RealClimate logo

The Younger Dryas comet-impact hypothesis: gem of an idea or fool’s gold?

There was a paper in Science last week that has gotten quite a bit of press. It reports further evidence in support of the idea that the Younger Dryas — a distinct period towards the end of the last ice age when the deglaciation in the Northern Hemisphere was interrupted for a period of about 1300 years — was caused by a barrage of comets hitting North America.

When the first papers on this came out last year, we expressed skepticism. We remain skeptical and our reasons remain unchanged. But we think it is worth saying a bit more on this, because the reporting on this issue has largely ignored just how big an idea this is, and therefore how much more work would need to be done before it could be taken very seriously.

For background, see the good article by Kenneth Chang in the New York Times, which, however, does not address our main concerns with the hypothesis.

The brief history is that in 2007, Firestone and others published an article in PNAS showing evidence of various materials that may be diagnostic of extraterrestrial origin (and hence an impact) in layers of sediment dating to 12.9 thousand years ago, just before the beginning of the Younger Dryas cold event. Now, in a Brevia piece in Science Kennett and others show further evidence: “abundant nanodiamonds in sediments dating to 12.9 ± 0.1 thousand calendar years before the present at multiple locations across North America.”

According to Richard Kerr’s news item that accompanies the article in Science, at least some experts are skeptical that Kennett and others have really found nanodiamonds, or that, even if they have, they are necessarily evidence of an impact. But we don’t claim enough expertise in nanodiamond detection or interpretation to have an opinion on this aspect, so let’s give them the benefit of the doubt. Suppose there really was an impact (or impacts) at the right time in the right place. We’d still be skeptical that this was a trigger for the Younger Dryas.

Among our reasons for skepticism (again, see our earlier post on this) there is a basic statistical problem. The problem is — and this context is missing from most if not all of the articles we’ve seen on this — that explaining the Younger Dryas in terms of an impact leaves all the other rapid climate change events (the so-called “Dansgaard-Oeschger events”) of the last glacial period unexplained.* One would have to either accept the conventional ideas for the causes of these events, or, alternatively, one would have to propose that there was an impact not only before the Younger Dryas, but before each of the earlier events.

We recognize that it isn’t entirely an either/or situation. Indeed, the suggestion appears to be that a cometary barrage causes various kinds of havoc, including the ice sheet collapse that led to ocean circulation change (the most well-evidenced proximal cause of rapid climate change). But the point is that if these events can happen as part of the inherent variability of the ocean-atmosphere-ice-sheet system, then there is no need to invoke the impact hypothesis in the first place. And indeed it would be virtually impossible to show it was other than mere chance that comet impacts occurred at the right time, especially given that it would still be necessary to show that the ice sheet would care about comets, which we also consider unlikely (see the good discussion — particularly Mauri Pelto’s comments — on this over at the Open Mind Blog). On the other hand, if abrupt climate changes don’t happen on their own — if they only happen due to extraterrestrial causes — then one would want to see evidence of impacts for at least a few more of them, not just one. That would be a truly exceptional paradigm-breaking discovery, going against just about everything we think we know about the system.

We emphasize that we are not saying “the Younger Dryas can’t have been caused by a comet, unless all the Dansgaard-Oeschger events were caused by comets”. We’re saying that we see no need to invoke such an hypothesis, so the level of proof required for this extraordinary idea will need to be extraordinarily strong. So far, it doesn’t appear that that is the case.

Think about it. If it turned out that rapid climate change events are caused by comets, it would imply the climate system is far more stable than we thought, that abrupt climate change events are not part of the inherent variability of climate during glacial periods. That would perhaps allay fears that we could be pushing the system towards an abrupt climate change in the future. On the other hand, it would also suggest that cometary impacts are far far more common than we thought. Now that would be news. Perhaps further research by Kennett, Firestone and others will indeed show that to be the case. We’re not, however, holding our breath.

*Not to mention that there is an event similar to the Younger Dryas at the end of at least one other glacial period, “termination III” (see e.g. Carlson et al., 2008).

252 Responses to “The Younger Dryas comet-impact hypothesis: gem of an idea or fool’s gold?”

  1. 101
    Jim Cross says:

    I think some climate scientists have a degree of investment in the DO explanation. It is something that has happened multiple times and doesn’t seem random like a comet. It also plays into the “tipping point” narrative – warming reaches a certain point and suddenly bad things happen.

    Keep in mind that the article has to do with the finding of nanodiamonds in the YD sediment layer. Black mats, megafauna extinction, and the YD may be able to be explained in ways other than invoking the comet hypothesis, but I’m not sure the nanodiamonds can be explained as easily, alhough there is some controversy over whether the kind of nanodiamonds found are what would be expected from a comet.

  2. 102
    mauri pelto says:

    Ike excellent post #80. David Benson your doggedness is searching for answers serves us all well. If we had an index for the degree to which a post stayed on topic, this would have to rank at the top, great stuff. I want to return to the idea of Lake Agassiz inducing the YD regardless of what would have released the meltwater. The proponent of this initial idea was Broecker. He is concerned by the recent information not finding an identifiable pathway for this water and given the volume of water needed it would leave evidence. Broecker misses the point that even an under ice release would leave plenty of evidence of its passage. . Pay attention to the last two paragraphs. This more cautionary viewpoint is the result of several detailed studies that are not finding anything convincing about a drainage of Agassiz at the start of the YD

    We are getting ahead of ourselves to think about the shattering impact on the ice sheet of an impact, when we have not been able to document the flood event itself. That is conjecture built on conjecture not a strong foundation. It is also worth noting that an ice dam in this case is not likely some narrow delicate thing. We have an ice sheet advancing up

  3. 103

    First impacts of celestial bodies sufficiently large to inflict hemispheric-scale damage are (thankfully) rare.

    ‘Rare’ doesn’t tell us anything. The question is rather how often do they occur. Now we know they occur fairly regularly on millennial time scales, since we have two pinned down at 12,900 and 4800 years ago.

    Those are just the two we have hard evidence for. Surely there must be more, but one every 5000 years or so sounds about right, and fits the evidence thus far.

  4. 104
    Hank Roberts says:

    > how often … fairly regularly … we have two pinned down …

    Wait — two data points suffice to determine … what?

  5. 105
    Hank Roberts says:

    > the removal of most large herbivores resulted in
    > increased deposition of organic matter …

    Interesting idea. Few people realize how incredibly efficient grazing animals are in wildland numbers — removing the large herbivores would be followed by an overgrowth of the tallgrass prairie, buildup of fuel and then episodes of increased wildfire. Just speculation.

  6. 106

    Wait — two data points suffice to determine … what?

    Hemispherically catastrophic impacts on millennial time scales, I thought I was fairly clear on that statement.

    Add that to catastrophic earthquakes on yearly time scales and global warming on decadal time scales, and global economic collapses on centurial time scales.

    Compare that to where we are on mitigating any of this.

  7. 107
    Ray Ladbury says:

    Thomas, I’m afraid the mechanism for the putative YD trigger seems a bit vague. It seems to me that they’ve had to make some rather tenuous arguments to explain the YD as resulting from impact. What is more, even if we take two events in 13 K years, that puts the mean periodicity at anywhere from 2000 to 15000 years.

    I 100% agree wrt the misplaced emphasis for mitigation efforts.

  8. 108

    It seems to me that they’ve had to make some rather tenuous arguments to explain the YD as resulting from impact.

    I wasn’t making that connection, but clearly there is one that must be more fully refined. The big problem here is that the melting ice sheet washed away much of the evidence, and they are working the peripheries of it.

    Just like the rising seas washed away whatever was happening on the Bahama plateau for 100 thousand years.

    Clearly a catastrophic impact occurred at 12,900 B.P. Clearly also a catastrophic impact occurred at 4800 B.P.

    The evidence for these things is irrefutable, what remains is getting people out there to find more of them, and determine most probable impact scenarios for them.

    Captcha : inevitable value

  9. 109
    Jim Cross says:

    Mauri Pelto

    One of your links has a abstract that says this:

    “During the life of Lake Agassiz, 4 of the 5 largest catastrophic outbursts occurred at 12.9, 11.7, 11.2, and 8.4 ka cal yrs BP, which released 9500, 9300, 5900, and 163,000 km3 of freshwater, respectively (0.30, 0.29, 0.19, and 5.2 Sv if released in one year). Because these freshwater additions to the oceans occurred near the start of the three largest cooling events during this period of deglaciation–the Younger Dryas, Preboreal Oscillation, and the “8.2 ka cal yr cold event”–they may have been the trigger for changes in thermohaline circulation.”

    If I read this right, the amount of fresh water released during the 8.2 ka event was over 16 times what was released prior to the YD. Yet the 8.2 ka event was much less severe and shorter than the YD. Does this pose a problem for trying to explain the YD solely with Lake Agassiz release, whatever may have caused it?

  10. 110
    Jeffrey Davis says:

    I have to agree with ccpo at 98. It reads like a repeat of the wrong end of the “CO2 leading/lagging indicator” argument.

  11. 111
    David B. Benson says:

    ike solem (80) — Broecker’s paper, linked above, states the delta methane is unique to Younger Dryas, not found otherwise in Antarctic ice cores. Quoting from the Firestone et al. PNAS paper, “As evidence for biomass burning, Mayewski et al. (42, 43) reported large ammonium and nitrate spikes in the Greenland GISP2 ice core at the onset of the YD. These GISP2 data are consistent with strong geochemical evidence in the GRIP ice core for massive biomass burning at the YD onset, especially a major ammonium spike, in association with peaks in nitrate, nitrite, formate, oxalate, and acetate (44).” This suggests some chemical removal?

    A couple of years ago Scientific American had a very good article on super-eruptions, mostly about Mt. Toba, 74,000 years ago. There was a bit about Long Valley super-eruption. Other than species with highly constrained ranges, super-eruptions do not cause extinctions. With regard to humans and Mt. Toba it was a near thing however, one creating a definite east-west genetic divide; see the book by Oppenheimer mentioned in an earlier comment.

    All — Do recall that Bison bison survived the large mammal extinction event. So it is distinctly odd that the American lion did not, especially as there is no evidence of humans hunting them. Again quoting from the Firestone et al. PNAS paper, “It is likely that some now-extinct animals survived in protected niches, only later to become extinct because of insufficient food resources, overhunting, climate change, disease, flooding, and other effects, all triggered or amplified by the YD event.”

    mauri pelto (102) — Thanks. One certainly prefers to have good evidence. There is what I take to be a deep submarine canyon connecting Lake Mickigan and Lake Huron on the northern Lake Michigan bathymetric graphic from NOAA’s NGDC. (There is also a couple of features which might be impact craters.) An indicator of massive flooding, if any, ought to be the depth and size of the sediment deposits at the foot of the St. Lawerence rapids just west of Montreal, although subsequent floods or glacial re-advances might well spoil the evidence.

    That only a few thousand years of weathering can remove evidence is shown by stages of

    where driving along the Clark Fork one cannot recognize what occurred there (at least I could not). Again west of Longview, the 13,000+ years of heavy rain together with sea level rise has left no sign of those Bretz floods, to my eye.

    In any case, I’ve now seen enough evidence to find the case of ET air explosions and impacts into the Laurentide Ice Sheet compelling. Whether — and the extent to which — this YDB event promoted proglacial Lake Agassiz rapid drainage (about 150 years, start to finish) remains open. Detractors, however, are going to have to find an alternate hypothesis which takes into account to coincidence of timing between YDB an the onset of Younger Dryas itself.

    [reCAPTCHA entones “is disturbing.”]

  12. 112
    SteveF says:

    Jim Cross says:

    If I read this right, the amount of fresh water released during the 8.2 ka event was over 16 times what was released prior to the YD. Yet the 8.2 ka event was much less severe and shorter than the YD. Does this pose a problem for trying to explain the YD solely with Lake Agassiz release, whatever may have caused it?

    An additional possibly relevant factor when comparing the 8.2ka event with the YD comes from work from Real Climate’s very own Stefan Rahmstorf:

    “Simulation of rapid glacial climate changes in a coupled climate model”

    The paper looks at the stability of various climate systems.

  13. 113
    tamino says:

    Re: #111 (David B. Benson)

    Detractors, however, are going to have to find an alternate hypothesis which takes into account to coincidence of timing between YDB an the onset of Younger Dryas itself.

    I won’t opine on whether or not the impact event was a cause (or partial cause) of the YD, but I will mention that one explanation for the (possible) timing coincidence is: coincidence. Statistically, it does happen — in fact, coincidences are inevitable.

    [Response: Yes. This is a key point I made in the post: “.. the point is that if these events can happen as part of the inherent variability of the ocean-atmosphere-ice-sheet system, then there is no need to invoke the impact hypothesis in the first place. And indeed it would be virtually impossible to show it was other than mere chance that comet impacts occurred at the right time ….” –eric]

  14. 114
    Hank Roberts says:

    Also interesting to look at what can destabilize a glacier, though all the current observations are on mountain sites, not continental glaciers. I don’t know how to compare a large bolide to an avalanche.

    Rock avalanches that travel onto glaciers and related developments, Karakoram Himalaya, Inner Asia

    “… Glacier movement increased sharply at and below the rock avalanche deposits and, within a few months, the glacier surged. A second surge occurred 2 years later. Major slope failures and debris flows were triggered beyond the ice margins, and ponded melt water led to small outburst floods. By 2005 the landslide material had been transported some 9 km, about one third of this distance in the surges. Eventually it was fully reworked to become less readily distinguishable from other heavy supraglacial debris. A large area of thickened ice persisted where the debris reduced ablation; a positive impact on mass balance equivalent to a 20% increase in annual accumulation. However, it occurred as a moving segment of the ablation zone. In 10 years it had, in effect, replaced the mass transferred in the surges. Data for the mid-1980s indicate the rock avalanches exceeded pre-existing supraglacial debris by roughly five times and, over a 30 year period, will equal almost 500 years of normal supraglacial transport to the glacier margins. Other impacts of the episode suggest a doubling of this contribution to denudation….”
    —-end excerpt—–

    Similar developments were observed at three other glaciers

    EOS also just recently had an article documenting aa change in frequency of mountain avalanches, and included one that destablilized a glacier.

    Whether a large bolide going boom overhead packs anything like the impact of many tons of local avalanche, I dunno.

  15. 115

    What is more, even if we take two events in 13 K years, that puts the mean periodicity at anywhere from 2000 to 15000 years.

    I’m sorry I didn’t address this comment earlier, but impacts aren’t periodic, nor are they even random, and they’re not even stochastic. Like global warming, the problem is nearly deterministic to first order. In theory these objects can be identified, cataloged and tracked nearly down to the individual, and lesser fluxes may be observationally verified on relatively short time scales. I think that lunar surface observations are helping here, and should be helpful in the future, as well as closer atmospheric observations and cataloging of fireballs.

  16. 116
    David B. Benson says:

    tamino (113) — Yes, that needed to be stated. What was a coincidence was that it appears that the Laurentide Ice Sheet was ‘surging’ off into the sea, calving rock-bearing iceburgs, rather like a small Heinrich event. Some have gone so far as to term Younger Dryas H0. In any case, this presumably somewhat preceeded YDB.

    But the short term effects of YDB included a massive release of black carbon, i.e., “soot”. Enough of that might well have encouraged accelerated melting of all the continental ice sheets. So possibly the dating is off and H0 was caused by the YDB explosion(s) and impact(s).

    Hank Roberts (114) — Thanks. The YDB events must have been far more destabilizing, except that the Laurentide Ice Sheet was quite flat. Still, your except is indicative of what might have occurred along the southern margins.

  17. 117
    William Astley says:

    In reply to Barton Paul Leveson’s comment #74:

    “William, the cores of planets are hot for two reasons: 1) Primordial heat, since planets form by accretion and the kinetic energy of the infalling material has to go somewhere. 2) Radioactive decay. For Earth, the planet we know best, those sources account very neatly for the known thermal structure of the planet, within the limits of present empirical knowledge. No need for exotic mechanisms.”

    I believe there is now consensus that it is not possible to explain the planetary heat balance anomaly with the radioactivity mechanism. At most ¼ of the current internal heat is due to gravitational collapse. The problem with the radioactive mechanism is an unrealistic amount of radioactive material is required which creates too much heat at early stages in the planet. (If the planet is too hot the core would not have solidified.)

    This article in the January, 2009 issue of Scientific American discusses the planets’ anomalously high internal temperatures. The article not one researcher hypotheses that dark matter could be the mechanism supplying the missing heat to the planets’ cores, however, experimentally detailed experiments have failed to directly dark matter by searching specifically for very small heat unbalance.

    Also, I find it hard to believe that an impact of a body would remove the dark matter from Uranus.

    …Jupiter, Saturn, Uranus and Neptune. If these massive worlds have gravitationally captured dark matter, then dark matter particles could smash into them—rare events but enough to heat up the gas giants and account for why the insides of these planets (and even Earth) seem hotter than known mechanisms can explain. It might also account for why Uranus seems anomalously cold—the planet is bizarrely tilted, perhaps because of a colossal impact, and Adler surmises that this collision might have knocked away most of the dark matter cloud that might typically have heated Uranus…

  18. 118
    Hank Roberts says:

    > impacts aren’t periodic, nor are they even random

    Well, starting with the number of near misses reported since people started watching, we ought to be seeing some numbers about likely frequency. Is the assumption that these objects are randomly distributed broadly across the solar system, and we’re just seeing those within limits of our immediate attention? or is there some argument made that the’re more common crossing Earth’s orbit? Or even that they’re more common in the plane of the ecliptic? If they’re perturbed out of the asteroid belt or Oort cloud, wouldn’t the perturbation throw them into any possible orbital plane?

    Well, off topic, pointer welcome if this is summed up elsewhere.

  19. 119
    jcbmack says:

    Tamino, Eric, Ray, it does not seem to be just coincidence, I draw the line there. The impact hypothesis needs to be further studied and ongoing data collection and exploration of the alternative hypotheses need to be looked at with greater emphasis, not written off as it could be a coincidence or that it might be statistically insignificant, this is not proper use of mathematics or science. Asteroids and comets can alter climate in and of themselves greatly for the long term, however, many points in the Earth’s history cannot be clearly demarcated to reveal clear lines of cause and effect. Look at 100 years ago at the huge explosion in Siberia; we still have hundreds of ideas and hypotheses, could have been comets or asteroids, both, tectonic plate shifts, or some other combination. Point is chalking this comet hypothesis or early working theory for the last glacial period borders on ludicrous without any solid data to support that assertion.

    I am currently scouring many papers of geologists, paleo-
    biologists,astronomers, Astrophysicists, paleontologists and many expert interviews from History channel,, PBS and PBS,com as well and I have to tell you the data is pretty compelling and the arguments well thought out.When I finish I will cite all my sources.

  20. 120
    jcbmack says:

    To be sure there has to be coincidence at times and it is important to see through insignificant occurrences and those that actually telling data, but this is very premature, such speculation about a very very feasible cause and effect relationship or as a supplemental occurrence of still great magnitude and importance.We are not, however arguing over the bible code.

  21. 121

    Hank wrote:

    “Few people realize how incredibly efficient grazing animals are in wildland numbers — removing the large herbivores would be followed by an overgrowth of the tallgrass prairie, buildup of fuel and then episodes of increased wildfire. Just speculation.”

    And not only the herbivores, apparently–it seems that such effects can ripple down all the way down from the top of the food chain:

  22. 122
    Ray Ladbury says:

    William Astley says: “Also, I find it hard to believe that an impact of a body would remove the dark matter from Uranus.”

    Wow, you’re a comedian, too!

    However, RE: core energy balance, a quick search finds:

    As with the climate, let’s try known physics before traipsing off into the unknown, shall we?

  23. 123

    we ought to be seeing some numbers about likely frequency.

    Actually it’s a fairly new science with observations only for 200 years, and only theoretical fluxes calculated. You have to remember that they are distributed in size as well as orbits, composition, etc, and a lot has to do with angle of entry, altitude of impact, etc. The fluxes of the larger objects is clearly underestimated, so for the last ten years or so we’ve been sitting around waiting for some data points, knowing full well we should be seeing them, wondering where they were, and thinking about oceanic fireballs and impacts, and finally fully realizing that possibly the flux calculations were wrong. Then someone hit on the idea of monitoring lunar impacts by their optical emission. My criteria for a data point was any large tracked fireballs with debris, visible cratering, etc. Finally there was on object that hit a mountainside in Europe somewhere, this Peruvian impact which finally restored our confidence in overall skill, and there was recently a large fireball in Canada with debris. That was a big boost for optical tracking, which is mostly done by amateurs. There is a long oral history of these things as well, so we need to prepare ourselves.

    The dragon will return to the forest, trust me. Rather than just waiting around for the statistics to reveal themselves, it’s much easier to track the larger ones. The statistics is just for the small and insignificant stuff, the regional, local and city wide catastrophes.

  24. 124
    PeterB says:

    I’m not sure how well vetted this site is:

    But I found the topic germain to this very interesting discussion.

  25. 125
    David B. Benson says:

    The site of Montreal underwater at the time of Yunger Dryas: “The sea lasted from about 13,000 years ago to about 10,000 years ago and was continuously shrinking during that time, since the rebounding continent was slowly rising above sea level. At its peak, the sea extended inland as far south as Lake Champlain and somewhat farther west than the site of Ottawa, Ontario.”


    and “After the Port Huron advance, the ice retreated above the Straits of Mackinac and exposed a lower outlet at Kirkfield, Ontario. This outlet lead into the Trent Valley from the Georgian Bay and into Lake Ontario. With this new, lower outlet, the waters of both Lake Michigan and Lake Huron were diverted from the Chicago outlet and into the Atlantic Ocean (Farrand, 1988). During this same interval the earliest lake formed in the Superior basin, Lake Keweenaw. This lake occupied about two-thirds of the basin.”


    so an established flow path from Lake Michigan-Huron to the North Atlantic appears to have existed at the time of Younger Dryas.

    If so, and assuing that a subglacial flow from Lake Agassiz across the current location of Lake Superior to the Straits of Mackinac area, I opine that little evidence for such might still remain.

  26. 126
    ike solem says:

    Re#111: “ike solem (80) — Broecker’s paper, linked above, states the delta methane is unique to Younger Dryas, not found otherwise in Antarctic ice cores.”

    Actually, the ice core records are synchronous, according to past research, with some variation in amount but not timing.

    Antarctic Ice Core Hints Abrupt Warming Some 12,500 Years Ago May Have Been Global (1998)

    James White, a paleo-climatologist at the University of Colorado at Boulder, said changes in stable isotope ratios — an indicator of past temperatures in the Taylor Dome ice core from Antarctica — are almost identical to changes seen in cores from Greenland’s GISP 2 core from the same period. “The ice cores from opposite ends of the earth can be accurately cross-dated using the large, rapid climate changes in the methane concentrations from the atmosphere that accompanied the warming,” White said.

    The atmospheric mixing time is very rapid (several hundred days?), and the lifetime of methane in the atmosphere is about one decade (modern conditions), so any increase in methane will be recorded globally, though gradients in amount can happen, which may indicate the difference between tropical and boreal increases or decreases. There is disagreement over the cause of the methane decreases and increases, but terrestrial ecosystems seem more likely.

    See also (pg 1336) for the methane ice core curves.

    Here is a study of the methane issue during four methane events in the ice core records from Greenland and Antarctica:

    …During these events, atmospheric methane concentrations increased by 200-300 ppb over time periods of 100-300 years, significantly more slowly than associated temperature and snow accumulation changes recorded in the ice core record…. We find no evidence for rapid, massive methane emissions that might be associated with large-scale decomposition of methane hydrates in sediments.

    In the case of the warming episodes – the lower tropical latitudes could be the primary drivers behind abrupt polar climate changes. Take a look at this study: (Science Jun 2008)

    “Greenland Ice Core Analysis Shows Drastic Climate Change Near End Of Last Ice Age”

    Both dramatic warming events were preceded by decreasing Greenland dust deposition, indicating higher tropical temperatures and significantly more rain falling on the deserts of Asia at the time, said White. The team believes the ancient tropical warming caused large, rapid atmospheric changes at the equator, the intensification of the Pacific monsoon, sea-ice loss in the north Atlantic Ocean and more atmospheric heat and moisture over Greenland and much of the rest of the Northern Hemisphere.

    “Here we propose a series of events beginning in the lower latitudes and leading to changes in the ocean and atmosphere that reveal for the first time the anatomy of abrupt climate change,” the authors wrote. White likened the abrupt shift in the Northern Hemisphere circulation pattern to shifts in the North American jet stream as it steers storms around the continent.

    That’s similar to the modern case – tropical clouds being a major uncertainty in climate models, for example, and warming at the poles driven mainly by heat transfer from the equator.

    The comet hypothesis doesn’t match any of that, being a cooling scenario. What it does look a lot like, however, is the “nuclear winter” scenario. For that, you’ll want to look to Rutgers & Alan Robock.

    Take a look at the nuclear winter scenarios, the most recent being here:

    Would that correspond to the effects of a rare swarm of comets air-bursting all across North America?

    I am starting to wonder if this is an archaeologically-inspired drama, with all the references to Clovis man and the overkill scenario. See Charles Mann, “1491”, Chapter 5 “Pleistocene Wars”. Basically, it now appears likely that humans were in the Americas as long as twenty or thirty thousand years ago, and in largish numbers, meaning that extinction of megafauna was likely a typically slow process (as elsewhere), not one driven by bands of recently arrived and overeager hunters in the classic Clovis overkill scenario proposed by C. Vince Haynes c. 1964. Note that the comet hypothesis goes along with the Clovis scenario, but avoids the need for rapid overkill.

    The place to look for climate change is things like ice core records and lake sediment cores – pollen types, isotope ratios, and so on. The human angle is obviously of great interest, as are the mammoths, but that’s not going to give you too much information about past climate change.

  27. 127
    Hank Roberts says:

    One thought, I recollect from about water under the Antarctic currently that subglacial water can definitely flow ‘uphill’ in situations there’s higher water/pressure somewhere and a solid confinement between rock and ice through which the water can escape. So I wouldn’t assume the usual “water flows only downhill” rule applies for sure when trying to figure out where drainage occurred.

  28. 128
    David B. Benson says:

    The main reason for a link to

    Feng Sheng Hu, Herbert E. Wright, Jr., Emi Ito, Kathryn Lease, “Climatic effects of glacial Lake Agassiz in the midwestern United States during the last deglaciation”

    is for the sketch map of proglacial Lake Agassiz II. That was one big lake!

    ike solem (126) — Thank you. I obviously was not sufficiently clear. According to Broecker’s paper no other such cold interval event shows a corresponding decrease in methane concentration. Another unique aspect of Younger Dryas, it seems.

  29. 129
    David B. Benson says:

    “The Lockhart Phase was terminated when the Kaministikwia route to Lake Superior was deglaciated, causing a rapid drop in lake level and the abandonment of the southern outlet (e.g., Fenton et al. 1983; Teller and
    Thorleifson 1983) (Fig. 1, outlet K).”


    David W. Leverington and James T. Teller, “Paleotopographic reconstructions of the eastern outlets of glacial Lake Agassiz”

    which is a web accessible pdf file to which I am not capable of posting a link.

    About Kaministikwia River, cf

    and its headwater is

    but other flow possiblities are suggested by this (long download time) geologic map of the area:

  30. 130
    David B. Benson says:


    appers to indicate flow to the Mississippi during the 12.9 ka — 11.6 ka period of interest, as well as the development of some flow to the east. This paper seems to ignore the possibilites of sudden flooding.

    Quite professionally done, as should be expected from a member of the Geological Survey of Canada.

    From just this paper, taken alone, one would find it difficult to support the idea of a sudden draining of Lake Agassiz initiating Younger Dryas. However, the connection between the latter draining out Hudson Bay and the 8.2 Kya seems to be quite well established. From that one can infer a causal connection for the earlier Younger Dryas initiation,

  31. 131
    David B. Benson says:

    “During the subsequent Moorhead Phase (about 11,000 to 10,100 14C yr B.P.), drainage was east to the Great Lakes region, via the Kaministikwia route to Lake Superior [7][8], and the lake became as large as about 190,000 km^2 and 20,000 km^3 [5].”



    This 2 page review paper is recommnded. I was earlier confused between the two dating systems, radiocarbon years and calendar years. Given the researchers and their propensity to write joint papers I can better appreciate the prior papers as well.

    The issue, as I see it, is how the lake water is to cross the continental divide just to the west of Dog Lake. Presumably the lake level was high enough, but there was a considerable amount of ice preventing flow to the east; see the map in this short paper. The ET impact hypothesis is that enough of this ice was seriously shattered, melted, vaporized or otherwise removed to initiate flow. To support this, one would need magnetic surveys of every one of the smallish lakes westerly of Thunder Bay. Then one would have to attempt to see if impact craters could be located and, ideally, dated.

    That no major flow channels seem to be present in this part of western Ontario is problematic. But I’ve been through there once on the Cnadain Transcontinental highway so I can assure you there is essentially no relief; it is flat country. So I am supposing that the flow channel(s) through the ice kept changing position; at that depp time the land would generally be sloping downhill to the north amd east. The assumption is that the flowing water kept eating away and the ice on its left bank and not so much into the bedrock.

  32. 132
    Chris G says:

    Just thinking.

    Predator and prey populations tend to achieve a sort of equilibrium. It could be that the equilibrium is near the sustainable harvest point of the prey. (Of course, there will be boom-bust cycles, but those could smooth out over time.) So, let’s imagine that the human and large prey animal populations have achieved this balance point about 13,000 years ago.

    Let’s carry on with the hypothesis that there was a large air-burst collision; something had to form the nanodiamonds. Further, let’s continue with the line that smaller animals would be better able to duck-and-cover than large animals. Under this line of reasoning, you suddenly have an imbalance between the smaller predators (man) and the large prey. It takes a generation or two to make cultural changes; so, the people are going to continue to hunt the large game and quite possibly drive them off into extinction where otherwise, without the people, they may have been able to recover.

    Now, if my understanding is correct, Clovis points are mostly spear tips. You can hunt large targets with spears; it is much harder to hit small targets with spears. So, if the large animals are becoming scarce, the people would have to shift to other technologies to survive.

    I can’t explain the survival of the bison; maybe it had to do with migratory patterns at the time.

  33. 133
    David B. Benson says:

    Chris G (132) — The Clovis culture peoples had bison for about 35% of their meat. With lots of bison and few people, I suspect that American lions were the primary predator of the bison.

    With regard to the mammoths (3% of meat), I seriously doubt that human predation had very much to do with the untimely extinction. Humans mostly hunted immature male mammoths; probably lions did as well. In any case this would not have changed the population dynamics of the mammoths by much.

    There is no evidence of the use of Clovis points after YDB. It could well be that none of those who knew how to make Clovis points survived YDB. The closest subsequent culture appears to be Fulsom point culture. The Fulom point is smaller, enough for bison but not for the (extinct) mammoths perhaps.

    But in addition, many smaller species of mammals and also birds also became extinct at the time of this event.

  34. 134
    Jim Eager says:

    Re David Benson’s links @125, coincidentally, this past summer I canoed much of one former outlet along parts of the French River, Trout Lake and the Mattawa River.

  35. 135
    David B. Benson says:

    Oops. With regard to perscentages, I seem to have misremembered. Here is a paragraph from the C. Vance Haynes, Jr., paper previously linked:

    “Of the 70 sites in SI Table 2, 56 (80%) have skeletal elements of the Rancholabrean megafauna directly underlying the YD black mat (Table 1). Approximately 38.6% have mammoth remains, 37.1% bison, 8.6% horse, 7.1% camel, 2.9% mastodon, and 8.6% other extinct-species remains, all on the Clovis-age surface, and only bison remains appear in the overlying YD black mat.”

    which actually dosn’t determine food proportions.

  36. 136
    jcbmack says:

    ike # 126, I like your post, but disagree with your conclusion a bit… The study of human populations at different time periods and correlating them with approximations of the climate along with other animal populations can be very useful. Archaeologists offer an important component of study to all sorts of things including possible climate and triggers/consequences thereof.

    I know what you are saying and can agree for the most part with your line of thinking, however, ice cores and such are not the only ways to get good inferences about climate long before the historical record. There is much dispute actually highlighted regarding the Pleistocene era in this month’s Scientific American (pop evolutionary psychology) and excellent articles on evolution. At any rate I enjoyed reading your post and your references; if you have more I would be happy to read them as well, but I think we are getting ahead of ourselves on this issue of YD, especially what limitations we have in regards to inferring climate changes that wiped out or changed life form bio mass and diversity.

  37. 137
    jcbmack says:

    Of course Ike my own view is influenced by my Biology training and the desire that more biological realities including paleontological be included in climate science in several applications. I am not saying that Biology or Archaeology is completely ignored, but rather that looking at more data from these areas may improve our understanding of paleo climate. There are pretty good dynamic vegetation models and certain other Biological facets being modeled as well. I think fossil data can be revealing even if incomplete and oftentimes the source of further questioning.

  38. 138
    Pt says:

    Ike, I’ve been following your articles with much interest. Do you have an explanation for why ammonia (in the ice record) will go up following a major forest burn event, or an impact for that matter?

  39. 139
    David B. Benson says:

    Spahni, R., et al., 2005.
    EPICA Dome C CH4 Data to 650KYrBP.
    IGBP PAGES/World Data Center for Paleoclimatology
    Data Contribution Series # 2005-078
    NOAA/NGDC Paleoclimatology Program, Boulder CO, USA.

    I checked, by eye, for an negative methane spike remotely similar to that of Younger Dryas, about -180 ppbv. The 8.2 kya event shows up with around -20 ppbv and no earlier excursions appear to be more than +-70 ppbv per sampling interval. Granted the resolution grows coarser the deeper the time, but I didn’t notice any other, roughly similar short negative spikes. In magnitude combined with short duration, Younger Dryas methane change appears to be a unique event in the last 650,000 years.

  40. 140
    Hank Roberts says:

    PT, if you look up “forest fire” and ammonia in Scholar, and poke around, you’ll find ammonia comes out into the air as one result of a lot of natural processes; forest fires would be one very rapid process putting a pulse of ammonia into the air. If you want the specific chemistry involved you’ll have to look deeper than I did!

  41. 141
    David B. Benson says:

    “Our proxies, combined with planktonic δ18O seawater and δ13C, confirm that routing of runoff from western Canada to the St. Lawrence River occurred at the start of the Younger Dryas, with an attendant increase in freshwater flux of 0.06 ± 0.02 Sverdrup (1 Sverdrup = 106 m3·s−1). This base discharge increase is sufficient to have reduced Atlantic meridional overturning circulation and caused the Younger Dryas cold interval.”

    from A.E. Carlson, P.U. Clark, B.A. Haley, G.P. Klinkhammer, K. Simmons, E.J. Brook and K.J. Meissner, Geochemical proxies of North American freshwater routing during the Younger Dryas cold event, Proceedings of the National Academy of Sciences 104 (2007), pp. 6556–6561:

    “… in records of atmospheric methane concentration mainly reflecting precipitation and temperature impacts on boreal and tropical wetlands …”

    from Anders E. Carlson, Why there was not a Younger Dryas-like event during the Penultimate Deglaciation, Quaternary Science Reviews, Volume 27, Issues 9-10, May 2008, Pages 882-887: (link in footnote of Eric’s main post).

  42. 142
    Jim Cross says:

    Regarding the black mats, there is an interesting article:

    “Strata above and below the black mat reflect drier conditions with lowered water tables, and buried features such as wells dug by humans or animals and dry spring conduits indicate fallen water tables and drier conditions. Fallen water tables commonly result from a relatively warm, dry climate, whereas black mats, as used here, may have accumulated under conditions of increased precipitation and/or colder climate when rainfall is more effective in recharging water tables because of reduced evapotranspiration.”

    Many posts here and the main article arguing for the comet itself ascribes the black mats to wildfires, presumably caused by a large comet explosion over N. America.

  43. 143
    Hank Roberts says:

    Anyone tested for DNA in the black mats?

  44. 144
    Jim Cross says:

    Adding some additional thoughts to my last post.

    To believe that the YD was caused by something like a DO event not caused by a comet, one has to accept the following scenario occurring within a few years or decades:

    1- Megafauna and clovis people are alive

    2- Something deposits a layer of magnetic microspherules and other possible ET evidence.

    3- The DO event occurs unrelated to the ET effect.

    4- The climate grows dramatically colder and wetter. (By the way, do any of the other DO events show a black mat possibly caused by a colder and wetter transition?).

    5- Megafauna and clovis people are gone.

    6- We begin 1300 years of relatively cold climate.

  45. 145
    Jim Cross says:

    Regarding the CH4 oddities with the YD.

    Much of the methane contribution to the atmosphere comes from the peat bogs of Siberia.

    There is some evidence that these peat bogs expanded greatly at the transition to the Holocene from the YD.

    Whatever happened may have caused a massive die-off and suppression of CH4 production at the beginning of the YD. The die-off might have caused an expansion of the peat bogs. This would have been followed by jump start at the transition to the Holocene possibly caused by the insolation increase which was ongoing (as long as we are looking for tipping points).

  46. 146
    jcbmack says:

    NH4 would also be given off from burning plant life and bacteria. Environmental Chemistry should clear up some basic issues; they are available on scholar.

  47. 147
    mauri pelto says:

    #141 In the Carlson paper two additional key points are 1. They note several times that there was a progression of increasing values from the start of the yd to 12,500 and a bigger change at this point. This does not argue for a catastrophic release, but the opening of a new outlet that continues to expand its drainage area. Second there Figure 1 map provides ice marginal boundaries that do not fit the best data we have which shows ice advance across Lake Superior overrunning several forests in northern Michigan post 12.5. This map from Dyke (2004), which is based on earlier work as it is just a compendium map allows for the Lake Agassiz drainage but as the other papers that postdate Dyke indicate this drainage way is not evident. The increased flow through the St. Lawrence at the start of the YD could also be related instead to Lake Warren-Lake Whittlesey changes as well.

  48. 148
    David B. Benson says:

    Jim Cross (144) — Whatever Younger Dryas cause, it was not a DO event. DO events are warming events, not cooling. One of the earlier comments includes a link to a paper regarding THC modes which I found quite helpful.

    mauri pelto (147) — Thank you. As best as I can make out Younger Dryas initiation requireed about 150 years to full development. That suggests a more gradual release of proglacial lake Agassiz from Lockahart stage to early Morehead stage.

    But indeed, there is no reason for the fresh water to have all (or any) originated from Lake Agassiz (except that the flow had to go somewhere). Indeed the freshening could have been enhanced by other releases.

  49. 149
    Jim says:


    Check posts #14 and #8 with Eric’s answers regarding warming and cooling.

  50. 150
    David B. Benson says:

    The introduction linked below gives a general overview of the retreat of LIS from the Great Lakes, but the details may no longer agree with evidence developed later. In any case, here is a quotation: “Figure 9: Lake Algonquin stage was initiated when the Trent valley outlet was dammed by a local readvance. (about 11,000 years ago)” wherein the figure clearly indicates flow to the Strait of St. Lawerence from all the Great Lakes.

    Even if Lake Superior was solid ice, it seems to me that all that is necessary to initiate eatward flow from proglacial Lake Agassiz is for the water to move over the continental divide a few tens kilometers west of Thunder Bay. Subglacial and ordinary flow does the rest.