materials of Firestone et al. (2007). However, what they actually did is modify the magnetic separation protocols in ways that were fatal to their study and cut important corners. Even though Surovell et al. (2009)’s fatal errors were established in the literature just 3 years after they were published (LeCompte et al. 2012), their study has been regularly cited as evidence against it, even to this day (Powell et al. 2022a; Holliday et al. 2023).

Meanwhile, a third independent attempt to replicate YD microspherules was underway; this group, led by Malcolm LeCompte had done their due diligence by contacting Allen West to consult extensively on the methods he used for the Firestone et al. (2007) analysis. LeCompte attended an American Quaternary Association conference in 2009 to present the preliminary results of his investigation, and was placed alongside Todd Surovell, who was also presenting his results. LeCompte, who successfully replicated the spherule evidence, showed Surovell his results, informing him that he had not performed the crucial step of sorting by grain size in his sample prior to analysis (Malcolm LeCompte, Pers. Comm. 2022). After the conference, a months-long email exchange took place between LeCompte’s group and Surovell, who repeatedly denied they had made fatal mistakes; the exchange ended when LeCompte challenged him to a side-by-side comparison of each protocol using Surovell’s samples (Malcolm LeCompte, Pers. Comm. 2022). While the two groups never ended up comparing their protocols, I undertook a research project with my university in 2022 to determine how modifications of the protocols between the Firestone et al. (2007), Surovell et al. (2009), and LeCompte et al. (2012) studies may have affected their results. Of course, the experimental data produced affirms the claims made by LeCompte et al. (2012) as to why Surovell et al. failed. While this study has not been published yet as additional work is underway, I hope to have it published by mid-2025.

While the LeCompte team never made direct comparisons between protocols, they did go back and further refine them in collaboration with the rest of the Firestone team to improve the results. They also examined additional sites, and attempted to describe why Surovell’s replication failed, eventually publishing their results in 2012. LeCompte et al. (2012) reported a successful replication of the microspherule results at multiple sites, though with some minor differences; the spherules found by LeCompte were mostly smaller than those from Firestone et al. (2007), but electron microscopy nonetheless revealed their morphology and geochemistry were consistent with formation during an impact event. The paper also articulated 5 major deficiencies introduced by changes that Surovell et al. (2009) made to the Firestone team’s protocols, each one compounding the issues of the others, as follows:

• Deficiency 1: Sediment samples were too thick. The average thickness of samples in Firestone et al. (2007) was 2.3 cm, while Surovell’s averaged 11 cm, or 5x thicker. The YD boundary impact layer is as thin as a few millimeters in some places. The thicker the sample containing the thin YD impact layer, the more it is diluted by sediment on either side that does not contain the impact material. This would result in less spherules being present in each subsample sprinkled onto the microscope slide, which would have significantly influenced their results. Interestingly, this deficiency has been ignored by almost every single failed YDB replication performed since.

• Deficiency 2: Inadequate aliquot size. The amount of material examined by Surovell per sample was significantly less than the original study. Surovell often only examined 2 slides of subsampled material (maximum 4), or around ~10 mg of the <1 mm magnetic fraction. In contrast, Firestone examined a minimum of 100 mg of the <0.15 mm magnetic fraction, or ~10x that of Surovell, up to the entire magnetic fraction of >200 mg. Analysing only a few slides per sample can significantly exacerbate the issue introduced by deficiency 1, causing an inability to find many microspherules.

• Deficiency 3: Insufficient size sorting of grains. Firestone et al. (2007) sorted their magnetic fraction into separate portions by grain size using a very fine sieve with 0.15 mm mesh. This allowed them to isolate most of the microspherules from the rest of the magnetic grains by removing uninteresting or oversized material. Surovell et al. (2009) only size-sorted at 1.0 mm, meaning that while they were looking for spherules under the microscope, comparatively giant sand grains were obscuring the much smaller spherules from view. Spotting spherules on a slide is already a difficult task, and Surovell et al. (2009) made it significantly more difficult by drowning their spherules in uninteresting grains. Obscuring spherules behind grains almost 7 times larger than them would have significantly exacerbated the issues introduced by deficiencies 1 and 2.

• Deficiency 4: Perfect sphericity requirement. Firestone et al. (2007) published images of 4 spherules in their original paper, two of which were perfectly spherical, and 2 which had surface features like bumps and divots, meaning they were not perfectly spherical. Surovell et al. (2009) unilaterally made the arbitrary and unscientific decision to exclude all spherules that were not perfectly spherical, meaning that they likely discarded the the best candidates for impact spherules. During my experimental evaluation of the protocols, only 40% of the spherule candidates I found were perfectly spherical, meaning the Surovell team may have discarded more than half of their candidates. This arbitrary decision demonstrates a complete and total lack of understanding of how spherules form during an impact event, and how they can be used as evidence for one; spherules formed during an impact event are often not perfect spheres but can take a wide variety of aerodynamic forms such as teardrops, dumbbells, ovoids, and more. In fact, based on previous studies, perfect spheres are probably one of the less common forms.

• Deficiency 5: No electron microscopy. Determination of how a spherule formed requires detailed examination at very high magnifications that most optical microscopes are not capable of. Scanning Electron Microscopy (SEM) enables characterisation of a spherule’s surface morphology and geochemical analysis of the spherule’s elemental composition, providing further indication of how it formed. Without SEM or similar microprobe analysis, no credible claim can be made regarding how a spherule formed, as all candidates formed by any process essentially look the same under an optical microscope.

Shortly after the publication of LeCompte et al. (2012), Mark Boslough made his first accusation of fraud against Malcolm LeCompte; LeCompte’s study was published as an independent replication, but according to Boslough, because one coauthor also appeared on the original Firestone et al. (2007) paper, to claim it as independent was fraudulent. Dr. Albert Goodyear’s role in the LeCompte paper was limited to consulting on the stratigraphic location of the Clovis layer at Topper; he was the principal investigator of the site when LeCompte collected samples from there (Albert Goodyear, Pers. Comm. 2022). This was also his contribution to the Firestone et al. (2007) paper; he was not involved in laboratory analyses or data interpretation in either paper (Albert Goodyear, Pers. Comm. 2022). LeCompte performed all sampling and magnetic separation himself, and none of the SEM work or data interpretation was performed by anyone affiliated with the Firestone et al. (2007) team (Malcolm LeCompte, Pers. Comm. 2022). Despite being debunked at the time and dropped; this accusation was recently dredged back up as part of the most recent propaganda campaign against the CRG (Boslough 2022b). Considering this, it seems reasonable to highlight the following deception perpetrated by Surovell. In their 2009 study, they specifically state that the study was designed to replicate the magnetic microspherule evidence presented by Firestone et al. (2007), and to test the impact hypothesis. However, just 5 years later, their story had changed. Surovell published a response to LeCompte et al. (2012) by sneaking it in through the back door of a 2014 paper by Holliday et al., hidden away in the supplemental information. In his response, he now claimed that rather than replicating the spherules and testing the YDIH, his study was originally designed to test the methods of Firestone et al. (2007). Of course, this is an outright lie; if the goal was to test the methods, they would not have changed them in significant ways (Figure 6). Conversely, if the goal was to replicate the presence of magnetic microspherules as originally stated, the only requirement is ensuring the protocols are suitable to detect them, which Surovell failed abjectly to do. In the worst case scenario, this lie constitutes an intentional deceit. Even in the best case scenario, Surovell et al. moved the goalposts after their protocols were shown as deficient, perhaps to save face after being embarrassed live on stage at an AMQUA conference while presenting their results; when the LeCompte team asked him why he didn’t size sort properly, he hung his head in shame (George Howard, Pers. Comm. 2019). In either case, this breach of academic integrity should be enough to completely invalidate any and all contributions by these researchers to the subject of the YDIH; clearly they are more interested in advancing some sort of agenda against the YDIH than maintaining their integrity.

Since LeCompte et al. (2012)’s replication, all further studies that sought to replicate YD microspherules have been successful using various iterations of the LeCompte protocols, and successfully. Between 2012 and 2013, multiple studies fleshed out the specifics of the magnetic microspherules and sought to advance the discussion around them. They characterised the physical and chemical properties of YD spherules; how they formed, and how they compare to other spherules formed by cosmic, volcanic, anthropogenic and authigenic processes (Bunch et al. 2012; Fayek et al. 2012; Wittke et al. 2013; Wu et al. 2013). One study also produced a model that suggests up to ~10,000,000 tons of impact material was deposited during the event (Wittke et al. 2013). Examples of material recovered and/or analysed during this study are shown in Figure 7. Essentially, the takeaway from the 2012/2013 spherule studies was that YD microspherules are somewhat novel among impact spherules from prior studies in terms of their chemical compositions, which vary between sites.

Spherules formed by cosmic impacts are primarily composed of terrestrial material that has been melted and/or vaporized before condensing and/or cooling rapidly; sometimes they are mixed with cosmic materials from the parent body at extreme temperatures in the impact plume (Bunch et al. 2012; Genda et al. 2021). Notably, impact spherules are distinct from other cosmic spherules, which are mostly micrometeorites and zodiacal dust that accretes to Earth constantly, or ablates from meteoritic bodies as they melt during atmospheric entry. Cosmic spherules from meteoritic ablation or cosmic dust are also found alongside impact spherules in the YDB. Israde-Alcantara et al. (2012a) replicated YD microspherules in central Mexico along with multiple other YD impact proxies, extending the range of the impact event far southwards. Their study prompted a furious response from critics (Blaauw et al. 2012; Boslough 2012; Daulton 2012; Gill et al. 2012; Hardiman et al. 2012), who had likely just finished their circle jerk over the 2011 “Requiem” paper and other garbage science. Israde-Alcantara et al. (2012b) responded with a comprehensive rebuttal to all 5 comments, highlighting several inaccuracies and backing up their own claims concisely and effectively. Many subsequent studies have further replicated a peak of magnetic microspherules at the YD boundary: Andronikov et al. (2016b), an independent Russian group, Kletetschka et al. (2018), a CRG-associated group from Eastern Europe, Pino et al. (2019) from South America and Teller et al. (2019) from North America, the latter two resulting from collaborations with the CRG. All these groups found enhanced concentrations of impact microspherules at the YDB, with none above or below. This demonstrates that despite the early teething problems before standard protocols were established, impact microspherules are consistently reproducible at the YDB. Of course, this can only happen if the researchers use protocols that allow them to be reproduced. Whether Surovell et al. (2009)’s unfortunate modifications to their separation protocols were borne of ignorance, cutting corners, a combination of the two, or something more sinister remains to be seen. To put it into perspective, Surovell et al. (2009) are the only group in 15 years of research to fail their replication. Remember, the Haynes and Pigati studies replicated their presence, they just offered alternative explanations. In a just world, given that their findings have been demonstrably invalidated, Surovell et al. 2009 would be retracted.

Carbon Spherules & Glass-Like Carbon

Just as with the magnetic microspherules, the first independent attempt to replicate the carbon spherule evidence presented in Firestone et al. (2007) was unsuccessful (Scott et al. 2010). In their study, they examined sediment ‘sections’ collected from the California Channel Islands and found abundant carbon spherules and glass-like carbon. The problem is, they found this material consistently throughout many different layers in their ‘sections’, and their analyses determined that the carbon spherules were fungal sclerotia and termite feces. Yes, unfortunately you read that correctly; according to Scott et al. (2010), the microscopic carbonaceous spherules that the YDB team had been examining at extremely high magnifications with highly sensitive instruments, that contained billions of nanodiamonds, are actually just bug poop. How could the YDB team have made such a catastrophic mistake? This was a damning indictment of the YDIH, the implication being that its proponents were incompetent, and their hypothesis was based on a misinterpretation of the evidence. As the reader can imagine, the headlines write themselves, and this finding was another major hit to the credibility of the YDB team and the YDIH in general. However, I feel the need to clarify something… Bugs do not poop nanodiamonds, that’s… not a thing. So how could Scott et al. (2010) have come to this conclusion?

Wittke et al. (2013) examined their data more closely and made a troubling discovery in the fine print of the methods section; their sediment ‘sections’ were composites, manufactured by combining several different samples taken from up to 7 km apart into a single stratigraphic section, reminiscent of Frankenstein’s monster (Figure 8). Essentially, without very precise chronological control, every single layer in their manufactured section could theoretically be from the same exact layer sampled in different locations and stacked on top of each other. Worse still, it is theoretically possible that none of the layers in the manufactured section, or even all of them, date to the YDB. An even more thorough examination by Powell (2022a) of this study and others found that the Frankenstein stratigraphic section used by Scott et al. (2010) were also used in 2 other studies, both of which failed to find any of the respective YD impact proxies they were searching for. As with Surovell et al. (2009), the question of intentionality must, at the very least, be considered; was this Frankenstein’s monster of a sediment ‘section’ manufactured as part of an agenda to discredit the YDIH? Or was it simply accidental, resulting from shockingly poor scientific methods? What purpose does manufacturing a composite sample from multiple sites up to 7 km apart serve? Why can the sample not be taken from a single location? Speaking as a recently qualified archaeologist and geoscientist, and a current PhD candidate in geoarchaeologist, this is highly questionable. In fact, the recent retraction of the Hopewell Airburst paper (Tankersley et al. 2022) was justified in part because of a similar error, though much less egregious.

The table published in Wittke et al. (2013), with the benefit of hindsight and further studies since Scott et al. (2010), suggests that at least 2 samples in the Frankenstein section at least partially captured the YDB. Samples 28 and 29c, taken from more than 7 km apart, both contain nanodiamonds in addition to carbon spherules. Contrary to Scott et al. (2010)’s assertion that the carbon spherules were ubiquitous throughout every sample, this clearly demonstrates that the carbon spherules in the 5 other samples are different to those found in the nanodiamond samples. Importantly, at the time of Firestone et al. (2007)’s publication, YDB carbonaceous spherules were already known by the YDB team to contain high concentrations of nanodiamonds. Unfortunately, the analyses on the nanodiamonds had not been completed in time for the 2007 study. Since the findings of Scott et al. (2010) and the other studies that relied on the same Frankenstein stratigraphic section, carbon spherules have a 100% replication rate at the YDB. Israde-Alcantara (2012a, 2012b) replicated the YD carbon spherules and found that, for the most part, they were distinct from fungal sclerotia:

“Lake Cuitzeo carbon spherules have “smooth, glassy, highly reflective interiors with no evidence of filamentous structure observed in in fungal sclerotia”, consistent with previous observations. Although some carbon spherules at some YDB sites may be charred sclerotia, it is clear that not all are, because they can also be produced from the burning of resinous wood, as in wildfires.”

I am lucky enough to have examined YDB samples under the microscope myself, specifically for the purpose of isolating carbon spherules. While I have not confirmed the presence of nanodiamonds in any of my spherules, I have noticed there are at least 2 distinct types or species of spherule in my sample. One species has a hard shell that makes it difficult to crush, ranges from spherical to subspherical, typically between 0.3 and 0.5 mm, light vesicular interior when crushed, dark in color with a scabby appearance, and one example I found was fused together resembling a dumbbell. The second species is much larger, typically >1 mm, often perfectly spherical (but not always), light in color, a much smoother and more hom*ogenous surface, and very easy to crush (Figure 9).

Of course, without detailed examination using multiple instruments, my observations are just anecdotal, but at the very least they suggest there are multiple different types of spherules in YDB samples. Unfortunately, it seems Scott et al. (2010) did not examine them closely enough to notice a difference. Even if, for some reason, the YD carbon spherules containing nanodiamonds really were fungal sclerotia or bug poop, the presence of diamonds within them still provides evidence for an impact event. In this case, they could theoretically have been formed within existing spherules of bug poop/fungal sclerotia that were affected by the airburst shockwave. Remember, only a thin upper layer of sediment is shocked by an airburst as it disperses through the atmosphere, while in crater-forming impacts the shockwave penetrates deep into the bedrock.

There has not been a significant amount of work done on the carbon spherules; they are primarily of interest due to the nanodiamonds they contain more so than the spherules themselves. It has been acknowledged since day one that carbon spherules (Figure 10) like those presented in Firestone et al. (2007) can be formed by many processes, including normal forest fires. However, “normal”, non-YDB, pre-industrial carbon spherules have never been shown to contain nanodiamonds. Carbon spherules are not definitive evidence of a cosmic impact event, but when they contain nanodiamonds and are found alongside other impact proxies in the YDB layer, an impact event is the most parsimonious explanation. Thus, carbon spherules, particularly those containing nanodiamonds, still stand as evidence for the YDIH. However, what do other studies have to say about the presence of nanodiamonds at the YDB?

Nanodiamonds

As mentioned earlier, Firestone et al. (2007) mentioned that preliminary studies had shown that YDB carbon spherules appear to contain high concentrations of nanodiamonds. Subsequent studies (Kennett et al. 2009a, 2009b) announced the discovery of multiple nanodiamond allotropes (species/configurations) in the YDB at multiple sites. At the time, the YDIH was enjoying widespread interest, including from network television, and the discovery of nanodiamonds inspired PBS to fund an expedition to the Greenland ice sheet to look for nanodiamonds in the YDB. The expedition resulted in a NOVA documentary titled “Megabeasts’ Sudden Death”. The expedition was led by glacio-geochemist and Firestone et al. (2007) coauthor, Paul Mayewski, whose extensive knowledge of the ice sheet allowed him to locate the YDB in the ice and retrieve samples. The YDB-aged ice was found to contain very high concentrations of nanodiamonds (Figure 11); the sight was overwhelming for James Kennett, who was moved to tears, describing the find as one that every scientist hopes to experience, a true ‘eureka’ moment. The findings were subsequently published in the well-respected Journal of Glaciology (Kurbatov et al. 2010). However, behind the scenes, faceless critics were (allegedly) quietly emailing executives at PBS accusing the authors of spiking samples with lab-made nanodiamonds. Shortly after, the first published attempt to replicate nanodiamond evidence (using the Frankenstein samples from Scott et al. 2010) unsurprisingly did not find them (Daulton et al. 2010). Being the third failed replication in two years, this was enough to spook the invertebrate executives at PBS, who removed the documentary from their website and programming schedule; yet another major blow to the YDIH when it was still in its infancy. Today, “Megabeasts’ Sudden Death” can only be viewed on YouTube, preserved on several different channels including my own.

All subsequent studies by the YDB team (later becoming the CRG) and independent, even adversarial groups, have successfully replicated and characterised a significant and widespread nanodiamond peak at the YD boundary. The consistent replication of discreet YD nanodiamond spikes in samples that never passed through the hands of any CRG member rules out the possibility of fraud. An independent team successfully replicated the YD nanodiamonds at Lommel, Belgium, yet they claim they don’t provide unique evidence of an impact; they cite no supporting study or description of specific processes, but nonetheless feel comfortable claiming the nanodiamonds may have formed during wildfires (Tian et al. 2011). However, if that were the case, they should be ubiquitous throughout the geological record, yet they are not; even if it were argued that they only form under special circ*mstances in a wildfire, the obvious response should be “yeah, maybe like a wildfire that was caused by an impact or airburst?”. One year later, van Hoesel et al. (2012) further replicated the presence of nanodiamonds in the Usselo Horizon, the European version of the “Black Mat” (Haynes et al. 2008), but claim they post-date the YD boundary; they justify this by invoking the uncertainties inherent to radiocarbon dates, and the old wood effect, both of which are explored in a later section. However, analysis of their data and methods by Martin Sweatman reveals a few major issues with their interpretation and modeling; the reason they appear to post-date the YDB is because of undue and disproportionate weighting given to dates from several centimeters above the Usselo Horizon (Sweatman 2020a). All charcoal samples from within the Usselo Horizon layer, at the same depth where the nanodiamonds were found, date to the YD onset. One particularly egregious error made by van Hoesel et al. (2014) was the failure to re-calibrate several radiocarbon ages from other YDB sites that had been calculated using different calibration curves, effectively rendering them incompatible with others (Kennett et al. 2015).

Major advances in nanodiamond evidence came thick and fast in the next few years. In addition to the carbon spherules, Israde-Alcantara et al. (2012) replicated YD nanodiamonds in central Mexico and defended their interpretations well in their response to the barrage of criticism they encountered. Bement et al. (2014) confirm the presence of cubic nanodiamonds at Bull Creek, Oklahoma, advocating for their use as evidence for the YDIH, but acknowledging that claimed hexagonal allotropes are more consistent with graphene/graphane. They also report a couple of other nanodiamond spikes from late Holocene and modern sediments. As nanodiamonds or analogous materials are produced as a byproduct of many industrial and laboratory processes, their presence in modern layers is not surprising, and the late Holocene spike may be linked with an unknown airburst event. Kinzie et al. (2014) published a comprehensive review of nanodiamond evidence, finding that nanodiamond spikes are present in the YDB layer at 24 sites across 3 continents, replicated by 6 independent groups (Figure 12).

Allotropes of nanodiamonds found throughout the various studies include cubic nanodiamonds, hexagonal nanodiamonds analogous to lonsdaleite, n-diamonds and i-carbon, which are unique carbon allotropes with properties very similar to nanodiamonds (Kinzie et al. 2014). They conclude the notion that these nanodiamonds were formed by other processes are, while perhaps reasonable in a vacuum, are only tenable when ignoring all context of the other proposed impact proxy evidence found at the YDB. Such a notion also does not explain their presence in the same layer on multiple continents. In 2016, Molly Sexton completed her Master thesis, supervised by Leland C. Bement, wherein she characterised the stratigraphic context and performed a textural analysis of Bull Creek nanodiamonds from the YDB. Sexton found that alternative explanations for how YD nanodiamonds formed lacked merit, and the nanodiamonds she analysed most likely formed by chemical vapour deposition (CVD), consistent with a cosmic impact, citing prior studies coauthored by Daulton to support her claims (Sexton 2016). The findings that YDB nanodiamonds in Oklahoma were formed by CVD mirrors the conclusions of Israde-Alcantara et al. (2012a, 2012b) from central Mexico.

Finally, in 2017, Daulton et al., the only group who were unable to replicate YD nanodiamonds back in 2010 published their own ‘comprehensive review’ of YD nanodiamond evidence (Daulton et al. 2017a). They claim that all prior studies by seven different groups all misidentified graphene/graphane aggregates as lonsdaleite, and that they do not offer unique evidence of an impact; in other words, ‘only we are competent enough to reliably identify nanodiamonds, everyone else is mistaken’. They go on to confirm the presence of cubic nanodiamonds at the YD but cannot resist throwing in that they probably weren’t formed by an impact, because no other ‘recognised’ impact markers have been found associated with them. To be clear, these are weasel words. Implicit in this assertion is that no other claimed YD impact proxies are ‘officially’ a.k.a. ‘not by our side, the real experts’ recognised as impact markers. They further claim that the majority of YDB sites where nanodiamonds are reported only contain n-diamonds, and these are not conclusive evidence of an impact. They conclude by demanding the invention of more robust methods for identifying and characterising nanodiamonds before they are willing to accept that proposed YD nanodiamonds are really diamonds but suggest no such methods. They also begrudgingly admit that the YDB nanodiamond peak does strongly suggest some kind of ‘unique event’ occurred at this time but refuse to admit the possibility of an impact (Daulton et al. 2017a). Later in 2017 the same group published a 2-page summary of these findings in a separate journal (Daulton et al. 2017b).

A recent study from Chinese scientists certainly seeks to shake things up; essentially they claim that despite being studied and reported on for more than 30 years, n-diamonds do not actually exist. Instead, the unexplained (002) ‘forbidden’ reflections used to diagnose n-diamonds, are in fact an illusion introduced by spatial resolution limitations of the instrument used to analyse it (Su et al. 2022). They show that moving the instrument while observing a c-diamond causes it to transform into an n-diamond with the forbidden (002) reflection. Earlier studies have suggested similarly; that n-diamonds are really just c-diamonds with defects in their lattice, or variations in thickness (Nemeth et al. 2015). This has significant implications for the entire YD nanodiamond debate, and particularly for the findings of Daulton et al. (2017a); if, as they claim, most YDB sites only contain n-diamonds, and all n-diamonds are just c-diamonds, that means all YDB sites contain c-diamonds, nullifying Daulton et al. (2017a)’s argument that YDB ‘nanodiamonds’ are not diamonds. Given this development, and that Daulton et al. (2010) is the only failed replication of YDB nanodiamonds, probably because their Frankenstein samples did not contain the YDB, their results should be taken with a pinch of salt. Thus, the presence of nanodiamonds at the YDB has been robustly established, and still stands as evidence of the YDIH.

Scoria-Like Objects & Other Impact Products

Perhaps one of the strongest lines of evidence for an impact event at the YD onset are scoria-like objects (SLOs), a term coined by Bunch et al. (2012) for a special type of impact melt glass. SLOs are characterised by shock-fused vesicular siliceous glass which has similar textures to volcanic scoria, hence the name. Often, they comprise several fused, irregularly shaped and sub-rounded glassy objects, and contain higher concentrations of silicon, aluminium and calcium than microspherules, and less iron (Bunch et al. 2012). While these can be similar in appearance to magnetic microspherules, they are not always magnetic, as this depends on their iron content. Their silica-rich composition is a product of terrestrial sediments that were entrained in the impact plume at temperatures up to 2,200°C, the boiling point of quartz (Bunch et al. 2012). Their surface morphologies are also different from magnetic microspherules, typically lacking the characteristic dendritic crystalline textures formed as iron cools rapidly. They are geochemically distinct from to known cosmic, anthropogenic, and volcanic spherules, and overlap closely with known impact materials (Figures 13 & 14).

As a bolide detonates in the atmosphere during an airburst event, temperatures hotter than the surface of the sun (>4,700°C) occur in the convective plume (Bunch et al. 2012). By-products ranging from solid or plasticized impactites to molten or vapourised material occur in different regions of the plume; the dynamic processes occurring during the impact can rapidly eject terrestrial material in all directions and draw some of it further up into the plume, where individual grains can collide repeatedly, producing “accretionary and collisional features” (Bunch et al. 2012). Because airburst shockwaves are dispersed through the atmosphere, shock metamorphism produced during airbursts is minimal; the ejecta they produce is mostly plucked from the top few centimeters of surface material and entrained in the plume (Svetsov & Wasson 2007).

While microspherules have now been found at more than 2 dozen YDB sites, SLOs have only been found at a few; sites that contain SLOs in addition to microspherules are interpreted to represent ground-zero sites for localised airbursts at the YDB. Bunch et al. (2012) report SLOs from 3 sites: Blackville in South Carolina, Melrose in Pennsylvania, and Abu Hureyra in Syria (Figure 15), suggesting there were at least 3 large airbursts over North America at the YDB; other suspected ground-zero sites have been discovered since but have not been analyzed as thoroughly as those reported in Bunch et al. (2012). These include the Bowser Rd. mastodon site in New York (LeCompte et al. 2017), and the Pilauco site in Chile (West et al. 2020), Mt. Viso in Venezuela (Mahaney 2023), and more unpublished sites including Newtonville, New Jersey.

Following the publication of Bunch et al. (2012), one group responded a few years later with claims that the melt products from Abu Hureyra were formed by normal housefires, and do not require a cosmic impact to produce (Thy et al. 2015). However, they do not mention the other two sites in North America that produced identical proxies; even if they did, there is no evidence for any sort of houses at these other sites, and so there is no justification to claim those analogous SLOs were formed in housefires. In Thy et al. (2015)’s study, contrary to everything you have just read in this section, they claim:

“…there has been no geochemical assessment on the composition and formation of the siliceous spherules, as provided by the present study.”

Clearly this is untrue; in their very next sentence they cite studies that assessed the geochemistry and formation of the siliceous spherules from Abu Hureyra. However, they raise quite a few good points in their paper regarding potential reasons these objects can have been created at lower temperatures than claimed by Bunch et al. (2012), which required a lot of meticulous work to refute. One such claim is that:

“…in multi-component systems characterized by high SiO₂ content, this melting temperature will be greatly reduced by the presence of other oxides, particularly of alkali metals such as K₂O and Na₂O, resulting in initial melting occurring often well below 1000°C.”

Five years later, following additional analyses and extensive melting experiments, the CRG published a very detailed response to Thy et al. (2015), entirely devoted to Abu Hureyra (Moore et al. 2020). Melting experiments were conducted on sediment and SLOs from the same levels Abu Hureyra in laboratory furnaces between 1,100°C and 1,850°C at intervals of 50°-150°C (Figure 16). The only other known natural process capable of producing temperatures exceeding 1,850°C other than a cosmic impact/airburst is lightning, but they can also be produced during nuclear detonations. Their experiments showed that SLOs with plant imprints, smoking gun evidence produced when the molten material was splashed onto vegetation, were formed at minimum temperatures of between 1,200° and 1,300°C. Furthermore, in laboratory melting experiments, bulk sediment showed no sign of melting at 1,100°C; smaller quartz grains began to exhibit partial melting at 1,300°C (Figure 16), which is lower than its normal melting temperature. Moore et al. (2020) attribute this to a ‘fluxing’ effect caused by limestone and soda ash in the sediment, agreeing with Thy et al. (2015)’s explanation for lower melting temperatures than claimed in Bunch et al. (2012). However, at 1,700°C, the melt glass produced in the experiments began to closely match those found at the site; clusters of white quartz remained on the surface of the transparent/translucent brown melt glass. This melt glass remained molten until it began to solidify around 1,200° to 1,300°C, and the same was true in experiments that re-melted melt glass retrieved from the site. Additional geochemical analyses of YDB SLOs from Abu Hureyra revealed melted grains of monazite, chromite and chromium-rich magnetite that melted, and perhaps even boiled; their presence indicates temperatures of between 2,000° to 2,600°C.

In total, Moore et al. (2020) has 63 pages of evidence and 46 figures showcasing the various melting experiments; very high magnification electron micrographs and geochemical maps of melt glass and SLOs from Abu Hureyra offer strong refutations to every single one of Thy et al. (2015)’s claims. Moore et al. (2020)’s arguments and results are airtight, and not a single issue has been raised regarding their evidence in the 3 years since it was published. Thus, melt glass and SLOs stand as two of the strongest lines of evidence supporting the YDIH. After most of this article had been written, three additional studies detailing impact evidence from Abu Hureyra were published, including the first evidence of shocked quartz published by YDIH proponents.

Platinum Group Elements

Enhanced concentrations of iridium and other platinum group elements are often regarded as good evidence of an impact event; the quintessential example being the K-Pg boundary impact, where very high levels of iridium were found in the same layer globally (Alvarez et al. 1990). When Firestone et al. (2007) was first published, their suite of proxy evidence included enhanced iridium concentrations at the YD onset at several of their sites. Specifically, elevated iridium was primarily detected in individual magnetic grains or impact spherules ranging from 2 ppb (± 90%) to 117 ppb (± 10%), with bulk sediment from 3 sites found to contain detectable iridium (Firestone et al. 2007). Of all sites where magnetic grains produced iridium anomalies, no similar anomaly was detected above or below the YD onset. They clarify that bulk sediment iridium concentrations are near the detection limit of the neutron activation analysis that was used to find them. It is possible, perhaps even quite likely, that the neutron activation was detecting very high concentrations from a few individual magnetic grains within the bulk sediment, like the grains that reportedly contain ~117 ppb of iridium. In addition to iridium concentrations, they also reported elevated levels of other rare earth elements associated with impact events such as nickel and chromium (Firestone et al. 2007).

Unfortunately for the YDIH, another premature nail in the coffin resulted from the first attempt to replicate Firestone et al. (2007)’s iridium spike, which was reported by the authors as being unsuccessful. Paquay et al. (2009) examined samples from the black mat at multiple sites, including some from Firestone et al. (2007), but reported finding no enhanced iridium signature. They also tested for other platinum group elements (PGEs) and gold in the Black Mat but did not find any (Figure 17). They also made incorrect assumptions about the nature of the impact scenario, which like had direct influence on their conclusions against the YDIH.

Regarding the methods used by Paquay et al. (2009), they differed both in the types of sample tested, and the geochemical method used for the analysis; they exclusively tested bulk sediments rather than individual grains, and used Inductively Coupled Plasma Mass-Spectrometry (ICP-MS) instead of neutron activation. While admittedly ICP-MS is more suited to the work than neutron activation, it uses much smaller sample sizes than neutron activation, and usually also requires bulk digestion using extremely dangerous chemicals. The different methods and sample preparation requirements inherently result in a lower chance of the analysed sample containing individual spherules with high iridium concentrations; while neutron activation sample could have been quite large, up to multiple kilograms, while ICP-MS usually takes samples less than 1 mL. Bunch et al. (2009) published a comment highlighting very high uncertainties in their data (up to ± 140%) and reproducibility issues (varying by up to 400%). However, most importantly, they highlight that their conclusions contradict their data; upon closer inspection, Paquay et al. (2009) did in fact replicate YDB iridium spikes up to >300% above background in the YDB, found alongside microspherules and nanodiamonds at both Murray Springs and Lake Hind (Bunch et al. 2009; Figure 18).

A later study by an independent author Marshall (2012) replicated a similar concentration of iridium, exceeding 300% above background in the YDB at Bodmin Moor in southwest England. Another independent team later detected elevated PGE concentrations including iridium in several magnetic microspherules, and in bulk sediments (Andronikov et al. 2016a). While it is safe to say that a slightly elevated concentration of iridium exists at the YD onset, it is obviously not particularly strong. However, there is a much stronger, much more reproducible geochemical signature that occurs at the YD onset that has breathed new life into the YDIH debate since 2013.

The late Prof. Wally Broecker was a pioneer in the field of paleoceanography and a close colleague of James Kennett of the CRG; together they did much of the early work on the YD in ocean cores. Before he passed in 2019, Broecker was the pre-eminent expert on all things Younger Dryas in the field of paleoceanography. He was initially very skeptical of the idea that it might have been triggered by a cosmic impact and was “shocked by [Firestone’s] grandiose claims” (Broecker 2017). His deep-seated skepticism of the YDIH even led to him make some unfortunate disparaging comments about Kennett in a 2011 exposé on the YDIH (Dalton 2011):

“It’s all wrong,” said Broecker, if not “very likely total nonsense. But he never gives up on an idea”… Kennett seems fixated on the Younger Dryas, Broecker added, ‘He won’t listen to anyone. It’s almost like a religion to him.”

In 2017, he wrote a “Broecker Brief (#3)” detailing insider knowledge and how his personal beliefs and position regarding the YDIH had evolved over time (Broecker 2017); these briefs, of which there were many, appear to have be his way of setting his scientific affairs in order before he passed. In brief #3, he explains that he encouraged Alan Zindler and Stein Jacobsen of Harvard University to search for iridium at the YD onset in the Greenland ice cores, anticipating a death blow to the YDIH when none was found. He describes his pleasant surprise when Jacobsen handed him a manuscript reporting a platinum anomaly of up to 300 times background levels, with “rise and decay times that were consistent with an air blast” (Broecker 2017). While Broecker does not mention in the brief whether he ever apologised to Jim Kennett for making such disparaging comments about him, he became a full-throated believer in the YDIH; in his mind, Kennett was vindicated in his belief that an impact had triggered the YD cooling, and had maybe even converted Broecker to his ‘religion’:

“My take on this is that the Greenland platinum peak makes clear than an extraterrestrial impact occurred close to the onset of the YD. Perhaps the object was vaporized in the atmosphere accounting for the shape of the platinum peak. …Although I don’t for a minute believe that this impact did in the mammoths and the Clovis people, I do think that it triggered the YD. …I can’t buy it’s a coincidence.”

The paper was then published in the Proceedings of the National Academy of Sciences in 2013 by Petaev et al. (2013). The following quotes from the paper provide valuable insight into the potential origins of the platinum:

“The Pt concentrations gradually rise by at least 100-fold over ~14 y and drop back during the subsequent ~7 y. The decay of the Pt signal is consistent with the ~5-y lifetime of dust in the stratosphere. The observed gradual ingrowth of the Pt concentration in ice over ~14 y may suggest multiple injections of Pt-rich dust into the stratosphere that are expected to result in a global Pt anomaly.”

and

“The Pt anomaly is accompanied by extremely high Pt/Ir and Pt/Al ratios, indicative of a highly unusual source of Pt in the ice…” “Materials with high Pt/Ir ratios and essentially no Al are known among magmatic iron meteorites. Finding a terrestrial Pt-rich and Ir, Al-poor source is difficult. Most volcanic rocks have elevated Pt/Ir ratios, although not as high as in iron meteorites, but Pt/Al ratios are very low.”

and

“Circ*mstantial evidence hints at an extraterrestrial source of Pt, such as very high, superchondritic Pt/Al ratios at the Pt anomaly and its timing…”

and

“Until the question about the nature of Pt-rich material and the means of its delivery to the ice is resolved, an extraterrestrial source of Pt appears likely.”

The only reasonable conclusion from the sections reproduced above is that the origin of the Pt spike is indeed unresolved. While this essentially means that Petaev and coauthors are just as clueless as anyone else, they clearly state that absent an alternative explanation, an extraterrestrial source of the Pt is the most likely. It is particularly disappointing therefore, to see that he joined the “comprehensive refutation” paper as a coauthor. While this is clearly nothing to do with his actual thoughts on the source of the Pt, and everything to do with getting Mark Boslough to stop sending him harassing emails every few days demanding that he make a public statement against the YDIH. Petaev’s contribution to the refutation is quite milquetoast, and basically just re-iterates his previous interpretation more conservatively.

Regardless of Petaev downplaying the YDIH like a hostage at gunpoint, the YD Pt anomaly has since been established as global in nature, just as they predicted in their 2013 paper (Figure 19). Its confirmation as global in nature serves to debunk pseudoscientific claims made in response to the original paper by Mark Boslough. Prior to any attempt to replicate it, or any other information whatsoever regarding its nature, Bostradamus predicted that the Greenland Pt would prove to be a local signature and would prove to not be global in distribution. Further, he suggested it must have been caused by the Cape York impact, despite admitting that the known Cape York Pt/Ir and Pt/Al ratios are completely different to the GISP2 anomaly (Boslough 2013). Because the Cape York impact is believed to have been a smaller localised impact, if the Pt turned out to be global, it can’t have been from the Cape York impact, and later work explored the potential relationship between the Cape York and the YDIH and found it unlikely.

Regardless of whether the YD platinum provides evidence of an impact event, which is yet to be conclusively established, YDIH proponents have taken Petaev et al. (2013)’s suggestion that the Pt anomaly should be found globally and ran with it; such an anomaly with a ~7-14-year signature could be used as a very high resolution geochemical datum that is significantly more precise than radiocarbon and other traditional dating methods. Thus, the Pt spike is precisely what is needed to make any reliable determination of the nature of synchroneity at the YDB; this datum, if shown to be global, should be a minimum requirement of any study making claims about the YD onset. The first comprehensive study seeking to test its utility as a geochemical datum anomaly found it in YDB sediments at 11 archaeological sites across North America in varying concentrations between 0.3 to 65.6 ppb (Moore et al. 2017; Figure 20). While the average crustal abundance of platinum is 0.5 ppb, the 0.3 ppb spike from Johns Bay is higher than background at that site.

Establishing a global geochemical datum that precisely marks the YD onset fully resolves all issues of chronological uncertainty and misalignment of data, such as in Gill et al. (2009). However, in order to establish the utility of such a datum, it would need to be replicated much further afield than North America. Luckily, in 2019 it was replicated in several studies at very high latitudes in South America and South Africa, and again in North America. Pino et al. (2019) replicated a 10-ppb spike accompanied by a 100:1 Pt/Pd ratio in YDB sediments from Pilauco, southern Chile (Figure 21). The platinum spike at Pilauco is precisely coeval with a peak in impact microspherules, an abrupt collapse of pollen and seeds, and a major spike in charcoal representing biomass burning (Figure 21).

Another independent group was encouraged by James Kennett to search for the Pt in their pre-existing sediment cores from Wonderkrater in South Africa at the YD onset; this group also found a 10-ppb spike coeval with YD cooling in their pollen-based temperature reconstruction (Thackeray et al. 2019; Figure 22). These two studies provide confirmation that the platinum spike really can be used as a global datum; if it has been replicated at such high latitudes in both the Northern and Southern hemispheres, it should be replicable everywhere, provided the appropriate analytical techniques are used.

Building on his 2017 replication of the platinum in archaeological sediments, Chris Moore of the CRG published a multi-proxy study from a lake core from White Pond in South Carolina (Moore et al. 2019). They reproduced the Pt signature and demonstrated that it is coeval with both a decline in fungal spores that feed on megafauna dung, and a major spike in carbon representing the biomass burning event (Figure 23). Notably, the White Pond platinum anomaly closely mirrors the one from Petaev et al. (2013), with a small spike just prior to the YD onset followed by a major spike just after it. This study further confirms that the platinum spike definitively marks the YD onset in ice cores, lake cores, and archaeological sediments on at least three continents in both hemispheres.

By 2021, platinum had been very well established as the geochemical signature of the YDB, regardless of whether it originated from an impact event. However, this fact had apparently eluded a particular Japanese team, who decided to search for ‘YDB iridium’ as part of their palaeoclimatological reconstruction at Lake Suigetsu (Nakagawa et al. 2021). They incorrectly and somewhat sarcastically describe a high concentration of iridium at the YDB in lake sediments in many places in North America as “The principal line of ‘evidence’ to support…” the YDIH. If the reader has made it this far, they will know this is clearly not the case. Iridium, at least compared to platinum, has not been extensively replicated at the YDB, and most of the places it has been replicated were not in lake sediments. Nakagawa et al. (2021) made no attempt to search for platinum at the YDB, but even if they had, they would not have found any; the analytical method they were using for their core geochemistry was totally incapable of detecting any rare-earth elements, let alone iridium or platinum.

Prior to the experimental evaluation of magnetic separation protocols I conducted, I was originally planning to attempt to replicate the global platinum anomaly in lake sediments from Australia. During the planning process, a variety of analytical methods were explored, including an advanced ‘Itrax’ core scanner, which uses X-ray fluorescence (XRF) to efficiently measure elemental concentrations at high resolution throughout the entire core. However, the author was informed by the technicians at the Australian Institute of Nuclear Science & Engineering that XRF core scanners are not capable of detecting low ppb concentrations of platinum. Even if a significant spike was present in the core, it is still such a low concentration that it’s below the detection limit of the instrument. Because of this, XRF is generally considered semi-quantitative and inappropriate for measuring rare earth elements (Rodriguez-Germade et al. 2015; Gregory et al. 2019); clearly the authors of Nakagawa et al. (2021) did not inquire about the capabilities of the instrument they were using, because if they had, they would have known it was unsuitable for this type of work. To be reliable, REE concentrations obtained using XRF need to be calibrated using much more precise methods: Neutron activation, which requires large samples and permanently irradiates it, and ICP-MS, the gold standard, which has a detection limit in the parts per trillion and works well with incredibly small samples but destroys them. Regardless of their intent, this study amounts to a drive-by hatchet job; both their study design and their understanding of the hypothesis in general were significantly flawed. This has been a recurring theme throughout the 15-year history of the hypothesis, and clearly it shows no signs of improving. A common criticism of the YDIH is that its proponents are primarily a small group of collaborators, with only a handful of independent researchers being convinced of its veracity. Papers like Nakagawa et al. (2021) are the reason why; proponents understand the claims being made by the hypothesis, and know what is required to properly test it, while opponents consistently demonstrate the opposite. The Nakagawa paper has at least one YDIH critic, a member of the van Hoesel team, as a co-author; they are likely the one who suggested investigating the YDIH as a side quest within the greater project, as it was clearly tacked on as an afterthought.

Part III

Human Population & Cultural Change

While the debate amongst the geoscientists over purported YDB impact proxies was beginning to take off, another discipline was already in uproar over Firestone et al. (2007)’s audacious claims. It was the archaeologists, and they were pissed. Firestone et al. (2007) cited the rapid disappearance of the Clovis technocomplex, along with various other archaeological evidence to demonstrate the potential human effects of the YDIH. The pushback was immediate and sometimes deranged; multiple critics tried to keep a straight face as they claimed that human populations barely even noticed the YD was occurring (Buchanan et al. 2008; Holliday & Meltzer 2010; Meltzer & Holliday 2010). This claim might even take first prize as the most ridiculous ever made in the history of YDIH debate. If ancient Indigenous populations barely even noticed the significant temperature swings and radical changes to the environment as claimed, this bodes well for the future of our species; the worst projections for anthropogenic climate change are comparable to the magnitude of climate changes associated with the YD onset and termination, but far less abrupt. Thus, according to these morons, the worst possible projected consequences of climate change over the next century will be barely noticeable by modern civilisation. Of course, the idea that humans, particularly Indigenous people, barely noticed the widespread and significant environmental and ecological consequences is nonsensical and remarkably ignorant, especially for archaeologists; Indigenous people are the most in tune with nature among our species; they have their fingers on the pulse of nature, and are aware when even the slightest changes occur, because their survival depends on it. Furthermore, the archaeological record very clearly shows significant disruptions all over the world at the YD onset, from the termination of the Clovis technocomplex (Figure 24) in North America (Newby et al. 2005; Anderson et al. 2011), the Magdalenian, Federmesser & Hamburg cultures in Europe (Enloe 2001; Weber et al. 2011; Pettit & White 2012; Burdukiewicz 1986), and significant influences on cultures in the Levant (Moore & Hillman 1992). Clovis fluted projectile points vanish from the archaeological record at the YD onset, with different styles of points appearing shortly after. The quarries where people procured their raw materials for Clovis points were mostly abandoned, suggesting that if the people weren’t killed, they were at least forced to move elsewhere (Figure 24). More recently, much work has been done by Martin Sweatman to link the YDIE with the Neolithic revolution (Sweatman 2017, 2019) and carvings on pillar 43 at Göbekli Tepe, which may represent a zodiacal datum marking the YDIE (Sweatman & Tsikritsis 2017).

In addition to affecting occupation patterns and material culture dynamics, a catastrophe like the YDIE would be expected to produce a significant genetic signal in humans, but such a signal had never been detected. However, groundbreaking work over the last decade has revealed significant global and regional Y-chromosome and population bottlenecks occurring near the YD onset (Karmin et al. 2015; de Pablo et al. 2019; Sepulveda et al. 2022). The discovery of this Y-chromosome genetic bottleneck represents a multi-dimensional ‘smoking gun’ for the YDIE. In 2015, 101 authors from 67 institutions gathered or contributed their existing data to a paradigm-disrupting study of 456 geographically diverse high-coverage Y-chromosome sequences, 299 of which were new (Karmin et al. 2015). According to their data, beginning around 10,000 years ago, the ratio of males to females dropped to 1:17, meaning there were 17 females for every 1 male. They offer several potential explanations for this, most of which are unconvincing or defy logic, such as male-driven conquest, population collapse following the advent of agriculture in Europe, and culturally driven sex-selection in offspring. Unfortunately for the authors, these explanations only work on a regional scale and are not applicable in a global context; the most parsimonious explanation is that the simultaneous GLOBAL Y-chromosome bottleneck (Figure 25) resulted from a synchronous GLOBAL catastrophe.

There is, however, one major issue to overcome with this study if it is to support the YDIH; their signal seems to intensify around 10,000 years ago, ~2,800 years after the YD onset, following a much slower decline that appears to have begun around the YDB. At first glance, this would appear to be a problem for linking the signal to the YDIE, but there is some wiggle room here. Their chronology, which is based on a ‘molecular clock’ estimate, may be slightly off, meaning the bottleneck signal can have occurred at the YD onset. I do not purport to be a geneticist by any means, and I have little understanding of the mechanics underpinning molecular clock chronologies. However, like the YDIH, the molecular clock chronology is merely a working hypothesis based on the fundamental assumption that proteins in DNA undergo mutations at a consistent rate throughout time and between very different species (Lee & Ho 2016). Within the molecular clock model, DNA mutation rates do not, have not, and cannot change over time. If they can, the assumption would be that we have it 100% worked out; both are very bold claims with good odds of being wrong. While this rickety paradigm underpins all modern genomic models and research, many scientists are unconvinced, and there are competing hypotheses (Huang 2008; Hu et al. 2013). As the YD impact event is the only known global catastrophic event capable of producing this genetic signal in reasonable proximity, it is not unreasonable to suggest a potential relationship. Today, geochemical impact proxy evidence that marks the YDB can be used for chronological calibration at individual sites and stratigraphic correlation over long distances. In much the same way, Karmin et al. (2015)’s Y-chromosome bottleneck should be used to calibrate the molecular clock for humans.

So, what do other studies say about a population or genetic bottleneck at the YD onset? De Pablo et al. (2019) demonstrates a sustained period of population decline and stagnation from ~12,900 BP following a period of exponential population growth between ~16,600 and ~12,900 BP. Later, as the temperatures and environmental conditions became more hospitable in the early Holocene (~10,200 BP), populations began to increase again. These findings contradict Karmin et al. (2015), whose Y-chromosome bottleneck supposedly peaked at this time; their molecular clock chronology is incompatible with the archaeological record, at least in this region. Unlike the genomic studies, de Pablo et al. (2019) use summed probability distributions (SPDs), a relatively new statistical method that has been touted as a useful tool for reconstructing past population dynamics; SPDs are based on the amount of organic carbon deposited, and more importantly, sampled and dated from a site, and their utility for reconstructing past population dynamics has been challenged (Carleton & Groucutt 2020). Because this method is often used in archaeology, the findings of de Pablo et al. (2019) are used as evidence here. However, issues with SPDs will be discussed in more depth later in this review.

South America

A groundbreaking paper by Sepulveda et al. (2022) presents genomic evidence for a previously unknown South American population that existed prior to 18,000 years ago, completely re-writing the genetic history of Pleistocene South America. They conclude that a significant ‘loss of lineages’ occurred at the YD onset, and particular sub-lineages that survived the cataclysm had expanded and diversified by its termination. Unlike the Karmin et al. (2015) paper, their chronology matches nicely with the YD onset. For reasons unknown, their findings have been largely ignored by prominent pontificating paleoanthropologists with expertise in genetics; this significant discovery should be making waves throughout the media, but because it doesn’t fit within the prevailing paradigm, and offers support for the YDIH, to date it has been largely ignored.

Traditionally, the earliest widely accepted occupation of South America based on archaeological evidence was around 14,500 BP at Monte Verde in Chile; even this date has been controversial, as it pre-dated the earliest accepted (at the time) occupation dates from North America by >1000 years (Meltzer et al. 1997; Dillehay 1997). For almost a century, the dominant paradigm for human occupation of the Americas has dictated that humans first entered North America through the ice-free corridor between the Laurentide and Cordilleran ice sheets (Arnold 2002; Braje et al. 2017) approximately 13,500 years ago. If humans first came to the Americas through this corridor ~13,500 years ago, how could they have been on the southern tip of South America ~1,000 years before that? It is a good question, and Monte Verde was one of the first cracks that appeared in this paradigm, which has since been largely abandoned in favour of the “coastal migration” hypothesis (Erlandson et al. 2007; Davis & Madsen 2020; Praetorius et al. 2023).

Genetic evidence of human populations in South America prior to 18,000 BP is all well and good, but “where is the archaeological evidence for this supposed population?”, the reader may ask. In fact, there has been a significant body of evidence published in peer-reviewed journals by archaeologists since the mid-2010s that fits very nicely with this new genomic evidence. The following late Pleistocene archaeological sites in South and Central America may have been occupied by the population Sepulveda et al. (2022) claim to have detected prior to 18,000 BP:

• The Boqueirão da Pedra Furada in Piauí, Brazil, with occupation beginning at least 20,000 BP, and perhaps as early as 45,000 BP (Dillehay 1997; Santos et al. 2003; Boeda et al. 2014; Lahaye et al. 2015).

• The Toca de Tira Peia site in Piauí, Brazil, with evidence of occupation from at least 27,000 BP (Lahaye et al. 2013; Boeda et al. 2016)

• The Toca do Sitio do Meio site in Piauí, Brazil, with periodic occupation from at least 35,000 BP (Boeda et al. 2016; Mota & Scheel-Ybert 2019)

• The Arroyo del Vizcaino site in Uruguay, with occupation from at least 30,000 BP (Fariña 2017)

• The Santa Elina site in Brazil, with occupation from at least 20,000 BP (Vialou et al. 2017)

• The Chiquihuite Cave site in Mexico, with occupation from at least 26,000 BP, perhaps as early as 33,000 BP (Ardelean 2013; Ardelean et al. 2020, 2022; Boeda et al. 2021;)

• The Monte Verde site in Chile, with reasonably well-accepted occupations from ~18,500 BP (Dillehay et al. 2015), with early work hinting at occupation up to 33,000 BP (Meltzer et al. 1997).

While much of the evidence from these sites has been poorly received by critics (Chatters et al. 2021; Coutouly 2022; Holcomb et al. 2022), Sepulveda et al. (2022)’s genomic evidence strongly supports the idea that these sites were formed by humans. This new data cannot be ignored and represents a significant threat to the prevailing paradigm of human occupation in South America. In the 1990s, archaeological evidence suggested human occupation at Monte Verde as early as 30,000 BP (Dillehay 1997; Meltzer et al. 1997). Monte Verde was among the first sites to produce evidence for human occupation in Pleistocene South America and was met with a cacophony of harrumphs and sneers of derision; it was instantly dismissed by archaeologists who refused to even visit the site (Dillehay 1997; Meltzer et al. 1997). At the time, American archaeology was firmly in the grip of the Clovis-first dogma, and so these dates were quibbled, denied, and gradually whittled down until a more palatable date of ~14,000 BP was begrudgingly accepted; unfortunately, its ramifications are only just now beginning to be digested. As the grip of Clovis-first on the discipline has gradually weakened and scientific methods have improved, proponents have retaken some ground, and today occupations in North America around ~15,500 BP are more well-received (Davis et al. 2022). Most recently, Dillehay et al. (2021) have strongly defended against many recent (non-peer-reviewed) attacks on their work at these sites. Commenting on another recently discovered Pleistocene occupation, Boeda et al. (2021) note that the latest work at Chiquihuite cave in Mexico by Ardelean et al. (2020, 2022) strongly supports findings from their own sites.

The most recent assault on these Pleistocene South American sites attempts to cast doubts on the idea that they were even occupied by humans; instead, they attribute the lithic assemblages at the Pleistocene-aged sites to capuchin monkeys (Agnolin & Agnolin 2022). To do so, these authors needed to intentionally disregard the additional archaeological context of these sites, a ‘number of features’ which they gloss over entirely as follows:

“Pedra Furada is probably the most well-known and debated site,having a number of features that have been attributed to anthropogenic origin and genesis for Pleistocene levels (32,000–50,000¹⁴C yr BP)” (Agnolin & Agnolin 2022).

Naturally, despite the relative weakness of the argument, this study has received a warm reception by many opponents of the work being done at these sites by Boeda and colleagues. However, now there is strong genetic evidence of human presence in South America coeval with claimed occupations at these sites; unless the claim is that these people were archaeologically invisible, the notion of human occupation at these sites must be seriously considered and treated with more respect. That is, of course, unless they want to claim that Sepulveda et al. (2022)’s genomic evidence was obtained from the ancestors of present-day human-capuchin hybrids. Not long after this study claimed that the lithics were produced by Capuchins, highly compelling evidence from the Santa Elina rock shelter in Central Brazil was published in the Proceedings of the Royal Society B. Pendants made of bones from Glossotherium phoenesis, an extinct megafaunal species of giant sloth, with intentionally drilled holes have been securely dated to 25,000 years old (Pansani et al. 2023). Thus, unless Capuchins were manufacturing jewelry 25,000 years ago in Brazil, this is definitive proof that humans were the ones occupying these sites.

North America

The late Pleistocene history of North America is viewed similarly; evidence for dozens of pre-Clovis occupations have been reported over many decades (Leakey et al. 1968; Cinq-Mars 1979; Dillehay 1997), only becoming somewhat palatable in the last 5-10 years. The best-documented example of this is Jacques Cinq-Mars’ claims of human occupation at Bluefish Caves, Yukon, far-west Canada, by 24,000 BP (Cinq-Mars 1979). His claims saw him effectively ‘cancelled’ from archaeology; he was laughed out of conferences and ultimately bullied out of academia (Pringle 2017), but the most recent analysis of his claims has largely vindicated him (Bourgeon 2021). His treatment by his peers was among the more brutal examples; many other archaeologists have dared to step outside the prevailing paradigm to propose early occupations and suffered significantly less, but this behaviour is still unacceptable, and is detrimental to the advancement of science. Another famous example is Louis Leakey, whose family name is among the most widely revered in the discipline of archaeology; even the Leakey name did not shield him from being largely deplatformed after advocating for human occupation in the Calico Mountains of California between ~40,000 and ~120,000 BP (Leakey et al. 1968; Dillehay 1997). In summary, the halls of archaeological history are littered with the deceased careers of those who had the temerity to present evidence that challenged the prevailing paradigm.

Importantly, there is a distinct and mysterious absence of human remains dating to Clovis or pre-Clovis times in the Americas. Remains of only two YD-age humans have ever been found, which itself demands explanation (Chatters et al. 2014; Becerra-Valdivia et al. 2018); the Clovis-age skeletons of a male infant named Anzick-1, and a teenaged female named Naia. Where are all the skeletal remains of the North American people from around the YD onset? If their populations were large enough to wipe out millions of mega mammals within a few decades, would their remains not be plentiful? How can anyone claim anything with certainty regarding the people who lived in the Americas at the YD onset based exclusively on fragmentary evidence of their material culture? Could this apparent lack of human remains be indicative that some catastrophe befell these people? Claims that the YD onset, impact event or not, had no effects on global human populations simply are not tenable and are indicative of ideological possession. In short, the idea that significant effects on human population around the YD onset has been ‘debunked’ is unsupportable based on the literature. While Clovis-first had already been largely discredited when Firestone et al. 2007 was published, many subsequent findings in the Americas between 15,000 and 130,000 BP have all but confirmed the argument for pre-Clovis occupation in the Americas throughout the late Pleistocene.

The most recent work has revealed human occupation at Rimrock Draw rockshelter in Oregon as early as 18,250 years ago, based on very secure radiocarbon dating of camel-tooth artefacts found below a volcanic eruption dated to 15,000 years ago. The paradigmatic vacuum left by Clovis-first is gradually being filled, dragging even the most zealous dogmatists kicking and screaming with it. A little epistemic humility goes a long way, and the discipline of archaeology is sorely lacking in this department, as evidenced by the visceral response to the work of Graham Hanco*ck.

Megafaunal Extinctions

It has long been recognised, since well before Darwin’s Voyage of the Beagle in the 1830s, that some sort of catastrophic event had been responsible for the extinction of the mega mammals of the Americas (Pavid 2018). In the Voyage of the Beagle, Darwin (1845) made the following observations:

“The mind is at first irresistibly hurried into the belief of some great catastrophe; but thus to destroy animals, both large and small, in Southern Patagonia, in Brazil, on the Cordilliera of Peru, in North America up to Behring’s [sic] Straits, we must shake the entire framework of the globe”

and

“These remarks, I may be permitted to add, directly bear on the case of the Siberian animals preserved in ice. The firm conviction of the necessity of a vegetation possessing a character of tropical luxuriance, to support such large animals, and the impossibility of reconciling this with the proximity of perpetual congelation, was one chief cause of the several theories of sudden revolutions of climate, and overwhelming catastrophes, which were invented to account for their entombment.”.

See Also

Notice catalographique « ICOM Committee for Conservation, ICOM-CC : 11th Triennial Meeting, Edinburgh, Scotland 1996 : preprints »WSJ What’s News - WinampSheet music: Exploration (Saxophone Quartet: 4 saxophones)Morning World from Monroe, Louisiana

The Younger Dryas impact hypothesis affirms Darwin’s interpretation; the YD impact event really did shake the entire framework of the globe as between dozens and hundreds of comet fragments rained destruction all over the Earth. Darwin’s catastrophist interpretation of the megafaunal extinctions has been echoed by many others throughout history, such as archaeologist Frank Hibben (1968), who did extensive work on the megafaunal remains found in the muck deposits of North America:

“The Pleistocene period ended in death. This was no ordinary extinction of a vague geological period which fizzled to an uncertain end. This death was catastrophic and all-inclusive… The large animals that had given the name to the period became extinct. Their death marked the end of the era. But how did they die? What caused the extinction of forty million animals?”

The latest work suggests multiple episodes of catastrophic emplacement may have occurred over the last 40-50,000 years; Hagstrum et al. (2017) recovered high concentrations of impact-related microspherules from inside the skulls of megafauna found in the muck (Figure 26), and tusks embedded with meteoritic material have also been found (Dalton 2007; Hagstrum et al. 2017). Unfortunately, the topic of the enigmatic Alaskan muck deposits, some containing the remains of thousands of animals per acre. is far too broad and deep for inclusion in this review, and it is largely beyond the scope of this review.

The scientific consensus since at least the 1960s has been that some sort of catastrophic event was responsible for the megafaunal extinctions, but the specifics of the catastrophe have always been a point of conflict; around this time, several competing theories vied for paradigmatic dominance. Much to the detriment of scientific knowledge, Paul Martin’s “Overkill” hypothesis (Martin et al. 1967) won out in the end, in part due to how well it meshed with the Clovis-first hypothesis; essentially, one hypothesis was built on the foundations of another hypothesis, a sort of paradigmatic symbiosis. The basis for Overkill’s credibility was a handful of kill sites for proboscideans and bison, and that many species seemed to disappear rapidly and very close to when humans had supposedly first arrived. Despite today’s Indigenous hunter-gatherer populations living in harmony with nature, taking only what is needed from the land that nurtures and sustains them, “Overkill” proponents envision a band of rapacious savages sweeping across the land, butchering everything that moves in an all-consuming blitzkrieg (Martin 1973). According to the “Overkill” fantasy, they hunted down every herbivorous species to extinction, including the horse, sloth, tapir, camelid, and even the last glyptodont, an armadillo the size of a Volkswagen Beetle, without knowledge of, or care for, the consequences. The loss of herbivorous species then supposedly led to the extinction of the carnivorous species, who relied on the herbivores for food (Grayson & Meltzer 2003).

The problem is, even to this day, only a handful of unambiguous kill sites have been found for many of the species that disappeared (Grayson & Meltzer 2003; Meltzer 2015; Meltzer 2020; Bampi et al. 2022), and while there are numerous kill sites for proboscideans and bison, many species have none. Because of the distinct lack of evidence to support it, if Paul Martin’s “Overkill” hypothesis were proposed in today’s scientific environment, it would be dismissed as pseudoscience; ‘Twitter scientists’, who rush to debunk the claims of anyone declared to be in the “out group”, aka claims that contradicts the prevailing paradigm, would be outraged. Thus, the “Overkill” hypothesis should be the next ivory tower to fall in the gradual reclamation of the prehistory of the Americas from decades of dogmatic gatekeeping. As early as 2003, there were already calls to abandon it based on its lack of scientific merit; a requiem for “Overkill” was declared (Grayson & Meltzer 2003), much to the chagrin of ‘paradigmatic holdouts’ (Fiedel & Haynes 2004).

A key factor for acceptance of the “Overkill” hypothesis is that the American megafauna were naïve to the dangers posed by humans, which only makes sense if you subscribe to the Clovis-first hypothesis. Yet, these same proponents accept that this megafauna freely travelled between the Americas and Eurasia, which was inhabited by humans, in times when the route between them was open (Pitulko et al. 2017; Praetorius et al. 2023). Moreover, as mentioned previously, human occupation in South America prior to 18,000 years ago suggests that megafauna there co-existed with humans for up to ~40,000 years before they went extinct. Even if only well-accepted human occupation in North America is considered (Davis et al. 2022), this still allows ~3000 years of co-habitation with humans. Other recent finds such as the Hartley mammoth demonstrates that they were being hunted as far back as ~37,000 years ago (Rowe et al. 2022), though acceptance of these dates is far from widespread. Then of course, there is the Cerutti Mastodon from 130,000 years ago in California, which despite widespread criticism, still has many important questions unanswered; if the bones were broken by modern highway constructions as critics claim (Ferrell 2019), why are the broken surfaces of the bones covered with thick layers of soil carbonate, suggesting they were broken in ancient times (Gruhn 2018). Furthermore, why does use-wear analysis reveal that anvils and hammerstones found at the site were used to break the bones (Bordes et al. 2020)?

A significant criticism of the relationship between the Younger Dryas cataclysm and the disappearance of the megafauna is that some species of megafauna began to go extinct in the ~10,000 years or so preceding its onset. However, in most cases, this assumes that the latest fossil found for each species represents the last living individual, which is wishful thinking to say the least. Furthermore, there are often significant uncertainties in radiocarbon dates of late Pleistocene megafaunal remains, which confounds efforts to determine extinction chronologies (Cooper et al. 2015; Emery-Wetherell et al. 2017). Even without these issues, proponents of the YDIH have never claimed that all megafaunal species went extinct in a split second as the whole world was destroyed. Even so, there does appear to be somewhat of a cluster of species that went extinct at the YD boundary (Haynes 2008; Barnosky et al. 2016; Stewart et al. 2021) Naturally, some sites also appear to contain evidence that certain species managed to live through the YDB event before disappearing around the onset of the Holocene (Steadman et al. 2007; Barnosky et al 2016). Again, this does not serve as evidence against a catastrophe at the YD onset, it just means that some species managed to survive it, and even these claims are debated (Politis et al. 2019). Thorough stratigraphic correlation of the Black Mat and megafaunal extinctions has demonstrated that the Black Mat marks the termination of the Rancholabrean megafauna (Haynes 2008), and that a catastrophic event was almost certainly responsible, and many groups agree (Faith & Surovell 2009). Most studies on the timing of extinctions rely on summed probability distributions (SPDs), an often-misused statistical method which will be discussed in more depth later. However, a new method specifically designed to overcome the issues with misuse of SPDs found that quite a few species succumbed to the drastic climatic changes at the YD onset rather than human interference (Stewart et al. 2021; Figure 27).

The most reasonable view of the megafaunal decline prior to the YD onset is that they were caused by a combination of human interaction, climate, and environmental change, and many researchers are coming around to this idea (Villavicencio et al. 2015). However, it seems likely that the abrupt climate change and environmental impacts induced by the YD impact either finished the job or left most species hanging by a thread. Due to the relative imprecision of pertinent records, it is simply not possible to claim anything for sure; while the evidence for an impact event at the YD onset is now very strong, claims regarding its role in extinctions and other ecological changes are unresolved. It seems reasonable to speculate, however, that the catastrophic nature of the event is at least partially responsible both for the extinctions and may even have contributed to the issues with data precision around the YDB.

Biomass Burning & Vegetation Dynamics

It is now well established by innumerable sources that near-simultaneous vegetation changes occurred throughout much of the world at the YD onset, particularly in the Northern hemisphere. Such changes occurred in a geological instant in Northern Europe (Engels et al. 2022), North America (Shuman et al. 2002), and China (Chen et al. 2020). While controversial, evidence also suggests these changes affected much of the Southern hemisphere, particularly South Africa (Truc et al. 2013). The cause, character, and chronology of these changes has long been debated, but the latest studies have offered unprecedented insight into nuances of the YD cooling. Some consider “climate change” on a generic level to be responsible; changes in atmospheric and oceanic circulations drove hydrological and temperature changes that created favourable conditions for new vegetation types, all unfolding gradually and causing dynamic changes (Mayle & Cwynar 1995; Brauer et al. 1999). Many groups examine the issue more closely, going further to consider the secondary causes or consequences of the climatic shifts, such as shifting fire regimes and faunal distributions driving the changes in vegetation dynamics. Some groups attribute changes in all three of those areas to anthropogenic intervention; humans, they say, swept across the land burning much of the vegetation, destroying faunal habitats, causing extinctions, and modifying the fire regimes (Bird et al. 2008; Pinter et al. 2011b). Yet more groups attribute changes in vegetation dynamics and fire regimes to the trophic reorganisation caused by the megafaunal extinctions; when large herbivores went extinct, the vegetation types they were previously consuming were left to grow out of control (Marlon et al. 2009; Gill et al. 2009). This, combined with increased aridity associated with cool and dry conditions, led to more frequent wildfires. Some believe it occurred the other way around; rapid changes in vegetation dynamics destroyed the food source of the mega herbivores, causing them to die off or move away (Rozas-Davila et al. 2016). One group even appears to claim that changes in vegetation dynamics caused the YD cooling; changing surface albedo led to the changes in precipitation and evapotranspiration known to have occurred at the YD onset, which is an interesting interpretation to say the least (Renssen & Lautenschlager 2000). As the reader can probably tell, there is significant confusion in the literature as to the order in which these changes occurred (Figure 28). However, if scientists were to consider the possibility of Earth having been struck by a disintegrated comet and taking some reasonably sized hits, that would offer a neat and parsimonious explanation; both for the changes themselves and the apparent inability to resolve them.

The latest high-resolution vegetation records indicate a simultaneous change in vegetation throughout northwestern Europe over a period of just a few years (Engels et al. 2022; Figure 29). This study specifically aims to address the uncertainties inherent in even the highest-resolution radiocarbon chronologies, which make it impossible to assess regional lead and lag timing in ecosystem responses, as these can occur in less than a decade (Engels et al. 2022). They further claim that these uncertainties in radiocarbon chronologies are the cause of the conflicting theories regarding the timing and speed of ecosystem disruptions at the YD onset.

To resolve these issues, they combined high-resolution sedimentological, geochemical and palaeoecological proxies from Lake Hämelsee in Germany and compared their results to revised or re-modelled similar records from Meerfelder Maar, Hässeldala Port and Kråkenes (Engels et al. 2022). They identified 5 volcanic events via tephra in the Hämelsee record and combined them with varve-counting and radiocarbon dating to produce a ~10-year resolution through the late Ållerød into the early YD. Importantly, this record is anchored using tephra from the Laacher See eruption, dated to 13,006 ± 9 BP, which also provides valuable insight into the relationship between the eruption, the YD onset, and the proposed cosmic impact (Engels et al. (2022). The LSE tephra was dated with such high precision by Reinig et al. (2021), who reconstructed the dendrochronological and ¹⁴C records of three birch trees entombed within it and compared it with known records spanning the Ållerød-Younger Dryas boundary. When compared with known records, the YD onset is a clear and distinct event that occurs approximately 200 years after the LSE tephra; this has been repeatedly corroborated in European lake varves, where 200 varve layers, each representing 1 year, are consistently counted (Reinig et al. 2021) between the two events. The following quotes summarise Engels et al. (2022)’s findings:

“The results indicate that the environmental impact of climate cooling was more severe than previously thought and illustrates the sensitivity of natural terrestrial ecosystems to external forcing.”

and

“The observed dates for the onset of the YD instead suggest that the impact of large-scale cooling was regionally synchronous across northwest Europe, arguing for abrupt and rapid vegetation response to climate forcing…”

The low temporal resolution of earlier studies limits their utility for precise interpretation of synchroneity of the vegetation changes at the YD onset, but plenty of studies clearly demonstrate their magnitude. A comprehensive analysis conducted by Shuman et al. (2002) of eastern North American pollen records produced between 1969 and 1996 demonstrates the magnitude of changes in vegetation dynamics throughout the continent quite clearly (Shuman et al. 2002; Figure 30).

They find that, contrary to gradual changes in the prior millennia (in most records), the YD is characterised by abrupt and drastic changes in almost all vegetation types, on both local and continental scales (Figure 30). The YD even saw species of vegetation that had not previously existed in North America colonising their new habitats; in some regions along the North Atlantic coast, the cooling initiated a return to vegetation dynamics consistent with prior epochs, but in most other places, the changes were distinct among earlier and later periods. This shows that YD vegetation changes were abnormal and unprecedented among prior D-O oscillations. They find that:

“The YDC [Younger Dryas Chronozone] vegetation patterns demonstrate (1) rapid ecological responsiveness to abrupt climate change and (2) spatially varied patterns of YDC climate change.”

and

“The vegetation responses also show evidence of pervasive vegetation sensitivity to rapid climate change… …Vegetation responses included rapid long-distance range shifts (>300km/century), as well as local changes in abundance.”

Drastic vegetation changes at the YD onset are also reported throughout China. Chen et al. (2020) reconstructed the pollen record from Xingyun Lake in Yunnan Province and compared their results to 35 other sites from previous studies (Figure 31). Their data showed that in high latitudes, trees were largely replaced by herbaceous vegetation at the YD onset, reversing at its termination. They determined that both the abrupt cooling at the YD onset and warming at its termination caused significant changes in vegetation dynamics throughout China, but were unable to draw conclusions regarding the magnitude of the changes (Chen et al. 2020).

“A more pronounced response to severe cooling [is revealed] in high-latitude regions in the north, and muted responses to lesser cooling in low-latitude regions in the south. The vegetation in mid-high latitude regions responded sensitively to changes in both temperature and precipitation, whereas vegetation in low-latitude regions was affected mainly by temperature. There was a lagged vegetation response to climate changes.”

While the idea that YD cooling affected the Southern hemisphere simultaneously with the Northern hemisphere used to enjoy widespread support, it has fallen out of favour in recent years. It is now understood that cooling in most Southern hemisphere records better corresponds with the Antarctic Cold Reversal, which is thought to have occurred prior to the YD in most regions. However, this is not the case at Wonderkrater in South Africa, where pollen records show a significant cooling event occurred at the YD onset, inferred by significant changes in vegetation dynamics; a rapid decline in pollen from plants that flourish in more tropical conditions occurs simultaneously with a surge in pollen from cold-adapted plants corresponds at the YD onset (Truc et al. 2013; Figure 32).

A paper by Gill et al. (2009) in the journal Science examined the relationship between vegetation dynamics, fire regimes, and megafaunal extinctions around the terminal Pleistocene, and specifically claims to have debunked the YDIH. They concluded that large herbivores gradually became extinct after declining for several thousand years prior to the YD, and then the masses of uneaten vegetation produced significant fuel loads and enhanced the natural fire regime (Gill et al. 2009). However, Dr. Martin Sweatman has found several issues and potential errors with their data (Sweatman 2020b; 2021a). First and foremost, the age-depth model on which they base their claims has no error bars to show the uncertainty in the data; not even undergraduate students get away with not including error bars (Sweatman 2020b). This may be because their Supplementary Information reveals uncertainties of up to ~2000 years in their radiocarbon dates; because they are addressing an abrupt, short-lived event; if they had included their uncertainties, their conclusions would have been unsupportable. Additionally, their Appleman Lake time series graph (Figure 33) displays several inconsistencies with most other literature of its kind from the same period. In no small part due to the massive uncertainties in their dating, it is fair to suggest that their data may be misaligned by more than 1000 years. As Dr. Sweatman points out, their graph shows a large spike in charcoal that closely correlates with the final decline of megafaunal dung spores (Figure 33), closely followed by a prolonged period of significant changes in vegetation dynamics (Sweatman 2020b; Figure 33).

While the highlighted section is slightly longer than the ~1200 year duration of the YD according to the age axis, this is not unexpected, as sediment deposition is rarely, if ever, constant over long periods; the age axis is built on a handful of radiocarbon dates with large uncertainties taken from the sediment core, and sections of it may be stretched or compressed out of conformity with the true age. For example, there may be 30 cm of sediment between the first and second radiocarbon dates, which might be 1,000 years apart, and only 15 cm of sediment between the next date, which might also date to 1,000 years later than the previous one. This can be further complicated by depositional hiatuses where sediment deposition is paused for long periods, or erosional discontinuities where previously deposited sediments are washed out before deposition resumes. Both processes can cause ‘jumps’ in time up to several thousand years in the span of a few centimeters of sediment. This means the data needs to be either stretched or squashed to fit into the axis, rather than the axis being stretched. Given the ‘sloppy’ (Sweatman 2020b) nature of the paper at large, the charcoal signal and megafaunal decline may even be misaligned from the rest of the data and correspond much more closely with the highlighted section (Figure 33).

To reinforce their argument, they attempt to compare their Appleman Lake results to two other sediment core records taken up to 1000 km away in New York, claiming they match closely (Gill et al. 2009). The author agrees that these two records share close similarities with Appleman Lake, but these similarities are not good news for Gill et al. (2009). Both of those records demonstrate significant issues with their radiocarbon chronologies, with uncertainties up to ~1400 years, and even a >1000-year dating reversal, highlighted in red boxes (Figure 34); a date from ~145 cm returns an age of ~13.3-13.6 ka followed by a date from ~120 cm of ~14.0-14.7 ka, with another date from ~70 cm of 13.3-13.6 ka (Figure 34). Soon after the date of ~13.3-13.6 ka from ~70 cm is a major change in vegetation accompanied by a major biomass burning episode and megafaunal decline (shown in green box B) suggesting the YD most likely occurs at ~50 cm in this core (Figure 34).

The only correlation between Gill et al. (2009)’s data and these other two sediment cores is their significant chronological misalignments to what is clearly the YD onset; both graphs show that the megafauna disappeared at the same time that vegetation rapidly changed and major biomass burning episodes occurred, and their chronologies are highly unreliable. Gill et al. (2009) joins the list of scientifically meritless papers published against the YDIH early in its history. The merits of papers have been artificially inflated by critics and proudly presented to the public, like Rafiki displaying the newborn Simba to the masses of animals below Pride Rock, as being fatal to the YDIH; only when examined more closely is the infant prince of the pride lands revealed to be just a lump of elephant dung (Figure 35).

Another study released in 2009 that has long been championed as having debunked the YDIH specifically addresses claims about changes in fire regimes around the YD onset; unlike Gill et al. (2009) and other scientifically meritless contributions, this paper is not too bad. Marlon et al. (2009) used 35 charcoal & pollen records from lake sediment cores throughout North America to examine the relationship between changes in fire regime and the YD onset. Interestingly, they seem intent on focusing on whether fire regimes around the YD onset, a significant cooling event, are comparable to fire regimes at the YD termination, a warming event of higher magnitude:

“…The well-documented rapid climate changes of this time alone may have triggered increased fire at a regional scale. To separate these effects, we compared the response of fire during intervals of rapid climate changes at the beginning and end of the YDC. Fire-episode events that occurred during the transitions into and out of the YDC were identified in both the high- and low-resolution records to determine whether fire episodes, regardless of magnitude, were more likely to occur at 12.9 ka than at 11.7 ka.”

Their data show a clear increase in fires at the YD onset, which decreases during the YD, and then gradually increases into the early Holocene (Figure 36). While the “peak” at the YD termination at first appears higher and narrower than at the onset, the reader must pay close attention to the data; that peak is based on between 0 and 1 dates (records), while the elongated peak at the YD onset is based on up to 4 times as many dates (Figure 36). This means there is a lot more data supporting the ‘onset’ peak than the ‘termination’ peak. Despite this being interpretable as supporting the idea that an impact event at the YD onset may have triggered the fires, they conclude the opposite. Because their data shows fire frequency gradually stepping upwards between 15 ka and ~13 ka, with the largest peak (prior to ~10.5 ka) occurring around the YD onset, they claim there was no continent-wide burning event at the YD onset. Instead, to explain this peak at the YD onset, they invoke:

“…noise and local variability, human activity, or megafaunal declines”,

and explain away a cosmic impact trigger by stating that:

“…such patterns are more likely a result of spatially complex climate controls and/or vegetation changes.”

They conclude with:

“…we find no convincing evidence that the observed changes in fire activity were caused solely by changes in human or herbivorous megafauna populations.”

In other words, despite the changes in fire regimes being anomalous and interpretable as a major biomass burning event, they prefer to invoke vague and complex processes rather than entertain a cosmic impact event. These are weasel words that essentially boil down to preferring their own opinions over those of others. It is obvious that impact events occur less often than normal wildfires, so of course it is statistically more likely that any given fire was caused by non-impact processes; this has no bearing whatsoever on whether any given fire was not by an impact. If there is good reason to believe a major impact occurred at the YD onset, any major fires coeval with the impact are more likely to have been caused by that event. This argument is essentially what the bulk of criticism against the YDIH boils down to; of course, each different geochemical impact proxy or ecological consequence, taken in isolation, can be explained away by other processes! But that is not how science works, especially not archaeology, where consideration of the entire context is the most basic requirement of interpretation! If they are all considered in context with each other (Figure 37), the most likely and parsimonious explanation is a catastrophic impact event. How many geochemical signals and ecological consequences need to occur simultaneously before an impact can be considered to have been responsible? To the most zealous critics, anything short of video evidence of the impact occurring would still be insufficient; some of them will find a way to rationalize that it is a deep fake created by Allen West.

As the reader may recall, the effects of the YD were supposed to have been minimal in the Southern hemisphere, where cooling should not have occurred. However, a collaboration between the CRG and South American researchers excavating a paleontological site in Patagonia produced evidence of significant vegetation changes and a major spike in charcoal at YD onset (Pino et al. 2019, Figure 38). This paper is one of the most thorough and robust papers in the history of the YDIH; it conclusively demonstrates that impact-related microspherules, a major platinum anomaly (discussed later), a major biomass burning episode, major changes in vegetation dynamics, and the megafaunal extinctions occur simultaneously (Pino et al. 2019). It also presents a detailed explanation of the YD impact scenario.

Despite their flawed conclusions, Marlon et al. (2009)’s comprehensive biomass burning study was generally of high quality; if the CRG wanted to establish widespread biomass burning at the YD onset, they would have to produce thorough and convincing evidence in response. So, after 9 years of research building upon Marlon et al. (2009)’s work, Wendy Wolbach, a top expert on impact-triggered wildfires at the K-Pg boundary, led the CRG in publishing a colossal 2-part response. They examined charcoal and soot records from 152 lakes, marine & ice cores, and terrestrial sediments from four continents, and found the largest peak in biomass burning in the latest Quaternary occurred at the YD onset (Wolbach et al. 2018a, 2018b). Using the same methods as Marlon et al. (2009), they show that in 30% to 44% of the records, fire frequency is highest or second highest in the 400-year window that encompasses the YD onset (Figure 39). Furthermore, they demonstrate that 66% of records across four continents exhibit at least one major charcoal peak within 150 years either side of the YD onset, and 83% have major charcoal peaks within 300 years either side (Figure 39). The average uncertainty of radiocarbon ages used in the study is ~179 years (Wolbach et al. 2018b).

Furthermore, they demonstrate prominent peaks of multiple coeval geochemical proxies of biomass burning at the YD onset; aerosols like ammonium (NH₄), nitrate (NO₃) formate, acetate & oxalate from ice cores (Figure 40) and other proxies from other sources indicate that up to 9% of the Earth’s terrestrial biomass was burning simultaneously at the YD onset. According to their calculations, this may have exacerbated, or even caused the significant cooling documented at the YD onset; the massive injection of smoke and soot into the atmosphere may have caused an impact winter, blocking out the sun for up to 6 weeks.

The impact winter alone would have had significant effects on vegetation dynamics, without even factoring in the burning itself; more than a month of darkness would significantly impact vegetation in the affected regions. Wolbach et al. (2018a, 2018b) used a global platinum anomaly from the YD onset in the GISP2 ice core discovered in 2013 as their datum for the YD onset, which will be discussed in depth in the next section. While the two-part study contains many graphs, a deep-sea core taken from ~2500 m deep off the Pacific coast of Papua New Guinea clearly demonstrates the massive scale of burning at the YD onset (Figure 41). Interestingly, the monstrous spike of charcoal around 50 ka is not mirrored in any of the other records such as ice cores, and as far as the author is aware, has not been well-researched. However, it roughly correlates to several known impact events; the ~1200 m Barringer Crater (Meteor Crater) in Arizona (Nishiizumi et al. 1991; Osinski et al. 2015), securely dated to ~50 ka, and the ~1900 m Lonar Crater in India, loosely dated to ~50 ka (Sengupta et al. 1997; Nakamura et al. 2014). Whether the charcoal spike was caused by the results of this bombardment episode or some other catastrophe is unclear, and not relevant to the YDIH. The original study that produced this data attributed both the 50 kya and the YD peak to anthropogenic fire regime changes but was published prior to the proposal of the YDIH (Thevenon et al. 2004). If they had known of a potential catastrophe coinciding with the YD peak, their interpretation may have been different.

An 26-page response was published 2 years later by Holliday et al. (2020) criticising a few key aspects of Wolbach et al. (2018a, 2018b)’s work; the bulk of their criticisms are merely rehashing their prior efforts at quibbling earlier work on the YDIH and are largely irrelevant. One of the few claims addressing Wolbach et al. (2018a, 2018b)’s data is that some ice core data appears to have been misaligned by between 10 and 50 years, and ‘smoothed’, resulting in small peaks appearing larger in their graphs (Figure 42).

In response, Wolbach et al. (2020) state that the alleged ‘misalignment’ may have resulted from Holliday et al. (2020)’s misunderstanding of the age scale of the data, which was originally plotted by Paul Mayewski (Mayewski et al. 1997), a coauthor of Wolbach et al. (2018a, 2018b, 2020); essentially, Holliday et al. (2020) claimed they know the data better than the scientist who originally plotted it. However, Wolbach et al. (2020) do not directly address the accusation of smoothing the data. While this is unfortunate, smoothing of the data does not amount to scientific malfeasance, does not significantly influence their interpretations or conclusions, and does not modify the data in any way, just its visual presentation.

While this is the only criticism of the actual data presented in Wolbach et al. (2018a; 2018b) in their 26-page response, there is one more relevant, particularly egregious claim made by Holliday et al. (2020); they claim there is no evidence that impacts can trigger wildfires:

“Wolbach and colleagues’ (2018a) attempt to make a case for burning at the YDB based in part on the assumption that an impact would trigger widespread fires, but the evidence for a link between extraterrestrial impacts and wildfires is weak…”

This is patently absurd; of course, impacts (particularly airbursts) can cause wildfires, and are documented to have done so, such as the Tunguska airburst of 1908, which flattened >80 million trees and caused intense wildfires (Jenniskens et al. 2019). They are well-established as being capable of melting rock at temperatures of several thousand degrees Celsius, which are sustained far longer than similar (lower) temperatures produced by lightning, one of the primary and most frequent ignition sources for natural wildfires (Krause et al. 2014). Also, note the intentional fallacy by Holliday et al. (2020), who phrase it as ‘an impact’ despite Wolbach et al. (2018a) providing the one of the most detailed explanations of the proposed YD impact scenario that has ever been published; they specifically invoke a swarm of airbursts, which are distinct from impacts, discussed in more detail later. Most recently, an excellent article was published in Science (though only because they don’t cite the YDIH a single time) providing some of the best evidence yet of massive wildfires and megafauna extinction at the YD onset (O’Keefe et al. 2023). Martin Sweatman has covered this study in detail and provides top tier analysis of its importance for the YDIH on his blog, Prehistory Decoded; I would like to direct everyone to his blog, as he has also written extensive rebuttals to the ‘comprehensive refutation’ paper. In summary, recent research thoroughly and conclusively demonstrates that large-scale biomass burning truly did occur over much of the world at the YD onset; extraordinary biomass burning still very much stands as evidence supporting the YDIH.

Part IV

Impact Scenario

Some common criticisms of the YDIH are that the proposed impact mechanism has been inconsistent (Holliday et al. 2023). Firestone et al. (2007) originally proposed two potential impact mechanisms as follows: One scenario posits that the Laurentide ice sheet was impacted by one or more comets of ~2 km diameter, with the ice absorbing the energy of the impacts and preventing the formation of significant craters in the bedrock below. The other posits that the Earth encountered a stream of debris from an already-disintegrated comet, resulting in a shower of impacts and airbursts of varying sizes over much of the world (Firestone et al. 2007). While other impact mechanisms have been explored and proposed in the 15 years since publication (Israde-Alcantara et al. 2012; Petaev et al. 2013; Usatov 2020), one of the two original scenarios is still favored, though with slight modifications as the hypothesis has evolved. The fact that multiple different impact mechanisms have been explored in the 15 years since its proposal is not a valid argument against the YDIH. As new data comes to light, hypotheses are adjusted; this is simply how science works.

The Taurid meteor stream (Figure 43) was formed by the remnants of a disintegrated centaur comet, perhaps >100 km in diameter, that most likely entered the inner solar system sometime within the last ~30,000 years; the timing of its entry was calculated by reverse engineering the dispersal of observable debris (Steel & Asher 1996; Ferrín & Orofino 2021). Every year, Earth crosses the multi-million-km cross-section of the debris stream, sweeping up any fragments in its path, resulting in rapturous meteor showers in the night sky. However, it hasn’t always produced just a pretty light show; early in disintegration cycles, comet fragments can be quite large, gradually becoming smaller over time due to hierarchical disintegration (Napier 2019). Today, at least 88 distinct cometary bodies are linked to the initial fragmentation event, each one making its own way through the complex, forming their own sub-complexes of fragments (Ferrín & Orofino 2021). As debris from the Taurid Complex (TC) has been raining down on Earth periodically for ~30,000 years, it is self-evident that Earth has, on occasion, encountered particularly dense and violent regions of the complex, resulting in larger impacts (Napier 2010). The ‘violent meteor shower’ scenario is the preferred explanation for the evidence found in the YDB layer and is likely responsible for other impact events periodically throughout the development of the complex. Such events may have formed the Alaskan/Yukon muck deposits (Hagstrum et al. 2017) and a recent study provided strong evidence for multiple large airbursts over the Atacama desert just 300 years after the YD onset (Schultz et al. 2021). The Taurids are also thought to have been responsible for the Tunguska airburst of 1908 (Kresak 1978; Napier 2010).

Napier (2019) calculated that 10 encounters with the denser areas of a Taurid-like debris complex within a 15,000-year disintegration cycle yields a ~90% chance of a ~6,000 megaton bombardment event. Another of his models predicts that bombardments of ~6500 megatons are expected every 17,000 years, with ~5000 megatons bombardments expected every ~3,800 years (Napier 2019). He concludes as follows:

“…there is a reasonable expectation of one or more brief meteor ‘hurricanes’, with intensities far beyond modern experience, in the course of disintegration of the progenitor. Enough meteoric smoke may be created during such encounters to generate sudden cooling of some years’ duration, along with widespread wildfires. The terrestrial upsets at the YDB of 12900 BP, and the simultaneous collapse of early civilizations around 2350 BC, may have been triggered by events of this character.”

It is important to remember that Earth-crossing debris streams from decayed giant comets are not hypothetical; arguments for their existence, or the rate at which they occur, are not ephemeral statistical models. Most of the major meteor streams are still producing airbursts on Earth today, as they have done for the last tens of thousands of years. Centaur comets >100 km in diameter can enter the inner solar system in Earth-crossing orbits on geologically relevant timescales from a variety of sources (Asher et al. 1994; Emel’yanenko et al. 2005; Napier 2015; Lisse et al. 2020). They likely pose a far greater threat to Earth than stray asteroids and contribute a significant percentage of hazardous objects to the inner solar system over long periods (Clube et al. 1996; Napier et al. 2015). Wolbach et al. (2018a) modelled the orbital evolution of a Chiron-like centaur originating in the Saturn/Uranus region for their study, demonstrating that it became Earth-crossing multiple times within a period of ~40,000 years (Figure 44).

While Petaev et al. (2013) made it clear that the novel geochemistry of the global platinum anomaly cannot have resulted from the impact of a comet, they acknowledged that it is highly unusual, and may represent the debris of a destroyed proto-planet:

“One of the plausible sources of the Pt spike is a metal impactor with an unusual composition derived from a highly fractionated portion of a proto-planetary core.”

Their claims as to the species of impactor rely on the fundamental assumption that the fledgling discipline of impact science has a good grasp of the chemical composition of all potential impactors in our galaxy; while it is true that no cometary material has yet been recognised as having compatible geochemistry, the fundamental assumption is almost certainly untrue. Too often in science, because funding depends on perceived confidence and certainty, scientists tend to be reluctant to admit they do not know something; ‘knowledge’ is simply what has been discovered so far, and there is a distinct, deeply ingrained cultural resistance to epistemic humility and informed speculation. I see it as perfectly reasonable to assume the existence of novel impactors that we currently do not recognise, especially concerning interstellar or Oort cloud objects, particularly comets. The Oort cloud, whose existence is merely a hypothesis (Zwart et al. 2020), as it is too far away, and thus too dark for us to observe, is hypothesised to contain all sorts of novel objects; its expected population is the remnants of destroyed proto-planetoids formed during the birth of the solar system (Levison et al. 2010; Zwart et al. 2020), likely with a wide variety of novel geochemical compositions. Approximately half of inner Oort cloud material, and a quarter of the outer material, could be non-native to our solar system, captured from the outer edges of other solar systems from our star’s birth cluster (Zwart et al. 2020). In fact, there are several objects in our solar system known to have novel geochemistry; the giant >220 km asteroid ’16 Psyche’ is ~90% metal, including iron, nickel, platinum, and gold (Petrescu 2020). Another ‘potentially dangerous’ (Ipatov et al. 2016) ~220 m diameter asteroid ‘2011 UW158’ with an exceptionally fast rotation (~37 minutes) contains ~$5.4 trillion of platinum (Gary 2016).

Observational astronomers, planetary scientists, and dynamicists all use different attributes to classify objects as comets or asteroids (Jewitt 2012). Furthermore, traditional understandings of the distinctions between comets and asteroids have become increasingly blurred in recent years (Gounelle 2011), particularly with the discovery of active asteroids. Active asteroids are now known to undergo fragmentation events like comets, and some well-known asteroids may even be the extinct cores of comets that have shed their volatile contents, becoming dormant (Hartmann et al. 1987; Fernandez et al. 2005; Babadzhanov et al. 2015). Admittedly, there is no known evidence of objects with similar geochemical compositions to the proposed YD impactor currently residing in the TC. However, extant TC debris includes asteroids (Asher et al. 1993), and objects with intra- and inter-heterogeneous compositions (Ferrín & Orofino 2021). Intra-heterogeneous objects are heterogeneous in composition within the same object, while inter-heterogeneous objects are multiple objects, presumably from the same parent body, that have heterogeneous compositions compared to one another. If the YD impactor originated from the TC, it may have been one of these extinct cores, with the extant debris being the remnants of the Progenitor Comet Encke’s outer crust (Clube & Napier 1984). Nobody can claim to know what giant comets contain in their cores (Cochran et al. 2015); this can only be hypothesised based on current observations and understandings, which are constantly updated as new discoveries are made. For example, for more than 40 years, comets were confidently ‘known’ to be ‘dirty snowballs’, composed mainly of ice with a minority dust component (Sykes & Walker 1992). However, now they are understood to be more akin to ‘snowy dirtballs’; conglomerates of whatever debris they agglomerate as they travel through space, containing dust-to-gas ratios that vary by a factor of 30 or more (Miles 2016). There is absolutely no reason why a giant Oort comet cannot have swept up and incorporated a large fragment, or many smaller fragments, of a Pt-rich proto-planetary core during the billions of years of its janitorial career.

One major reason for the ‘violent meteor shower’ being the preferred scenario is that no impact crater has yet been definitively linked to the YD onset, though many have been proposed as potential YDIH candidates. These include, but are not limited to: the Bloody Creek structure in Nova Scotia (Spooner et al. 2010), the Iturralde structure in Bolivia (Malkova et al. 2013), the Roseau structure in Minnesota (Rodriguez 2020), the Corossol crater in Quebec, (Lajeunesse et al. 2013), the Pagasitic Gulf in Greece (Dietrich et al. 2019), the Hiawatha crater in Greenland (Beech et al. 2020), and the Brushy Creek feature in Louisiana (Heinrich 2003; Webb et al. 2018). The lack of a large, established crater at the time of the YDB, combined with the discreet, localised ejecta fields, leads proponents to favour a bombardment of airbursts, where the impactors detonate in the atmosphere, forming patches of impact material on the ground; crater-forming impacts, where the impactor reaches the surface intact and forms widespread ejecta fields, are expected to produce more uniform distributions of impact proxies. Stony chondrites or metallic asteroids, particularly larger ones, often make it to the ground and form craters, and are therefore less likely to have been the culprit. As mentioned, discreet areas of impact material such as SLOs/melt glass near the multiple suspected ground-zero sites are suggestive of an airburst or small-scale impact origin (Bunch et al. 2012; Moore et al. 2020). This is a sticking point for Mark Boslough, Critic-In-Chief of the YDIH, known for creating computer models of airbursts. Boslough persists in claiming that the preferred YDB impact scenario is impossible, while demonstrating a complete and utter misunderstanding of the claimed scenario that borders on willful denial of the evidence (Boslough 2012; Boslough 2013; Boslough et al. 2012; 2013). This screenshot (Figure 45) of a recent comment by Boslough on social media clearly demonstrates that he still, to this day, argues against an impact scenario of his own invention, rather than that proposed by YDIH proponents (Napier et al. 2013).

Rather than the ‘violent meteor shower’ scenario, Boslough consistently argues against a scenario akin to the 1994 ‘string of pearls’ impact of Shoemaker-Levy 9 into Jupiter (Boslough et al. 2013). That comet fragmented as it crossed the Roche limit of Jupiter, torn apart by its gravitational pull, shortly prior to impact (Boslough et al. 2013), rather than a prolonged disintegration event. Boslough rightly claims that this scenario is physically impossible for an impactor into Earth, as Earth’s gravity is insufficient to tear such an object apart. However, no YDIH proponent has ever proposed such a scenario; on the contrary, astronomer proponents such as Bill Napier’s team, have published on the ’violent meteor shower’, or “Coherent Catastrophism” model for decades. The ‘British Neo-Catastrophists’ as they have been called (Morrison 1997), have long been derided by paradigmatic holdouts like Boslough and Morrison. These holdouts have never presented a convincing argument against the “Coherent Catastrophism” model, which enjoys widespread support in the astronomical community, and is strongly supported by recent work (Ferrin & Orofino 2021); instead, they just say that Napier and his colleagues are wrong, and resort to fallacies of association, equating them with acolytes of Velikovsky (Morrison 1997):

“While I believe that the British neo-catastrophists are wrong about the threat to Earth, their work is science, not pseudoscience. They are making their case to other scientists, and time will sort out who is right and who is wrong. They do, however, sometimes attract the attention of fringe elements. For example, the Society for Interdisciplinary Studies (SIS), a British group that espouses a skeptical philosophy but includes many defenders of Velikovskian ideas, is sponsoring a conference that features Clube and focuses on evidence for cosmic catastrophes in the ancient world. In fact, the work of Clube and Napier attracts many people who were once impressed by Velikovsky, such as Leroy Ellenberger, at one time a member of the Velikovsky inner circle and now one of the most outspoken critics of his current followers.”

The paradigmatic holdouts, led by David Morrison (1997), staked their reputations on “Coherent Catastrophism” being wrong more than a decade before the YDIH was proposed, which may be why they are so vehemently against it. Clearly, they understand the model of “Coherent Catastrophism”, but refuse to acknowledge it as the preferred YDB impact scenario. This is proven below, in a statement by Morrison in private communications (Figure 45) leaked by Boslough for his own gain from 2015:

“Scientifically, there has been no physically plausible suggestion of any way to produce a widespread comet shower over North America.”

Based on the evidence, this can only be an intentional misrepresentation; Morrison was a coauthor of the critical comment on Israde-Alcantara et al. (2012) that disparaged the YDB impact scenario as “inconsistent” and “defying physics” (Boslough et al. 2013), to which Bill Napier led the response to (Napier et al. 2013). In the response, titled “Decades of comet research counter their claims”, Napier and seven other YDB team members link the YDB event with the Taurid Complex, and the ‘Coherent Catastrophism’ model provides an explanation for the YDB impact event. Morrison knew his claim in this email (Figure 46) was not true for at least 2 years prior to making it.

Essentially, most planetary defense models are based on the false assumption that the flux of cosmic material into the inner solar system over time is, and has always been, constant, rather than variable; Napier et al. (2013) state the following:

“Fragmentation yields a power law distribution of mass with population index ~1.7, from which interception of 10¹⁴ g at 30 km/s may yield several impactors with energies up to 5,000 megatons, fully adequate for surface melting… Current impact hazard assessments predict one such impact with recurrence time in excess of 50,000 y, but these assessments are based on the erroneous assumption of a steady-state comet population. The occasional injection of giant, short-period comets negates this assumption over timescales relevant to civilisation.”

Morrison was the first to quantitatively estimate the cosmic impact hazard (Chapman & Morrison 1994), and in the subsequent decades has consistently excluded ‘Coherent Catastrophism’ from his models, as recently as 2019 (Morrison 2019). Morrison and other paradigmatic holdouts have the most skin in the game; the false paradigm they have forged over decades stands to crumble if the YDIH, using Coherent Catastrophism as its vehicle, prevails.

Despite his blanket refusal to entertain the possibility of multiple airbursts at the YD onset, Boslough veritably fawned over a paper by Schultz et al. (2021) that reports multiple airbursts over Chile around 12,600 BP, just ~300 years after the YD onset. Episodic airburst events are entirely consistent within the ‘violent meteor shower’ bombardment scenario detailed above, yet Boslough vigorously denies that any similar events could have occurred ~300 years prior to Schultz’s proposed airbursts. Schultz was a coauthor on Firestone et al. (2007) and formerly a member of the YDB team, but disassociated himself shortly before publishing his 2021 paper, which does not cite a single paper supporting the YDIH. Today, despite the objections of paradigmatic holdouts, a ‘violent meteor shower’ raining airbursts over much of the world, perhaps annually over a 10-year period, remains the most likely scenario for the YD impact event. However, the recent discovery of at least one large crater (31 km or 19.2 miles in diameter) that could date to the YD onset could mean that the ‘violent meteor shower’ scenario is not required. For comparison, the dinosaur-ending Chicxulub crater is 180 km or 110 miles in diameter.

Hiawatha Crater

In late 2018, a massive 31 km impact crater was discovered below the Hiawatha glacier in northwestern Greenland (Kjaer et al. 2018). The data was published in the journal Science, but an earlier version of the paper had been submitted to a different journal and included the words “Late Pleistocene” in the title of the paper. Furthermore, a source familiar with the matter has stated that this original paper contained many citations to papers supporting the YDIH; all were removed prior to resubmission to Science, alongside allusions to a late Pleistocene age, presumably due to objections from reviewers, but nevertheless, those authors considered it a prime candidate for YDIH crater. From the outset, it is important to note that the YDIH does not live or die on the dating of the Hiawatha crater. The ‘violent meteor shower’ scenario is and has been the favored mechanism by most proponents from the beginning; it is by far the more likely of two scenarios that can explain the YDB evidence, with the other scenario being a single impact into the ice sheet. It is also possible for both scenarios to have occurred simultaneously.

So what exactly led this independent team of prolific Scandinavian glaciologists to link this crater to the YD so early in the process? There are several features of the Hiawatha crater that strongly suggest that it is incredibly young, some of which were known at the time of publication, with others published in subsequent studies:

• The oldest ice in the crater dates to the YD onset ~12.8 ka (Kjaer et al. 2018).

• Simulations suggested that the impact occurred into 1.5 to 2 km of ice (Silber et al. 2021).

• The hydration of impactites is consistent with a delayed inundation by water from melting ice, rather than rapid inundation as occurs during a submarine impact scenario, or no inundation as occurs during a subaerial impact (Garde et al. 2022).

• Recovery of impactites containing scorched organic carbon from Pleistocene/Pliocene conifers that grew in Greenland between 3 and 2.4 million years ago (Gustafsson 2020; Garde et al. 2022).

• Geothermal heat still emanates from the bedrock of the crater, producing basal meltwater on contact with the ice (Kjaer et al. 2018).

• The crater rim and central uplift are pristine and virtually untouched by erosion from the ice sheet flowing over it (Kjaer et al. 2018).

• The bottom ice layer is characterised as containing a layer of debris frozen in the ice ~10 m above a shallow groundwater table (Bessette et al. 2021).

A few months after Hiawatha was published, the ~36 km Paterson impact crater was discovered approximately 183 km upstream from Hiawatha (MacGregor et al. 2019). While a lot of work has been done on Hiawatha crater, very little study has been done on the Paterson crater, likely because it is entirely inaccessible below the ice sheet, while Hiawatha is on the edge. A 2020 study using Earth gravity models supports the idea that both Hiawatha and Paterson craters are indeed of impact origin (Klokocnik et al. 2020). Studies have explored potential relationships between the unknown Hiawatha impactor and the well-studied Cape York meteorite, which Boslough tried to link to the global Pt anomaly (Boslough 2013) discovered by Petaev et al. (2013), and found they are unlikely to be related (Gustafsson 2020; Beech et al. 2020). The potential of Hiawatha and Paterson being twin craters which formed during the same event has also been explored, and was also found unlikely, but possible (MacGregor et al. 2019; Beech et al. 2020).

Despite abundant evidence cited above suggesting that Hiawatha crater is incredibly young, two recent studies have used multiple different methods to date the crater to ~58 million years ago; just 8 million years after the dinosaurs were wiped out by the Chicxulub impact at the Cretaceous-Paleogene boundary. When mineral grains are modified by shock metamorphism, their radiometric ‘clock’ can be reset; using various radiometric dating methods on grains shocked by an impact can allow determination of how long ago they were shocked. Kenny et al. (2022) used two methods: Argon⁴⁰/Argon³⁹ dating on 50 grains from glacial sediments and, Uranium/Lead dating on shocked zircon grains from pebble-sized impactites. 29 of 50 grains returned Ar⁴⁰/Ar³⁹ ages of between 58 and 60 million years, though interestingly, the authors did not report any of the other ages; what ages did the 21 unreported grains return? The Pb²⁰⁶/U²³⁸ dates cluster around 1.915 billion and 58 million years ago, indicating an impact occurred 58 million years ago into 1.915-billion-year-old bedrock, consistent with the bedrock in this region of Greenland. A second study (Hyde et al. 2022) dated shocked monazite using Uranium/Lead dating and produced dates ranging between 59 ± 36 million years and 73 ± 15 million years, consistent with the results from Kenny et al. (2022).

This news was naturally well-received by the most vehement critics of the YDIH, who performed a victory lap around social media proclaiming the zombified corpse of the YDIH had once again been put to rest. However, there are several major problems with Hiawatha being this old:

• How did an impact that occurred 58 million years ago modify organic carbon from the Pleistocene?

• How are the impactites consistent with an impact into ice when there was no ice on Greenland 58 million years ago?

• Why does the oldest ice in the crater date to the YD?

• How is the crater in pristine condition after 58 million years of subaerial weathering and trillions of tons of glacial ice bulldozing it periodically throughout the many glacial cycles throughout the Pleistocene?

• How is the crater still emitting remanent geothermal heat after such a long time? – It is still metaphorically smoking!

Despite the self-evident fact that Hiawatha is indeed an impact crater, when it was first published in 2018, long-time YDIH critic Christian Koeberl spoke out against the idea that the crater was formed by an impact in media reports (Patel 2018). In this article, Koeberl states the following:

“Even if one would agree that there are some shocked quartz grains, one swallow does not make a summer, just like one rock from an unknown source with a few shocked quartz grains does not make an impact crater”.

I wholeheartedly agree with this assessment, and notes that it also applies to the impactites and sedimentary grains that have been dated to 58 million years ago. These grains were sampled from the same location that produced the impactites containing impact-modified Pleistocene carbon; from the banks of the spillway produced by meltwater flowing out from beneath the glacier (Figure 47). As Koeberl succinctly puts it, dates from a few grains are not sufficient to unambiguously date the crater.

Notably, as shown in Figure 48, Paterson crater is directly upstream from Hiawatha, meaning impactites from Paterson can easily have been transported into Hiawatha and washed out through the spillway alongside material from Hiawatha; only drilling through the ice and sampling the bedrock itself can produce a reliable date for Hiawatha.

Based on the geomorphology of Paterson compared to that of Hiawatha, it has been eroding much longer; a larger percentage of material from Paterson would be expected to have washed out of Hiawatha’s subglacial spillway than from Hiawatha itself, as it was transported by the ice and dumped into Hiawatha. Given that impactites of both Pleistocene (2.6 Mya to recent) and Paleocene (66-56 Mya) ages have been sampled from there, and cannot possibly have resulted from the same impact, Hiawatha crater must be drilled. While the dates from 58 million years ago put a damper on the idea that Hiawatha dates to the YDIH, this idea is far from debunked. Regardless, the YDIH is not reliant on Hiawatha crater for validity or veracity.

Part V

Volcanism and the Younger Dryas Onset

Over the years, volcanism has occasionally been invoked as an alternative to the idea that various cosmic impacts have caused extinctions in the past; to this day there are still a minority who argue that the super-eruption of the Deccan Traps in India was responsible for the extinction of the dinosaurs, rather than the Chicxulub impact (Pal et al. 2020). Particularly in the last 5 years, volcanism has been invoked multiple times as potentially being responsible for the YD cooling, extinction event, and some of the geochemical signatures at its onset; it is counter-intuitive that this idea has gained traction recently (Baldini et al. 2018; Green 2019; Cheng et al. 2020; Jorgeson et al. 2020; Sun et al. 2020, 2021; Abbott et al. 2021), as the evidence supporting such a notion has never been weaker. The abnormally large Laacher See eruption (LSE) was previously thought to have occurred very close to the YD onset, but the most comprehensive work, published recently, has firmly placed its eruption ~200 years before the YD onset (Engels et al. 2022). Even if the new high-resolution chronology is wrong, and it did occur near the YD onset, their close proximity is not evidence that it was the trigger; there were at least 82 large bipolar volcanic eruptions between 60,000 and 12,000 BP (Svensson et al. 2020), and none of them are linked with millennial-scale events like the YD. In fact, one of the largest sulfur-rich eruptions in the last 2 million years (Innes et al. 2024) only resulted in minor cooling for a few years, and this should be kept in mind for the following sections.

A study by Baldini et al. (2018) proposed that the injection of sulfur-rich aerosols from the LSE provided the initial trigger for the cooling, which was then amplified by atmospheric feedback mechanisms. Notably, they admit in their paper that the LSE is now widely thought to have occurred ~180 years before the YD onset but justify their claims to the contrary using a decade-old study containing outdated science. At first glance, their claims could seem plausible; taken uncritically, the outdated chronology “strongly suggests” (Baldini et al. 2018) the LSE and YD cooling were synchronous. It is true that the LSE was indeed rich in sulfur, which has powerful albedo effects and is effective at reflecting and scattering solar radiation before it can warm the surface. However, the chronology used by Baldini et al. was developed using proxies from a different volcanic eruption, which were then correlated between Meerfelder Maar lake sediments and Greenland ice core records (Lane et al. 2013). Clearly, any chronology developed using proxies from other eruptions will be inferior to those directly developed using the LSE proxies. Essentially, Baldini et al. (2018) argues in support of the traditional chronology for the LSE, which has been falling out of favour for many years, and has finally been shattered by the latest work (Engels et al. 2022).

Other studies published since have reaffirmed that the LSE is distinct from the YDB, and thus the cooling and the proposed cosmic impact event, by approximately 200 years. For example, In Figure 49, Kletetschka et al. (2018) demonstrate the presence of a ~3 cm gap between the initial deposition of the LSE tephra and the appearance of cosmic impact materials deposited at the YD onset. This study from Stara Jimka, a lake in the Czech Republic, clearly indicates two things: that the cosmic impact ejecta from the YD impact event was not formed by the LSE, and that the impact ejecta, rather than the LSE tephra, corresponds to the YD onset (Kletetschka et al. 2018).

The work of Engels et al. (2022) played a key role in establishing the new chronology for the Laacher See eruption; it definitively shows that the YD cooling and associated vegetation changes began up to 195 years after the eruption (Figure 50). Engels et al. (2022) combine multiple proxies including lake varves (annual layers deposited in lake sediments), from multiple records, using the LSE tephra as a hard anchor; their methods allowed the most precise dating of the relationship between the LSE and the YD onset ever achieved. While their findings confirm the LSE has no relation to the YD onset, theirs is not the only study to have completely invalidated this idea.

A 2020 study by Svensson et al. is the most comprehensive effort ever made to correlate between different ice cores in the Northern and Southern hemispheres using volcanic geochemistry and diagnose abrupt climate changes during the last glacial (Svensson et al. 2020). Their analyses produced some interesting information that is very relevant to the debate over what caused the YD, and this is something they implicitly explore in their paper; in their Supplementary Information they plotted the GISP2 Platinum anomaly alongside the bipolar volcanic signatures detected in their study (Figure 51).

Importantly, their plot shows the GISP2 platinum directly correlating with the onset of YD cooling in the GISP2 and NEEM cores from Greenland, though not in the GRIP and NGRIP cores; the YD onset in these two cores more closely coincide with a large volcanic (acidity) signature that manifests in both the Northern and Southern hemispheres. It is also interesting to note that the Pt spike closely correlates with cooling in two of the three Antarctic ice cores, WDC and EDC, while the LSE correlates with warming in all three (Figure 50). The large bipolar acidity spike is more than likely representing the LSE, as there are no larger eruptions around this time capable of producing a global signature (Figure 50). While the reasons for the disparities between ice cores are complex and were likely the source of misunderstandings in previous studies, such as the chronology used by Baldini et al. (2018). The discrepancy is likely due to differences in the production or preservation of ice layers in different areas, which can occur for several reasons. Erosional discontinuities are well-known to occur in sedimentary sequences, and ice cores can be similarly affected, but by different processes such as melting rather than wind or water erosion. In the relevant timespans, it is not unreasonable to assume that many annual or sub-annual layers may be erased or prevented from depositing over the years by various means. For example, if there is a particularly warm winter, there will be less ice deposited and more ice melted. Melting rates can also vary within the space of a few meters; even shadows from cliffs etc can affect the spatial distribution of local surface melting during summer. All four major Greenland ice cores were taken from somewhat close locations; NEEM is closest to the Hiawatha crater, northwest of NGRIP, which is north of GISP2 and GRIP, which are quite close together near the summit of the ice sheet in central Greenland (Figure 51). Despite the relatively close proximities of the different cores, it is clear that there is some variability in ice layer deposition based purely on the variable depths of identical horizons between them; for example a layer deposited in the same year may be 7 mm thick in one core but only 2 mm thick in another, and when dealing with hundreds of meters of core, correlation between them becomes very complex. The depth of the YDB in all four cores varies between 1442.83 m in NEEM and 1710.70 m in GISP2, and the deposition rate between the onset of GS3 and the YD differs drastically (Figure 52); this indicates that ice layer dynamics became more unstable around the YD onset. Of course, it does make sense that as the ice flows downhill from the summit and spreads out laterally, ice layers will become thinner, but this process alone could theoretically erase layers; if there is differential ‘stretching’ between layers, this means individual layers could be missing. Whatever the reason for the differences, they clearly exist; anchoring the records using the volcanic signatures is certainly a legitimate and reliable method.

Sun et al. 2020

One major study linking the YD onset with volcanism was widely reported in the media as a significant challenge to the YDIH (Sun et al. 2020). The authors used the isotopic ratios of the platinum group element osmium to determine whether it is indicative of normal crustal levels or some sort of event; the ratios between ¹⁸⁷Os and ¹⁸⁸Os are known to distinctly differ between the two. Ratios of >1.11 are designated as ‘radiogenic’, or typical of normal crustal ratios, while ratios of <1.11 are designated as ‘unradiogenic’, consistent with extraterrestrial impacts or volcanic eruptions (Sun et al. 2020). The concentration of osmium in the ‘unradiogenic’ samples are then compared to average concentrations in CI-chondrites (486,000 ppt) and continental crust (30 ppt) and found to be inconsistent with both, suggesting a mantle origin, thus volcanism. However, CI-chondrites are a rare class of carbonaceous stony meteorites and are only tentatively linked to a cometary origin (Anders 1975; Campins & Swindle 1998); it is unclear why only this class of meteorites was chosen for the comparison and all others were excluded. Different classes of chondrites can have significant variations in rare earth element (REE) concentrations, both within the same class and between classes (Horan et al. 2003; Tagle & Claeys 2005; Riches et al. 2012), and so the exclusive use of CI-chondritic geochemistry is questionable. Though, to be fair, most studies usually pick one class of chondrite to compare their geochemistry to, so they should not have been expected to run additional comparisons for other classes. Importantly, the authors demonstrate their awareness that the proposed YDB impactor has novel geochemistry yet proceed to use CI-chondrite values anyway.

Based solely on the unradiogenic osmium ratio and ‘non-chondritic’ REE concentrations in their supposed YDB layer, Sun et al. (2020) claim that the YD was likely triggered by the Laacher See eruption. Of course, as the reader has just seen, this notion is simply unsupportable based on other work. Furthermore, according to their interpretation of the data, they have identified four additional volcanic events in the ~4000-year sedimentary record on either side of their claimed YDB (Figure 53). The number of events throughout the section is used to further justify this interpretation; if there are five events with strong signatures, they must be volcanic, because volcanic events occur much more often than impact events. Even assuming their data can be trusted, this argument does not make sense. Why can it not have been four volcanic events and one extraterrestrial event with a similar geochemical signature? Either way, it does not matter, because the following sections demonstrate that this paper has fatal methodological flaws that should have seen it rejected in peer review, and the authors even manipulated their data to suit their conclusions!

Close inspection of their data reveals that sediments were sampled during three distinct sampling episodes over three years in different areas of the cave using different methodologies: HC15 in 2015, HC16 in 2016, and HC17 in 2017 (Figure 54). This is not inherently an issue, but the problem is they combined all the data from each sampling episode into a single dataset; this can only work if specific measures are taken to ensure samples from each episode are comparable, which they did not take. Furthermore, the authors constructed their age-depth model using six radiocarbon dates taken throughout the section and filling in the gaps between the dates with linear interpolation. In doing so, they fall into the same trap as Gill et al. (2009) and countless other studies by assuming that sedimentation rates have been constant at Hall’s Cave throughout the entire ~4000-year section; this is almost never the case, yet this assumption has influenced a significant fraction of all geochronological reconstructions ever undertaken.

As the reader can clearly see (Figure 54), several depths (red and yellow boxes) were sampled multiple times between and within sampling episodes. For example, samples from 151 cm were sampled in all three episodes, and three times in 2017, for a total of five samples supposedly representing the same layer. However, their PGE abundances and osmium ratios are significantly different between samples taken from within what they claim is the same soil horizon; between the 5 included samples from 151 cm, their supposed YDB, the Os ratios range from 0.41 to 2.22, and Pt concentrations range from 64.3 ppt to 435.1 ppt (Figure 54). This means, according to their data, that the osmium ratio and concentration at the 151 cm YDB is simultaneously crustal and non-crustal, which is clearly not the case. Similar discrepancies occur at other depths with multiple samples such as 171 cm; the osmium ratio at this depth is 0.12 (non-crustal) and 1.48 (crustal) simultaneously, and the concentration of osmium differs by 13,990% between supposedly identical samples (Figure 54). Clearly something has gone wrong here; perhaps a closer look at their sampling methodology (Figure 55) can explain what happened.

All heights/depths in Hall’s Cave are obtained using a centralised datum established in 1986 (Sun et al. 2020), shown by the measuring tape in the left panel. Because the soil horizons undulate up and down (Figure 55), samples taken from the same vertical depth, but different horizontal locations, are actually capturing different soil horizons. For their sampling episodes to be comparable, they would need to discard the site-wide datum and use their claimed YDB layer as their datum; essentially, they would need to use stratigraphic boundaries (the transition between distinct layers) as their datum in the same way that volcanic events are used as anchors in other studies. This has been the method used by the YDB team since day one, as demonstrated in the figures from previous sections; typically, the widespread and immediately obvious stratigraphic boundary marking the YD onset is used as the 0 point.

The right-hand panel of Figure 55 demonstrates this issue clearly: each black rectangle represents a different sample year, and all were taken at the same depth based on the site-wide datum (yellow horizontal line). As the reader can see, in this example the YDB (black wavy line) is only just barely captured in the corner of one sample during one sampling episode. This may be why one of the 2017 samples from 151 cm contains a much higher concentration of Pt than the others at the same depth (435.1 ppt); perhaps that sample managed to just barely scratch the true YDB, while the others missed it entirely. This error is catastrophic to their interpretations; their data can only be considered reliable when all three sampling episodes are plotted separately and/or at least anchored by stratigraphy. This error also means that their five claimed volcanic signatures throughout the section are really three events at most, as three is the most that appear in the same sampling episode; at least two of their proposed events were duplicated between sampling episodes. Most interestingly, in the sampling episode with three volcanic events, none of them occur close to the YDB. As if this was not bad enough, there is another, potentially much worse issue with the data in this paper.

Close examination of the depth column in yellow (Figure 54), reveals a consistent sample resolution of 1-2 cm throughout the middle of the section, except for one anomaly. At 151 cm depth, there is an abrupt leap to 155 cm, highlighted in blue (Figure 54) that results in a gap of 4 cm. A sample from 153 cm would slot beautifully into this gap to complete the section, so why is there no sample (or samples) from 153 cm in their data? Because the interpretations and conclusions of this paper are based on correctly interpreting the layer in direct contact with the missing 153 cm sample, its inclusion is very important. Surely this must have been an accidental oversight; the authors, many of them well-respected scientists in their field, would not have intentionally omitted important exculpatory data that could negatively influence their interpretations, right? Perhaps we can learn something about the missing 153 cm sample from other investigations?

Luckily, the YDB at Hall’s Cave was examined for impact proxies by the YDB team back in 2009 (Figure 56), so what did they find? The study, led by Thomas Stafford Jr. (2009), principal investigator at Hall’s Cave, in collaboration with the early YDB team, found evidence of multiple impact proxies in the YDB layer; a discrete peak of nanodiamonds, magnetic microspherules, carbon spherules, and biomass burning proxies. So, which layer at Hall’s Cave were these impact proxies found in? The layer between 151 and 153 cm of course. The presence of multiple impact proxies at 153 cm demonstrates that the YDB at Hall’s Cave can be found at 153 cm depth; of course, this was only the case for the specific samples taken for the 2009 study, as it varies. Could this be the reason Sun et al. (2020) omitted this layer from their study? Perhaps if they found a large Pt spike at 153 cm, its inclusion may have discredited their interpretations and/or their conclusions. For obvious reasons, the presence of such a spike at 153 cm would have made it very difficult for the authors to claim the YDB occurs at 151 cm. Despite their brief discussion of the global Pt anomaly, demonstrating their awareness of its existence and its diagnostic utility for the YDB, the authors do not use the 151 cm spike as evidence for that being the YDB at Hall’s Cave. Doing so would have supported their case, as the Pt has been repeatedly replicated at the YDB, and using it as supporting evidence does not mean admitting it came from an impact event.

Because the study is ultimately devoted to testing the YDIH, it is unconscionable that the 153 cm sample was not included. Here are some of their conclusions that attempt to discredit the YDIH in favour of a volcanic trigger:

“The cause of the elevated HSE concentrations and the Os isotopic ratios in YD layer sediments remains equivocal and has been used to both support and negate the YD impact hypothesis. For example, Petaev et al. found a Pt enrichment accompanied with an extremely high Pt/Ir but Al-poor signature in the Greenland Ice Sheet Project 2 ice core at the Bølling-Ållerød/YD transition period, which they interpreted to be consistent with an ET impactor. Also, the elevated Pt abundance anomalies of 100 to 65,000 parts per trillion (ppt) at the onset of the YD in sites from North America is purportedly consistent with the Greenland ice core Pt data. Moore et al. found Pt and Pd/Pt anomalies in the YD basal layer in South Carolina.”

This quote demonstrates that the authors are aware of the widespread reproduction of the Pt anomaly at the YD onset, and that it has been proposed as a global geochemical datum marking the YDB. It also demonstrates they are aware that the YDB impactor appears to have novel geochemistry.

“The five unradiogenic Os peaks, including the YD layer, fall within a ~4000-year time interval. The unradiogenic ¹⁸⁷Os/¹⁸⁸Os ratio and HSE abundance data from Hall’s Cave sediments are inconsistent with the YD impact hypothesis. Alternatively, these levels contain cryptotephra and associated aerosols derived from large Plinian volcanic eruptions.”

As demonstrated above, their conclusion that there are five volcanic signatures in their Frankenstein sedimentary section from Hall’s Cave is unsupportable. Instead, there were three at most, and it is likely that none of them coincide with the YDB.

“The YD horizon correlates in time with the Laacher See eruption with a VEI of 6 and a 6.2-km³ eruptive volume that was dispersed throughout the Northern Hemisphere. Previously found YD markers, such as nanodiamonds and other wildfire products, are not necessarily solely impact-induced. Instead, these could originate from high-temperature, large-scale volcanic eruptions whose explosive conditions are capable of producing molten silica and carbon spherules and possibly nanodiamonds (Lonsdaleite).”

The authors cite van Hoesel et al. (2014) to support their claim that lonsdaleite can be formed during a volcanic explosion, and that other YDB impact proxies can have formed by other processes. That paper contains no evidence to support their claims, only conjecture. This is one of many examples of Sun et al. (2020) citing conjecture from other authors, with no supporting evidence, to support their own unfounded claims.

“These observations from the Hall’s Cave section also explain the lack of an Os isotope ET signature, or for the interpretation of a cryptotephra signature, at many YD locales across the Northern Hemisphere.”

This is simply not true; the authors ignore several studies that suggest an extraterrestrial Os isotope signature in the YDB layer at multiple sites, such as Beets et al. (2008) and Sharma et al. (2009). Prior to studies by Sun et al. (2020, 2021), only Paquay et al. (2009) had failed to locate an extraterrestrial osmium signature at the YDB.

“The results here have implications not only for the YD event but also other Pt and Ir enrichment events in Earth history and where other supposed bolide markers have been used to support impacts at those times.”

These authors are not satisfied with mysteriously omitting crucial data from their study that could potentially contradict their conclusions. They have the audacity to suggest that their results are so conclusive they should enable the calling into question of other previously established impact events in geological history! Fortunately, a former coauthor of this study who withdrew his name after submission due to the significant issues with the paper has reached out and provided substantial documentation including the original data from the study. The document package provided contains all stages of drafts, comments, and communications about the manuscript before and after the 153 cm sample was deleted and has been vetted to confirm their legitimacy; to say it looks very bad for the authors would be a significant understatement. Of course, the deleted sample contained by far the highest concentration of Pt in the entire record (1807 ppb), more than 4x the elevated Pt in the single 151 cm sample.

Attempts to contact the corresponding authors to ask them about the issues with the paper were ignored. A follow up email outlining our intent to raise the issue to the journal and the author’s institutions was met with indifference and hostility; one of them CC’d my academic supervisor in their response in an attempt to get me in trouble. Given the recent escalation in politically motivated retractions (Tankersley et al. 2022; Natawidjaja et al. 2024), it only seems fair that cases of actual scientific misconduct by the hostile faction should be brought to light. Therefore, requests for a formal inquiry and full retraction of Sun et al. (2020) will soon be launched with the author’s institutions, and the journal (Science Advances), and all the omitted data and documentary evidence will be made available to investigators.

Sun et al. 2021

These same authors published another similar paper from the Friedkin Clovis site in Texas (Sun et al. 2021). This paper has less obvious errors, but still contains some significant problems; they report PGE concentrations using duplicate samples, meaning two supposedly identical samples from each stratigraphic layer; while this method is fine, it did not go particularly well for them. The reader may remember from earlier how Bunch et al. (2009) highlighted reproducibility issues of up to 400% in Paquay et al. (2009)’s iridium concentrations; the average discrepancy between supposed identical samples from Sun et al. (2021) was 614%. Yes, that is correct; the average discrepancy between the duplicate samples used in Sun et al. (2021) to justify a volcanic origin for the YD event was almost 215% worse than the worst discrepancy from Paquay et al. (2009). The highest discrepancy between identical samples from the same layer was 19.7 ppt of osmium in one sample and 1985.8 ppt in the duplicate; that is a 10,078% discrepancy, making it 25 times more unreliable than the most unreliable result from Paquay et al. (2009).

The authors again invoke the ‘nugget effect’ to explain the significant discrepancies between purportedly identical samples and use the phenomenon to explain the significant discrepancies in their Hall’s Cave paper. However, the reader has just seen that the nugget effect is not responsible for those discrepancies; it is unequivocal that those discrepancies were caused by combining multiple sampling episodes, there is no denying that fact. The nugget effect is a real phenomenon, wherein platinum group elements and gold tend to clump together in sediment rather than distribute evenly like other elements. However, studies on recognising the nugget effect advise against invoking it to explain high variability between replicate samples without good evidence (Meisel et al. 2001):

“…unless it can be demonstrated that sample inhom*ogeneity is the major contributor to the high variability of an analytical result one should be careful not to mistakenly attribute this to a nugget effect.”

Sun et al. (2021) do not demonstrate this with any certainty; they use a single sample as evidence that the reproducibility issues at the Friedkin site were due to the nugget effect. All they demonstrated was that the nugget effect exists, not that it had any effect on the Friedkin samples. Furthermore, they specifically state that they avoided hom*ogenising their samples by way of grinding, as Paquay et al. (2009) did, to ‘preserve isotopic variability’. However, Meisel et al. (2001) state:

“…it is almost impossible to properly hom*ogenise samples, because the PGE-rich phases are either ductile or very hard. This effect challenges the analytical chemist when selecting representative samples by taking aliquots.”

However, short of analysing multiple replicates from the same sample and averaging the result, hom*ogenising a sample as much as possible is the only way to produce a result with interpretational utility. If each result from the same depth differs significantly, how can conclusions be made with any confidence? Which result is correct? What makes one result more correct than the other? How can one result be used to make a conclusion if another result from the same sample can be used to make a different conclusion? If conclusions cannot be made with any kind of certainty, there is no reason to even do the analysis. Furthermore, if the nugget effect results in such highly variable results, how can the authors claim their conclusions should be used to re-examine interpretations of past impacts as potentially volcanic?

Thus, their results are effectively useless for interpreting anything; their invocation of the nugget effect is an admission that their results are meaningless. This entire paper could even be a ham-fisted attempt to cover up their highly suspect omission of exculpatory data from Hall’s Cave. I have seen the documentation proving that the authors were made aware of the significant issues raised above during the publication process of the 2020 paper. Given these issues, and the significant issues in their 2021 paper outlined above, the conclusions of both studies should be taken with a grain of salt, if not entirely discarded.

Finally, the latest volcanism-based challenge to the YDIH by established critics was recently published (Montanari et al. 2024). They used the presence of volcanic tephra in a layer dating to 14,400 years ago to cast doubt on the YDIH. Confused? You should be – their paper is utterly nonsensical. Either the authors lack a basic understanding of soil formation, or they are attempting a quick and dirty hatchet job on the YDIH.

As shown below, (Figure 57), their date for the YDB (12,840 ± 122 BP) was taken from the very top of layer 2, a thick charcoal-rich layer, while the thin, orange-coloured volcanic tephra of layer 3 dates to 14,400 BP. Based on these 2 dates, it is clear that layer 2 formed over a period of ~1300 years, which is important to keep in mind. The layer 2 samples used for other analyses were bulk samples that span the entire 8 cm of layer 2, meaning they combined 1300 years of sedimentation into a single sample with a single date, and think this enables them to unravel an event that took place over a decade at most. The top of layer 2 is the most likely location of the YDB based on existing knowledge, and the date from there confirms it. Had they taken additional samples 1-2cm thick from the same level of layer 2 as the radiocarbon date to test for impact proxies, they may have been able to make some relevant conclusions about the YDIH, but they did not. Had they done so, there is no doubt they would have confirmed the presence of the platinum anomaly; despite the significant dilution of the thin YDB layer, layer 2 still contained the highest concentration of Pt in the record (2.2 ppb). Assuming the YDB is 2 cm thick, and the rest of layer 2 only has background levels of Pt, the 2.2 ppb would probably be more like 8-9 ppb, which is consistent with YDB Pt at many other sites. Clearly, rather than the entirety of layer 2 representing the YDB as they claim throughout, the YDB is marked by the drastic transition between the white carbonate later and the dark charcoal-rich layer. Thus, their study cannot serve as evidence either for or against the YDIH, despite a significant portion of the paper claiming that it can. Even if they did show that the YDB contained a volcanic signature, other recent studies show that the largest sulfur-rich volcanic eruption in the last 2.5 million years fell far short of causing millennial-scale cooling (Innes et al. 2024). This groundbreaking discovery adds another nail to the coffin of the volcanism hypothesis for YD cooling.

Radiocarbon Dating

Radiocarbon dating has been revolutionary for archaeology; it has become the benchmark and the gold standard for chronologizing sites and individual artefacts. So much so that without a radiocarbon or other similar radiometric date (or even the wrong kind of radiocarbon date (Natawidjaja et al. 2024)) any claims an archaeologist makes are often not taken seriously. Even with other radiometric dates such as optically stimulated luminescence, archaeologists will still find ways to disagree. Here is one example, in the context of the YDIH, that helps illustrate the biases of the more vocal critics: Holliday (2015) criticised the incorporation of a few Optically Stimulated Luminescence (OSL) dates, which typically have higher uncertainties than radiocarbon, into a Bayesian model that supports the YDIH. This is despite Holliday having previously dated the Odessa meteor crater using a similar hybrid model in which >70% of dates were OSL (Kennett et al. 2015b).

The paradigm of radiocarbon dating is based on two fundamental assumptions: 1. that the flux of ¹⁴C, a radioactive isotope of carbon, in the atmosphere is somewhat steady and 2. that the half-life, or decay rate, of this radioactive isotope is constant and unchanging. To summarise, as a living organism consumes carbon from the environment, incorporating it into their body, a small percentage of that carbon is radioactive ¹⁴C; when that organism dies, the half-life of ¹⁴C begins ticking like a clock (Figure 58; Capano et al. 2020). While the latter assumption is almost universally accepted as being true, we know the former is not always true. The atmospheric radiocarbon flux is dynamic, meaning radiocarbon dates need to be calibrated using a model that is updated and refined every few years, making older radiocarbon dates incompatible with newer ones without re-calibration (Engels et al. 2022). For example, at the YDB, there is a fluctuation in the ¹⁴C concentration that produces an anomaly of ~400 radiocarbon years in the space of ~100 calendar years (Fiedel 2011). These larger fluctuations that require correction can cause confusion when trying to compare dates published in different decades, particularly for the general public who may not realise this. Calibrated radiocarbon dates are usually reported in “BP”, or “before present”. What this actually means is “before 1950”, because around 1950 is when humans began significantly altering atmospheric radiocarbon concentrations by way of extensive nuclear testing. Uncalibrated dates are usually reported as radiocarbon years or ¹⁴C years.

Importantly, radiocarbon dates always have built-in uncertainties, often more than ±100 years, and this is the most important issue to consider in the context of the YDIH. When reading scientific literature, it is often easy to be fooled into thinking they are precise, both by how they are presented and how confidently they are used. However, the reported ‘date’ is usually the midpoint of the uncertainty. This means for a radiocarbon age of 34,034 ± 1,034 BP, the actual year the event took place could be any time between 33,000 and 35,068 BP. Furthermore, uncertainties are reported in confidence intervals (CI), often 95%; for the above example, if the uncertainty (±) of 1,034 years is at the 68% confidence interval, the 2,068 year ‘window’ becomes wider at higher confidence intervals. These inherent uncertainties can be further exacerbated by phenomena such as the old wood effect, which introduces a margin of error based on the age of the wood-producing plant when it died (Gavin 2001); a charred fragment of 1000-year-old Californian redwood can potentially produce a radiocarbon age up to ~1000 years older than when it died, depending on what portion of the tree was sampled. There are other similar effects that can influence the radiocarbon age of a sample, such as the hardwater effect, where calcium carbonate in a lake (fully depleted of ¹⁴C). This can interfere with the isotopic ratio, producing falsely old ages for plants and other organisms that consume it (Shotton 1972; Philippsen 2013). The consequence of these inherent uncertainties and confounding factors is that while radiocarbon dating can provide some level of support to archaeological interpretations of what probably was happening at the time, it is not suitable to justify claims that something did not happen. While critics scream from the rooftops demanding evidence of synchroneity (Jorgeson et al. 2020; Sweatman 2021; Holliday et al. 2023), the inherent uncertainty means they simply cannot be reliably used to disprove abrupt, short-lived events. The proposed date of the event falling within the dating uncertainty provides the strongest evidence possible in support of its synchroneity.

Radiocarbon & YDB Synchroneity

There are two main issues related to radiocarbon dating in the context of the YDIH: Age-depth models, and Summed Probability Distributions (SPDs), which were explored briefly in the megafauna section. Many topics discussed in the following section are also relevant to issues discussed in other sections, such as human population decline, megafaunal extinctions, platinum, and volcanism; they are instead discussed here specifically because they pertain to the issue of radiocarbon dating and synchroneity at the YDB. As discussed throughout, age-depth models developed using radiometric dating are subject to several confounding factors. For example, sediment accumulation rates are rarely constant; 10 cm of sediment may accumulate within a few years, while the next 10 cm may accumulate over 50 years, followed by 1 cm in 1,000 years, etc. To further complicate things, erosional events such as floods can remove layers of sediment that had previously built up, resulting in erosional discontinuities, or ‘time jumps’ equal to the sediment that was washed out. For example, if sediment deposited over a 1000-year period is eroded away, the interpretational utility of the section is decreased, as there is now a 1000-year gap in your record; such a gap may not even be noticed if the dating resolution is not sufficient. This can be further complicated if the erosional event selectively removes some material from the layer, but leaves others behind; erosion acts on smaller grains first, and can result in the entire fine portion of the 1000 years of sediment being removed, leaving behind the much larger material. This is one way that age reversals, such as in Gill et al. (2009) can come about. Finally, taphonomic processes such as reworking of sediments by various other means and bioturbation (e.g. organisms and plant roots) can produce anomalously old or young dates in a given layer of sediment. For example, older sediment being washed down a river and deposited in younger layers can cause younger layers to appear older, and burrowing creatures can bring younger sediments into older layers.

Unfortunately, for better or worse, age-depth models are still a staple of any archaeological or paleoenvironmental reconstruction, and absent a more precise and reliable method, they will continue to be so. Many of the early criticisms against the YDIH centered on casting doubts on the synchroneity of proxy evidence based on the uncertainty of radiocarbon dating (Boslough et al. 2012; van Hoesel et al. 2012, 2014; Meltzer et al. 2014). In response to these criticisms, a landmark study by Kennett et al. (2015a) sought to resolve, as best as possible, the issues surrounding radiocarbon dating of the YDB; they produced a robust (as possible) model demonstrating synchroneity between widely dispersed sites for the YDB using Bayesian modelling. Bayesian models have major advantages over other types of regression modelling for the following reasons, outlined by Kennett et al. (2015a):

• “Calculate and compare millions of possible age models (iterations), unlike regression algorithms that calculate only one.”

• “Integrate prior external information relevant to dating, e.g., the law of superposition (deepest is oldest).”

• “Identify outlying dates that are too young or too old [e.g., the old wood effect].”

• “Efficiently merge disparate data sets, e.g., from stratigraphy, archaeology, palynology, and climatology.”

• “Evaluate a cluster of dates for contemporaneity.”

• “Overcome some of the inherent biases of various dating methods that tend to favor some calendar dates over others.”

• “Present a robust statistical model that explicitly represents all modeling assumptions and data input.”

Kennett et al. (2015a) constructed their Bayesian model using 354 dates from 23 stratigraphic sections in 12 countries on four continents, in addition to the GISP2 platinum and six additional independent dates for the YD onset. Their combined model (Figure 57) found that the YDB, which contains impact proxies at 30 sites on four continents was deposited synchronously within the limits of radiocarbon uncertainty, between 12,835 and 12,735 BP at 95% confidence interval (CI). They modelled both the age of the YDB at each individual site and at all sites combined (Figure 59), finding that all models correlated to the YDB within the range of uncertainty, between 12,950 and 12,650 BP at 68% confidence.

Kennett et al. (2015a) make 6 conclusions they claim are supported by their results:

• “Bayesian analyses of 354 dates at 23 sites in 12 countries across four continents demonstrate that modelled YDB ages are consistent with the previously published range of 12,950-12,650 Cal B.P., contradicting claims that previous YDB age models are inaccurate.”

• “Bayesian analyses indicate that YDB dates could be synchronous within the limits of uncertainties (~100 y), contradicting claims that YDB dates are diachronous.”

• “Comparison with calibrated, unmodelled ages shows that Bayesian modeling does not significantly alter the calculated span of the YDB event.”

• “The ages of the 23 sites are coeval with the Younger Dryas onset in six records and with the age of deposition of extraterrestrial platinum in the GISP2 ice core at the Younger Dryas onset. This temporal relationship supports a causal connection between the impact event and the Younger Dryas.”

• “These analyses produced a more refined modelled age for the YDB event of 12,835-12,735 Cal B.P. at 95% CI.”

• “Although Bayesian analysis alone cannot determine unequivocally that the YDB is synchronous at these 23 sites, a single event is the most plausible conclusion, given the widespread presence of peaks in impact-related spherules, melt glass, nanodiamonds, and other markers that fall within a narrow temporal window of ~100 y.”

Despite limitations due to the inherent uncertainties built into radiocarbon dates, some groups have demonstrated a propensity towards taking them as gospel; they seek to use minor discrepancies between dates, within the margin of uncertainty, to disprove the synchroneity of the YDB. Remember, the YD onset is thought to have occurred over, at most, a few decades, which is almost always less than the radiocarbon dating uncertainty. In other words, the margin of error is larger than what it is trying to detect; usually when this happens in science, the result is discarded for unreliability. In the best-case scenario, radiocarbon dating can only be used to affirm positive evidence of an event occurring within a certain period; it cannot and should not be used with any certainty to provide negative evidence, or evidence that an event did not happen at a given time. For example, Breslawski et al. (2019) take issue with the fact that they were able to use other modeling software to move the YDB at a single site outside of Kennett et al. (2015a)’s combined model’s window. Not only that, but the dates used to justify this claim was published 4 years (Teller et al. 2019) after the study they are criticising, which used previously obtained dates that differ from the newer ones (Breslawski et al. 2019). Their two alternative models place the YDB at Lake Hind at least ~20 years and ~80 years outside of Kennett et al. (2015a)’s modelled YDB age (West et al. 2019). Breslawski et al. (2019) cite a study by Traschel and Telford (2017) to justify their claim that Teller et al. (2019)’s age-depth model is, in their view, likely inaccurate:

“Age-depth models can be fit with a variety of models and software packages that will yield different results”.

Ironically, they appear to have overlooked the title of their citation, which is “All age-depth models are wrong, but are getting better” (Traschel & Telford 2017); they also appear to have ignored earlier work by Telford et al. (2004) titled “All age-depth models are wrong: But how badly?”, wherein they reported discrepancies of up to 400 years between radiocarbon-based age-depth models and varve counting from the same section (West et al. 2019). If all models are wrong, what makes Breslawski et al. (2019) think that being 20-80 years off is relevant? The disparities between both models (60 years) is enough to dismiss their entire premise; if they had run the model 20 more times there would probably have been a reasonable number of them that placed Lake Hind within Kennett et al. (2015)’s window. Despite citing work that demonstrates the issues and uncertainties of age-depth models, the authors see fit to use their simple models built from a single site to discredit much larger and complex models built using hundreds of data points from sites all over the world. West et al. (2019) respond by stating:

“All age-depth models are simply probabilistic approximations that are likely to be inaccurate in most natural systems, given the fact that sedimentary deposition is variable and complex, and past rates are essentially unknowable with high certainty. All three Lake Hind models mentioned by Breslawski et al. differ from one another, and all are almost certainly inaccurate to some degree. Thus, differences of ~20-80 yr are statistically insignificant.”

They go on to address other fundamental assumptions regarding radiocarbon dating (emphasis added):

“Furthermore, Breslawski et al. assume an unwarranted precision and accuracy for the radiocarbon dates used to generate the age-depth models. While it is true that modern radiocarbon ages themselves are considered to have a high degree of certainty, that certainty applies only to the ages of the material dated but does not apply to the ages of deposition of the enclosing sediment. Dated in situ materials are easily moved around by bioturbation and/or reworked by erosion, frequently making the enclosing sediments appear either younger or older than they actually are.

Jorgeson et al. 2019

Another study critical of the YDIH based on radiocarbon simulations, by the same group of authors (Jorgeson et al. 2019), was published a few weeks after their response to the study at Lake Hind (Teller et al. 2019; Breslawski et al. 2019; West et al. 2019). The authors conducted a series of simulations with the ultimate goal of testing Kennett et al. (2015a)’s Bayesian analysis. I do not purport to be a statistician by any means, and so I am not in any position to counter their model with my own; instead, criticisms will be limited to other issues with their study that anyone capable of logic should be able to identify. There are some potential issues regarding their fundamental assumptions and the data being fed into their simulations that can be addressed. They begin as follows:

“While some events, such as volcanic eruptions, leave clear evidence for synchronous deposition, synchroneity is more difficult to establish for other types of events.”

This is based on the idea that volcanic eruptions deposit ‘tephra’ layers with geochemical signatures somewhat unique to that specific eruption, allowing them to act as extremely high-resolution chronological anchors. However, the reader will know by now that this also applies to the YDIH; rather than a geochemically distinct tephra layer, the YDB layer has a clear and well-established geochemical signature. Ignoring or downplaying this simple fact is the bread and butter of YDIH critics; they refuse to acknowledge the fact that spherule layers are used to correlate other impacts over widespread areas (Simonson & Glass 2004; Glass & Simonson 2012) in precisely the same way that volcanic tephra is used for eruptions.

On the one hand, they begin with the assumption that LSE tephra was deposited simultaneously at all sites, and so their LSE model is their benchmark for what a synchronous event should look like, to compare with their YDB simulations; this is of course a perfectly sensible assumption, and as they state, it is built into their assessment of their models:

“Since we assume that the LST was produced by a synchronous event, we conclude that simulation A1 does not fully capture the sources of uncertainty in the observed datasets”.

On the other, however, this luxury is not extended to the YDB, despite its obvious synchroneity based on the physical evidence of impact products and Pt geochemical datum. Thus, any interpretation of the differences between the models is already tainted by a very strong bias against the YDIH. The built-in assumptions are the most significant aspect of the models, as they are the lens through which interpretations of the models are made. The following is essentially what they assume: the LSE tephra was deposited simultaneously, therefore the results of model A, produced using LSE data, represent synchroneity. Thus, if the results of model B, produced using the YDB data, differ from model A, that must mean model B was not synchronous. The data used in each model is summarised below. In terms of the radiocarbon data used in their models, credit should be given to the authors, as for the most, part their treatment of the data is very fair; they wrote a small summary for each radiocarbon dataset they included, pointing out some inconsistencies between prior studies and demonstrating that they have thoroughly inspected Kennett et al. (2015a)’s data. They could certainly, if they had so wished, been more unfair in their data selection, but in almost every case they agreed with the inclusion of data in Kennett et al.’s model and their explanations for excluding lower-quality dates (Jorgeson et al. 2019, Supplementary Information). However, upon comparing the LSE and YDB datasets, some issues that could potentially affect their comparability become clearer.

All dates from the LSE tephra used in the model come from six sites within a ~30 km radius in Germany, while dates from the YDB are from 23 sites dispersed over more than 13,000 km. This simple fact on its own could account for much of the variation in the data; region-specific radiocarbon calibration curves for terrestrial material have been identified as a potential future development (Reimer 2021), made necessary because of differences in the regional and hemispheric distribution of atmospheric radiocarbon. Furthermore, several of the LSE dates excluded by Jorgeson et al. (2021) as unreliable show significant variation, some as young as 10,540 and as high as 11,370 BP (uncalibrated ¹⁴C years), with the majority of included dates being much closer to 11,000 BP. However, as they extended this fairness to Kennett et al. (2015a), allowing the exclusion of unreliable dates with similar variability that would have introduced further uncertainty, we can let this slide. Overall, Jorgeson et al. (2019)’s claims, presentation, and tone are mostly reasonable, at least much more than most other studies critical of the YDIH.

One of the more unreasonable criticisms of Kennett et al. (2015a) was that they rejected contradictory radiocarbon dates from their analysis ‘because they were contradictory’ (Boslough et al. 2015). However, in that case, they took issue with a single date of 207 ± 87 BP, produced from single carbon spherule; clearly this date cannot be related to the YDB, which was in direct contact with Clovis projectile points, which are not 200-300 years old. The likely explanation for such an anomalously young date is either bioturbation by plant roots, or contamination with modern material. Thus, the date was excluded on the basis that it cannot have been the date of the YDB, and its exclusion should not be even slightly controversial. Most importantly, the fact that only a single carbon spherule appears to have been dated by Boslough’s conspirators is suspicious, and if more spherules were in fact dated, why were the results never published anywhere?

Following Martin Sweatman’s (2021a) review of YDIH impact evidence, in which he points out several issues with their 2019 study, Jorgeson et al. (2021) responded to the criticisms levelled against them. Essentially, their response amounts to a game of semantics, claiming Sweatman misrepresented the claims made in their study; Sweatman’s response includes an excellent, though a touch hyperbolic, scenario that dissects Jorgeson et al. (2019; 2021)’s fundamental premise (Sweatman 2021b):

“…consider the flap of the wings of a butterfly, and the impact of a 10 mile-wide asteroid. Even if these two events occurred independently, say 13,000 years ago, it is not clear that they would result in a similar distribution of radiocarbon dates at their respective boundary layers. The butterfly’s wings might dislodge a seed, which could then be deposited and radiocarbon dated, but would otherwise not perturb the environment in any significant way. The asteroid impact, on the other hand, would alter the environment catastrophically, through a hierarchy of interlinked events and processes, many of which could lead to an increase in the distribution of radiocarbon dates relating to the event. Ancient forests might be felled, tsunamis, earthquakes and landslides might mix and redeposit soils, and old sources of carbon might be redistributed. Even if some of these catastrophic processes might be modelled, there will always remain some doubt about the suitability and completeness of such models. If even one of these catastrophic processes is not modelled adequately, the distribution of radiocarbon dates from the asteroid impact boundary layer is likely to show greater variance than those relating to the flap of the butterfly’s wings.”

Of course, in this scenario, the butterfly is analogous to the LSE, and the asteroid impact is analogous to the YD impact event. His allegory invokes a single large impact, but still applies to the ‘violent meteor shower’ scenario, as the repercussions are similar. Sweatman further states:

“Nevertheless, [Jorgeson et al. 2019] contend that their work brings the Younger Dryas impact hypothesis into doubt. Implicit in their view is their assumption that their modelling accounts adequately for all sources of uncertainty in the YDB radiocarbon measurements. Of course, this is unknown, and therefore their conclusion is not supported. An alternative explanation for the greater variance in the YDB radiocarbon data than they expect is that their modelling of uncertainty is inadequate.”

Dr. Sweatman (2021b) does a very good job of dismantling Jorgeson et al. (2019; 2021)’s entire arguments, which rely on the notion that it is possible to disprove something with radiocarbon dating. As stated earlier, radiocarbon dating is only able to establish positive evidence in support of a hypothesis or theory; it cannot be used to establish negative evidence, in an attempt to disprove a hypothesis. In the absolute best-case scenario for critics, it can be used to cast doubt on an idea, such as “the dating does not support the assertion that X people were occupying Y site at Z time”, but the ability to cast doubt on something does not mean that something did not happen.

Ultimately, Jorgeson et al. (2019) admit that synchroneity of the YDB is still possible and that their simulations merely suggesting it appears improbable:

“We agree that synchroneity is possible, but our simulations demonstrate that it is extremely improbable that a synchronous event could produce ¹⁴C measurements as dispersed as those in the YDB. The most parsimonious explanation for the large difference in clustering between the YDB simulation and YDB observations is that the observed measurements were deposited asynchronously over multiple years, rather than by a single event.”

If the reader has been paying attention throughout, particularly to the section outlining the impact scenario, non-synchronous deposition of the YDB impact proxies over more than a decade or even a few hundred years is valid within the model of coherent catastrophism; a series of widely dispersed local and regional-scale impacts over an extended period is quite likely based on the extent of the impact evidence. Again, the dates used for the LSE model to establish a baseline for synchroneity are all ~30 km apart and were deposited within minutes of each other, while YDB layers ~12,000 km apart could have been deposited as much as a decade apart; they were still part of the same (geologically) synchronous event, but may have taken place over days, weeks, months or years rather than seconds or minutes. On the note of geological synchroneity versus archaeological synchroneity, the conditions for geological synchroneity are entirely met; had the YDIH occurred before we evolved, it would be called the YDIT, as it would have been accepted without much pushback at all. In other words, it is mainly archaeologists and ideologically possessed associates who argue against the YDIH’s synchroneity.

Many of the early criticisms by archaeologists, such as claims there was virtually no effect on Paleoindian populations, or that they did not even notice the significant environmental changes occurring at the YD onset, are riddled with inconsistencies and ignorance. One such study (Holliday & Meltzer 2010) claims that there is no reason to attribute the transition from Clovis to post-Clovis technology to a catastrophic event; apparently there is no archaeological evidence for population decline or reorganisation at the YD onset. Their very first figure is supposed to reaffirm their statement but tells a different story (Figure 60). It should be immediately plain to anyone who even glances at this plot that some sort of anomalous event took place at the YDB; immediately following 12,800 years ago, the inherent uncertainties in the dates absolutely explode (Figure 60). Clearly, there should at least be room to argue that the massive inflation of radiocarbon uncertainties directly following the YD onset could itself be taken as positive evidence of a catastrophe at the YDB.

Summed Probability Distributions (SPDs)

Other criticisms of various components of the YDIH are based on Summed probability distributions (SPDs). SPDs have been used in their current form within archaeology for more than thirty years (Williams 2012), but similar analyses, those using radiocarbon data for something other than direct chronology, have been in use since at least 1969 (Carleton et al. 2020; Geyh 1969). They are most widely used to infer population dynamics through time, sea level dynamics, climate changes, and other palaeoenvironmental reconstructions. In archaeology, they have been used since the early 1970’s to infer population sizes and site occupation histories, becoming more widely used in the last 2 decades. Recent work by archaeo-statisticians has identified that SPDs have many issues, and those groups using them need to understand them at a high level to use them reliably and appropriately. This study (Carleton et al. 2021) pioneers what they claim is a more reliable method for producing SPDs, state the following:

“To address the issues of sample quality, several scholars have recently advocated carefully selecting radiocarbon dates based on the perceived reliability of the dates, quality of the dated material, and consideration of depositional context – together referred to as “chronometric hygiene.”

In addition to “chronometric hygiene”, sample size is an important factor:

“To overcome sampling issues, most scholars argue that large databases of dates are required, although little agreement exists regarding what constitutes a good sample size in this setting.”

A review of SPD methodology by Williams (2012) explores three key issues with how SPDs are used, recommending several baseline standards for future SPD analyses. Among them, Williams (2012) proposes a minimum of 500 radiocarbon dates should be used, with the sample size and the mean of standard deviations of the dates being reported. Probably, this would dismiss a significant fraction of SPDs ever conducted, as many only use a small dataset. Williams (2012) summarises the most important issues with SPDs:

1. “Intra-site sampling. At the site level, radiocarbon samples are usually selected to strategically frame a stratigraphic sequence and rarely constitute a representative sample of occupation events at a site.”

This refers to a common practice in science, where dates are often taken from strategic locations such as near stratigraphic transitions to provide the most information possible for the lowest cost, as opposed to from consistent closely spaced intervals throughout the whole section.

2. “Sample size. Results are known to be sensitive to sample size, but how many radiocarbon dates are needed for a robust and reproducible summed probability distribution?”

3. “Calibration effects. The radiocarbon calibration curve and calibration process are known to affect the form of summed probability distributions.”

4. “Taphonomic loss. Several studies have suggested that the statistical correction of summed probability curves is required to offset increasing loss of archaeological sites with age.”

5. “Comparison with other proxies. Although summed probability curves are widely used to reconstruct demographic trends or prehistoric occupation, issues (1)-(4) underscore the need for comparison with other archaeological proxies.

Now, armed with this knowledge, we can delve into the YDIH-specific criticisms based on SPDs. The first major objection to the YDIH came from archaeologists in 2008, who used SPDs to model population size around the YDB (Buchanan et al. 2008). They acknowledge early in their study that:

“Although the summed probability distribution method is capable of yielding important insights, it is not without shortcomings. One problem is that, although major peaks and troughs in a summed probability distribution can be reasonably interpreted in terms of demography, it is difficult to determine whether minor fluctuations are caused by changes in demography or reflect the “wiggles” in the curve used to calibrate the dates.”

Their SPD plot (Figure 61) shows a spike in population just prior to 13,000 BP, signifying the ‘arrival’ of the Clovis people to North America and a steady, but less rapid, rise in population between 13,000 and 12,800 BP. Then, just prior to 12,800 BP, population begins to decline and continues to do so for more than 100 years, leaving the population lower than at 13,000 BP. However, this decline is considered by the authors to be too small to be related to a global cataclysm, because similar troughs occur elsewhere throughout their plot, and they don’t see it as being consistent with a population bottleneck (Buchanan et al. 2008). They admit it is possible that the trough represents a population decline and that it therefore “…supports a weaker version of Firestone et al.’s hypothesis”. Of course, this fallacious claim is entirely subjective; Firestone et al. (2007) only claimed “…major adaptations and population declines among PaleoAmericans.”. There was certainly no claim along the lines of entire populations being decimated instantaneously, and the decline identified by Buchanan et al. (2008), taken as gospel, could certainly qualify as ‘major’. So what strawman conception of Firestone et al.’s hypothesis were these authors arguing against? As far as I am aware, this has never been clarified.

As this was the first major criticism of the YDIH, it attracted comments from five groups of authors, comprising well known American archaeologists indicating they were keen to explore the hypothesis, and YDB team members (Anderson et al. 2008; Culleton 2008; Jones 2008; Kennett et al. 2008; Kennett & West 2008).

As an aside, this was probably both a blessing and a curse; the over-excitement of archaeologists who lacked the proper training to properly test this bold new hypothesis may ultimately have led to its premature rejection. Certainly, in the case of Surovell et al. (2009), they dove into the project headfirst without a shred of research to inform their methods. At least most of the archaeologists who commented on Buchanan et al. (2008) stayed in their lane, contributing knowledge and data from their own field to the debate (Jones 2009; Anderson et al. 2011; Jones & Kennett 2012).

Anyway, back to the responses to Buchanan et al. (2008). Kennett et al. (2008) correctly highlight the following issues with the study:

• “Only ¹⁴C dates with measurement precisions <100 years, and preferably <60 years, should be used because larger error margins blur probability distributions; many dates had precisions from 200 years to >2000 years.”

• “Only bone dates processed with modern techniques [e.g., XAD or ultrafiltration] are valid because of catastrophic consequences of poor chemical preparation.”

• “Stratigraphic associations between radiocarbon dates and cultural residues need to be demonstrated; e.g., much of the purported pre-11,000 14C years evidence used is now discredited.”

• “Single-component sites do not have the same credibility as multiple-occupation sites.”

• “The potential for site discovery is not equal through time; destruction and preservation may vary by region and are determined by burial depth, depositional environment, ground water geochemistry, and site type.”

• “Cumulative probabilities of outdated and inaccurate radiocarbon dates from poorly defined archaeological contexts do not provide meaningful proxies of past human demographics.”

Most of their points are completely valid; the use of radiocarbon dates with uncertainties of more than a few hundred years are too coarse to appropriately resolve short-lived events such as a proposed cosmic impact. Though I would disagree that only specific types of dates should be included, even if they are the most precise in terms of low uncertainties.

Anderson et al. (2008) highlight that the radiocarbon database used by the authors to construct their SPD is incomplete; it ignores almost an entire region by including only a fraction of a potential 181 radiocarbon dates from the Southeastern USA. They note that in the Southeast, there is a systematic 250-300 year ‘gap’ in radiocarbon ages between 12,900 and 12,600 BP, consistent with the idea of a population decline, or at least prolonged site abandonment following the YD onset over large areas. They further highlight a distinct, well-established transition between lithic technologies at the YDB, and a major decline in the number of projectile points of each unique technocomplex. For example, 1993 Clovis points had been found throughout the southwest, but after they disappeared at the YD onset, only 947 non-Clovis full-fluted points had been found; this then ticked back up to 1717 unfluted points and 2594 Dalton points. A >50% reduction in the number of projectile points being deposited in the archaeological record, according to Anderson et al. (2008), is a more reliable indication of population decline and/or reorganisation than can be achieved using an SPD. Jones (2008) concurs that the archaeological record in California is consistent with a significant disruptive event at the YD onset but takes a slightly different approach. According to Jones, as of 2008, no radiocarbon dates from around the YD have ever been produced in California; they see the fact that the YDB layer is effectively absent from all Californian archaeological sites as being indicative of a “disruptive event” causing a “strong cultural unconformity” between the Pleistocene and the Holocene. Perhaps some sort of catastrophe could have been responsible for the widespread erasure of the archaeological record around this time? Kennett & West (2008) essentially just reiterate some information from Firestone et al. (2007) that contradicts the points made by Buchanan et al. (2008). Finally, Culleton (2008) offer several criticisms of the use of SPDs for a study of this type, some of which were also offered by Kennett et al. (2008):

• “…the nondescript summed-probability distribution is a corrupt demographic proxy. Their smooth curve is due to a low-precision ¹⁴C database (52% of dates have measurement errors greater than ± 100 ¹⁴Cyr, 25% ±200 ¹⁴Cyr), which spreads metaphorical “population” over several calibrated centuries, filling gaps and dampening variability.”

• “…a priori archaeological information in a Bayesian framework that could constrain these dates (e.g., stratigraphic relationships, diagnostic artifacts) are disregarded, and therefore Clovis dates contribute to Folsom population and vice versa.”

• “CalPal applies a smoothing algorithm to the summed-probability distribution which levels out several sharp peaks in the true distribution.”

• “The result is an insensitive, low-fidelity population proxy incapable of detecting demographic change. Testing predictions of prehistoric population change requires high-precision ¹⁴C dates, understood in their stratigraphic and cultural contexts, critically evaluated in a Bayesian model.”

Regarding the issue of incorporating additional context and criteria into radiocarbon-based modelling, Steele et al. 2010’s comments closely mirror the responses above:

• “More prosaically, we have also been reminded of the need to screen and process archaeological radiocarbon datasets using appropriate statistical and stratigraphic criteria, particularly if we are exploring a hypothesis of an abrupt and relatively short-lived extreme event”.

Building on their response to Buchanan et al. (2008), Anderson et al. (2011) published the results of an extensive population reconstruction using lithic artefact frequency as a proxy for population dynamics, discussed earlier. They found that multiple lines of evidence are consistent with at least regional population decline at the YDB. Anderson et al. (2011) used data from a country-wide database of 30,000 Paleoindian projectile points to demonstrate a major drop in the number of stone tools being produced after the YD onset (Figure 24). This drop in lithic production occurred simultaneously with the transition from Clovis to post-Clovis technologies (Figure 6). They also analysed 11 of the quarry sites where people were procuring their stone to produce Clovis points. Of the 11 Clovis quarries, only one has reasonable evidence of continued use directly after the YD onset; the people did not just change the way they were making tools, they abandoned most of the quarries where they sourced their materials (Figure 24). Anderson et al. (2011) also produced their own SPDs, but instead of just one SPD for the whole country like Buchanan et al. (2008), they produced plots for individual regions and compared them to others generated from regional, national, and even international data (Figure 62).

The authors concede that rather than a ‘major population bottleneck’, their multivariate evidence could be indicative of dramatic changes in settlement patterning due to population reorganisation. However, the SPDs clearly show that population effects were more significant in the Southeastern USA than the Northwest; this runs counter to expectations if the population decline was due to migrations to avoid the harsh glacial conditions of the YD, as such migrations are typically directed away from the poles to more tropical climates rather than towards polar ones. Furthermore, as discussed earlier, the most recent analyses (SPDs from Iberia and widespread Y-chromosome genetic bottlenecks) demonstrate significant effects on human populations at the YD onset (Karmin et al. 2015; de Pablo et al. 2019; Sepulveda et al. 2022). The totality of Anderson et al. (2011)’s evidence, including their 14 different SPDs, comprehensively refute the claims of Buchanan et al. (2008), which are based on a single SPD built using incomplete data, often with high uncertainties.

Regarding single SPDs based on incomplete data with high uncertainties, another SPD-based study by Boulanger et al. (2014) was recently used by the lead author to ‘debunk’ the YDIH on a Twitter thread (Boulanger 2022); Boulanger joined in the anti-YDIH circle jerk comprising all the usual suspects at the height of the furor over the Powell (2022a) premature rejection paper. While the study specifically concerns the “Overkill” hypothesis, and only mentions the YDIH to say they are ignoring it, this study is useful as a case study for how SPDs can go wrong. According to their SPD, Paleoindians did not arrive in the northeast until 12,850 BP (Figure 60), which contradicts most of the literature ever produced on both the arrival of the Clovis people and the megafaunal extinctions. Most notably, Boulanger et al. (2014)’s model replicates the initial arrival and post-arrival peak and trough shown in Buchanan et al. (2008), only steeper and offset in time. Interestingly, it shows that Paleoindian population plunges to its lowest post-arrival level directly following 12,800 BP, which coincides with the highest peak of megafaunal deaths (Figure 63). What this signifies is that the highest number of megafaunal remains being entered into the fossil record occurred directly following the YD onset, just as human occupation had reached its lowest point since initial Paleoindian colonization. Regardless, the near freefall decline of megafaunal populations when human populations are already at their lowest both human occupation and megafauna populations suggests a common cause. In other words, even though Boulanger et al. (2014)’s model is probably wrong, if it was correct it would still support the YDIH.

Generating an SPD, at least in Boulanger et al. (2014)’s case, seems to involve combining, or ‘binning’ and averaging all the individual radiocarbon ages from a particular site, which can be highly problematic and does not make sense archaeologically. In terms of the archaeological reality for any given site, this hom*ogenizes what would otherwise be interpreted as successive or continuous occupations throughout the whole stratigraphic section over hundreds or thousands of years into a single date, taken from the midpoint of the range! In other words, dates from two layers separated by hundreds of years would be mashed together to form a single date that is supposed to represent population dynamics. Not only that, but in the example below (Figure 62), somehow 12 radiocarbon dates with uncertainties ranging from 90 years to 275 years are transformed into a single date with an uncertainty of just 40 years; if someone were to look me in the eye and say with complete honesty that they do not find this concerning, that would be all the evidence I need to completely disregard everything that person had ever said. Not only that, but it would probably also lead me to actively disbelieve anything you tried to tell me. How can this method be the cornerstone of our archaeological understanding of human population dynamics over time? All I can say is that had better be an error, because if not, I am handing in my archaeologist’s whip and hat first thing tomorrow. As previously stated, the author does not purport to be a statistician, but this does not seem like it should be possible. How can such garbage science be seen as an appropriate method by any archaeologist? Any tradesperson understands that the smallest amount of error introduced at the start of a project can quickly snowball out of control.

As shown in their Supplementary Information, Boulanger et al. (2014) discarded a further 47 Northeastern radiocarbon dates associated with Paleoindians, often on questionable grounds; many of the dates they discarded were pre-13,000 BP, which clearly resulted in their model showing a much later arrival time than is now well established. At least one of the dates from the ‘Debert’ site, if not hom*ogenized with dates up to 1,000 years younger, would show that Paleoindians have been present there since 13,000 BP. Interestingly, while 12 dates from one site were hom*ogenized into a single date, 3 dates from the Hidden Creek site were hom*ogenized into two dates, for no apparent reason (Figure 64). Two of the dates from Hidden Creek returned near identical ages of 9150 ± 40 & 50 BP respectively and were combined into a single date with a ± 240 year uncertainty, while the third date was 1180 years older. Compare this to Debert, where there is 1,398 years between the youngest and oldest of the 12 dates with uncertainties ranging from 90 to 275 years are hom*ogenized into a single date with . The differences in how the data is pre-treated before being added to the model demands explanation. Notably, Boulanger et al. (2014)’s SPD for Paleoindians is constructed using only 25 composite radiocarbon dates constructed from 54 individual dates, and as detailed earlier, the minimum number of dates used in the construction should be at least 10 times this number.

Based on a cursory inspection of both their human population and megafauna population datasets, it seems entirely possible that the near freefall decline of the megafauna could have occurred simultaneously with that of the humans; it only does not because most of the uncertainties of dates used in the megafauna SPD are significantly larger than most of the human population ones. To accurately compare between two SPDs, ideally the uncertainties between the two datasets would be somewhat comparable. If this is not possible it is good practice to acknowledge the limitations of a study and the reasons why it may not be entirely accurate; not only so that people know to be cautions, but to demonstrate that the authors themselves understand the limitations of their conclusions. Even if Boulanger et al. (2014)’s SPD were to be taken as gospel, it still demonstrates a population collapse of almost 100% directly following the YD onset. The appearance of a ‘late arrival’ of Paleoindians to the Northeast could even be due to the catastrophe itself; its effects on the archaeological record are untold, and there is at least one suspected ground-zero airburst site in the region. Perhaps the absence of ‘Clovis’ technology, and the succession of ‘Clovis-like’ technology, could even result from the survivors of the event moving away from the area. Of course, that is all just speculation, and since their SPD is almost certainly wrong, unnecessary speculation at that.

Because of the significant issues outlined above, Boulanger et al. (2014)’s model and conclusions should be treated with extreme caution, if not dismissed outright. Furthermore, as they themselves admit, their study only pertains to one region; while this fact alone does not invalidate the study, it does mean it cannot serve to debunk the YDIH. After all, in the ‘violent meteor shower’ scenario, regional differences in the effects of the impact event are allowed, if not expected.

?

The State of the YDIH – Summary & Conclusions

The YDIH suffered from a premature rejection because the first empirical studies attempting to replicate several key impact proxies were unsuccessful for the reasons outlined in this article and multiple other review papers. These early studies were unable to properly replicate Firestone et al. (2007)’s results due to either significant modifications to protocols and methods, or because their sampling methodologies were inappropriate for the task. Why is there always some sort of suspect method like the Frankenstein samples used in not only one, but 3 studies that dealt major blows to the YDIH? Why is there always some incompetent suggestion thrown in randomly, like Surovell et al. (2009) only size sorting at 1 mm, or Nakagawa et al. (2021) using XRF to test for iridium anomalies? How can these be the people responsible for building the knowledge foundation of humanity? Why do I feel I have to decide between downright incompetence and malignant conspiracy?

The obsession of critics with the argument over synchroneity is unfounded, as synchroneity is impossible to determine with 100% proof due to the standard uncertainties of radiocarbon dating being larger than the window in which the YDB is expected to have been deposited. Furthermore, the statistical tools used to analyse the radiocarbon data are only as good as the data itself; in other words, garbage in, garbage out. This section has demonstrated that by its very nature, particularly near the YDB, radiocarbon dating is unfit for determining synchroneity with the required level of confidence. Instead, the most suitable method to establish YDB synchroneity is the global Pt signature, especially when used alongside the other YDB impact proxies, constituting an impact spherule layer. The critics’ obsession with synchroneity may even be one of the hot topics precisely because it is unable to provide the hard proof they demand, but it can sow just enough doubt among the uninformed to call it into question. Hanlon’s Razor states that we should not attribute to malice what can be explained by stupidity. While I would like to believe Hanlon’s Razor applies to the critics of the YDIH, particularly concerning radiocarbon dating and synchroneity, there are too many red flags. Remember, some of the main YDB critics are archaeologists, and it is their job to understand the minutiae of radiocarbon dating, yet they consistently misuse it nonetheless.

The only methods of subversion that critics have left are to seed doubt in the minds of the public and even their own colleagues. Of the approximately 96,000 words of Holliday et al. (2023)’s gish gallop, there are maybe half a dozen substantive arguments or good points, with most of the manuscript being devoted to attacking strawman arguments and trying desperately to associate the YDIH with other fringe ideas promoted by non-scientists. They even try to associate the CRG with lunatics like Michael Jaye, whose paper is the only one I refuse to include in the comprehensive YDIH bibliography, because I have never seen anymore so spectacularly wrong about so many things at once. Many of the arguments they masquerade as credible, such as the issue of synchroneity are simply red herrings that are literally unprovable to their standards; this is not only because the technology to prove them (eliminating radiocarbon uncertainties) does not exist, but also because they refuse to adopt our recommendations for a method that would only PROBABLY prove it, being the Pt geochemical datum. Any critic could go to any site containing the YDB right now, use appropriate sampling methods, obtain radiocarbon dates and test the sample for platinum. If the results show a high Pt concentration, and the uncertainty of the date is within the range of the YD onset, then and only then should further inferences be made about that site in terms of the YDIH.

Regardless, now that all the early instances of poor science from the first 5 or 6 years have been exposed and ridiculed, the YDIH has enjoyed a recent resurgence in support from the scientific community. Even with the supposed ‘comprehensive refutation’, this should mean the road should be somewhat clear for the YDIH, right? Well, unfortunately this is not yet the case. Over the last two years I have been involved in several research projects with the CRG, and we have been trying to publish our results, which include some of the most important evidence supporting the YDIH ever found. One paper, led by Chris Moore detailing shocked quartz and platinum from the YDB at 3 US sites, was just rejected for the second time over disagreements over the shocked quartz, despite comprehensive evidence gathered using more than half a dozen widely used methods. Its first rejection came about because the editor sent it to one of the YDIH’s top critics for their review, who rejected it on sight, providing no reasonable justification based on the evidence being presented in the paper, just citing the ‘comprehensive refutation’ paper. The second paper presents some airburst modeling we have been conducting to test how airbursts can produce shock metamorphism in quartz and other minerals, alongside additional YDB shocked quartz. While one reviewer was very fair, the other two were very hostile. One reviewer was incensed that we had not provided a full list of modeling parameters used in the project; they had clearly failed to even open the file containing our supplementary information, where they would have found the exact parameters they demanded. Both papers will now be submitted to ScienceOpen following the adoption of all relevant reviewer comments. This is a sad state of affairs for the peer review process, at least concerning the YDIH.

Some readers, mainly those who have been living under a rock, may find inferences to the ‘impact mafia’ somewhat hyperbolic, but I have seen it with my own eyes. It has gotten so bad that the CRG has joined forces with another group of European impact scientists to launch a new journal on the ScienceOpen platform called Airbursts & Cratering Impacts. Every submission is sent out to multiple reviewers who are willing to entertain the ideas in the paper, and great care is taken to ensure that no party has any role in handling their own submissions. Furthermore, the CRG is committed to at least submitting all papers to at least one other journal and incorporating all reasonable reviewer comments into the manuscript. Not only is the manuscript greatly improved by this, One major benefit of ScienceOpen is that people can submit their own peer reviews at any time and have them displayed alongside the paper, and the primary goal of the journal is to bypass the increasing censorship, to expose data to the light rather than have it killed in a dark room by faceless hacks, The final straw was the absolutely ridiculous justifications given for rejection of the 3-part Abu Hureyra study by members of the impact mafia. While having to rely on our own journal and taking the obligatory criticisms on the chin is obviously not an ideal situation, it is the only way to ensure that future work on the YDIH cannot be suppressed by bad actors, and a new generation of impact scientists can begin to bring this new paradigm home.

This article has demonstrated seemingly endless examples of the shockingly low quality of the scientific discourse surrounding the YDIH, but it has barely even touched on the non-scientific side. That is because the YDIH is a scientific theory, and speculations based on thin evidence by non-scientists as to the secondary and tertiary consequences of the cataclysm that go beyond the available peer reviewed evidence; the potential disappearance of a lost global civilisation is beyond the purview of the YDIH. Efforts by critics to simultaneously associate these speculations with the YDIH, and particularly with the CRG, are irrelevant to peer reviewed scientific evidence. I stop short of endorsing the idea of a lost global civilisation, but I will happily hear the arguments for and against it, and I refuse to reject the idea without a good reason; there is also some very intriguing lines of secondary evidence supporting such an idea. I will also vehemently defend the rights of those who investigate and argue for it to do so, as open-mindedness and open discussion are among the most important things for scientific progress… after funerals of course. So, has the YDIH been ‘comprehensively refuted’ as our most obsessed critics seem to believe? Should we all just stop doing science, give up and go home? The answer is a resounding no, and in fact, we are only just getting started.

Please stay tuned for a future addition to this series on the YDIH detailing the groundbreaking new shocked quartz evidence and our airburst modeling, after it is finally published.