Baillie: 540 AD climate event likely volcano, not cosmic; seven year glitch discovered in ice cores


The Tusk hates to see a good cosmic climate hypothesis die, but best it be at the hands of a catastrophist scientist and father of said theory. In a continuing demonstration of his intellectual integrity, true ring guru Mike Baillie has lowered the flag on the 540 AD event and recommended volcanoes as a better fit. It’s complicated as hell, but suffice to say that removing seven years from the annual layers of ice cores results in a match between known (but unidentified) volcanic eruptions and tree ring diminution.

There is already a good discussion thread with coauthor Jonny McAneney underway on a previous post. I hope the illuminating commentary there will move to this post so that comments will be available with the paper.

It has been evident for some time that a discrepancy has existed in the first millennium between evidence for volcanoes in Greenland (and now Antarctic) ice cores, when
compared with likely volcanic effects as witnessed by frost damage in American bristlecone pine trees; the offset being of the order of seven years with the ice dates being too old (Baillie, 2008). Here we have shown that remarkably consistent spacing between both the ice acidities and the frost rings allow additional documentation of this widespread offset. It has been possible to reconstruct how the ice cores from Dye3,25GRIP, NGRIP, NEEM, Law Dome and WDC06A are an integrated group, all offset, with only DML apparently retaining independence, and showing less of an effect.

~~From the conclusion

Baillie: 540 AD Climate Event Likely Caused by Volcano by George Howard

  • Steve Garcia

    Jonny –

    In time work like yours will be done to nail down the YD onset more precisely than exists at present. (Hahaha – you are invited to do it!)

    While van Hoesel did a sloppy job of it and quibbled about what she saw as about a 100-year discrepancy, some day someone will do a better job than she did. Was the YD onset at 12,800ya? Or 12,700ya? or 12,765ya?

    Perhaps the data to date it solidly exists already; perhaps not.

    Personally, I think that within a range of 100 years at a time that ancient is pretty darned good. With the proxies rarely able to time things as precisely as your paper did for those events ~530 AD and ~687 AD, much more exactness may not be possible. And with tree rings all but useless at 12,800ya, tree rings may be precluded as an aid. Ice cores are at this moment a bit under attack (though gently, IMHO), due to the wandering of gases up or down in the layers (…sorry, but the correct term escapes me at this moment…), so it is unclear how much help ice cores will be in narrowing it down further. Varves and perhaps alkenones will be able to help. Alkenones as I understand them are fairly new, so that is perhaps a window just now opening on timing events in the past. It IS quite clear that the YDB was a one-time dive in the climate and the granddaddy of them all, so every proxy should have the needed abrupt transition. it should be just a matter of time and the right people doing it.

    After all, it is not so long ago that climatologists and biologists were talking about the YDB being a 200-year onset, as opposed to the current thinking of substantially less than 100 years. As more work is done on fundamental cores of whatever kind, the YDB timing should be able to be nailed better.


  • Steve Garcia

    Jonny –

    In case it is something you’d missed, in Google books’ I ran across this September 2014 book “Volcanism, Impacts, and Mass Extinctions: Causes & Effects, edited by Gerta Keller and Andrew K.Carr” and on page 411 it has an article “Calendar-Year dating of the Greenland Ice Sheet Project (GISP2) ice core from the early sixth century using historical, ion, and particulate data”.

    They specifically address the resolution of the ice core data for the 536 AD climate event, with resolutions at the 3-year level. Your paper is done and published, but there may be info you’d like to compare to your assessments.

    I was able to access part of the paper, but not all, on Google books, but the authors do mention “dancing stars” in their abstract.

    The actual Abbott et al 2014 paper is at and is dated January 2014.

  • Jonny

    Hi Steve,

    I have a copy of that paper (though in pre-print form). This has been something we have been looking at, and indeed it was in the back of our minds while writing our paper. The problem with the work is that it uses the GISP2 core for analysis, which has its own chronological issues. I have mentioned it before, but there is a 14 metre section of trashed core (that is the core came up in ice cubes, not as a contiguous sample), which corresponds to about 70 years of missing chronology, right in the midst of the 6th century AD. In the GISP2 core, the gap is dated between AD 543.5 – 614.1 using their timescale. But the question of course is this, how do they know that the core continues on at a date of 543.5 if they have no layers to count above it? Hence using the ice core blind around this time is unwise.

    In light of our paper, it now becomes apparent that there is a second issue regarding the GISP2 core at this depth….it too is misdated, but rather than being too old, it appears to be too young. The reason I say this is that the AD 626 eruption, which produced an historic dust veil, hemispheric cooling, a frost ring at AD 627 etc, seems to be dated to 639.1-640.6 in GISP2. I say seems, since this is the largest sulphate spike in the 7th century in GISP2 and is closest to the 626 date. So it would seem that GISP2 in the early 7th century is actually 13-14 years too young.

    If we look at GISP2 below the 6th century gap, we see two eruptions 20 years apart dated to 507 and 528 (start dates in the ice). These have to be the same dates as sulphate peaks observed in the GICC05 constrained NEEM core at 495 and 515. Two reasons for this; one is that the limited space analysis fits (there are only two eruptions we know off in this century that are 20 years apart), and two, GISP2 AD 639 must be NEEM 619, which is AD 626, as explained above. To explain this point further. GISP2 above the gap is 13/14 years too young. So if we shift the GISP2 dates back 13/14 years below the gap, the two Sulphate signals become 493/4 and 514/5, which is the same dates as NEEM. But since we know from our paper that NEEM is ~7 years too old at this depth, meaning that GISP2 must be 6 or 7 years too young at this depth.

    But the important point about this then is that the 6th century gap in GISP2 is actually from AD 536.5 (or 537.5) to 601.1 or 601.1 (so lets say AD 536-601). SO GISP2 does not actually record the AD 536 or 540 eruptions.

    So what does this have to do with Abbott et al? Well, they use GISP2 core for their signal, and argue that GISP2 is 3 years too young in the early 6th century. Thus their dating is still out by a further 3 or 4 years. The knock on effect of this relates to a the next paper in that volume “D. H. Abbott, D. Breger, P. E. Biscaye, J.A. Barron R. A. Juhl P. McCafferty, “What caused terrestrial dust loading and climate downturns between 533 and 540 A.D.?” (George had slides from some presentations by Abbott on site, but they must have went up in flames)

    In that paper, the group appear to find cosmic markers in ice around their date of 536 (that is GISP2 – 3 years) and suggest an oceanic impact. But now knowing that they are still out by 3 or 4 years, this layer should have a true date of AD 532/533, which would fit in well with the historical reports of the commencement of dancing stars.

  • Steve Garcia

    Ah, yes, Jonny, I thought you’d have a bit to say. I am glad to ask a good question once in a while.

    Rectification is a proper exercise for any analysis that uses proxies. I see your efforts and Abbott’s as rectification.

  • Jonny

    New research published today in Nature shows that the ice cores are indeed too old in the 1st millennium as indicated by Mike Baillie and myself.

    Sigl et al. have used the two cosmic events at AD 774 and 993 which produced excess 14C in precisely dated global tree rings, and 10Be in ice cores, to constrain the NEEM core chronology, and confirms that the ice cores are too old by 7 years below AD 1000.

  • Steve Garcia

    Jonny –

    It’s excellent to see your work being supported so solidly. I wonder if the authors have any idea about your paper…

    The Abstract mentions the 6th century pandemic. Maybe I am forgetting something, but what vectors would have existed between eruptions and pandemics? I’ve got nothing on that one…

  • Jonny


    I dare say they would have been aware of our paper, since Gill Plunkett and John Pilcher are work colleagues of Mike Baillie, and Bo Vinther was one of our paper referees. And I know one or two others were aware of it, since we sent them copies of our drafts.

    The 6th century pandemic is the Justinian plague, which took hold in the early 540’s. It has recently been shown that this plague was linked to Y.Pestis, and so was the same as the 14th century Black Death. The link between eruptions and the pandemic is thought to be due to the climatic and environmental deterioration, resulting in food famine, creating a weakened population susceptible to contagion. Perhaps even a degree of increased population movement at the time could have spread the plague further. For example, richer communities would probably import food from further away, or communities may move closer to larger population centres to earn money for food.

    So the eruptions are not the vectors, but rather their climatic effects are the catalyst for the pandemic to take hold.

  • CevinQ

    Jonny, Steve,

    As far as Y. Pestis and the Plague of Justinian goes climate change is the causative factor.The effect it has on the main vector, the flea in its native environment, equatorial east Africa, is that atmospheric cooling has a causes a change in the flea.
    Y pestis lives in the gut of the flea, in its equatorial home, the feeds once and dies, not spreading the disease.
    But when atmospheric temps drop below a certain level, there is a physiological change in the flea, a blood clot forms at the entrance to the gut, preventing the flea from ingesting blood. The flea feeds but never really eats, so as the flea is essentially starving to death, it feeds on multiple hosts, while the blood clot keeps the blood out of the gut of the flea it doesn’t stop the passage of Y. Pestis to the host.
    There is also a roman factor involved, the Eastern empire started direct trade with equatorial east Africa in the early decades of the sixth century. For the first time goods and people from africa were carried directly to the empire. The spread of the plague can be traced directly to roman shipping and trade. The plague decimated roman Britain, while sparing the Angles and Saxons, who did not directly trade with the romans.
    The atmospheric cooling also figures heavily in the onset of the plague in the 14th cent. as it appears during a period of major cooling in northern Europe.
    And Jonny this statement, “The link between eruptions and the pandemic is thought to be due to the climatic and environmental deterioration, resulting in food famine, creating a weakened population susceptible to contagion. Perhaps even a degree of increased population movement at the time could have spread the plague further. For example, richer communities would probably import food from further away, or communities may move closer to larger population centres to earn money for food.”
    sums it up perfectly, especially the last sentence, as time get hard people move around and import more goods to make up for loss of production, thereby accelerating the spread of the disease.

    In the 14th century episode, there is evidence from british records, that there was another pandemic that just preceeded the Y.pestis outbreak. The population of Britain fell by 30-40%? between 1280 and 1320, so when the plague hit Britain it was an already depressed society, enhancing the soread of the disease.

    Jonny, are you guys aware of the lakebed mud core project at Clear Lake California, they planning on retrieving mud samples going back 200k years.
    Clear lake is the only fresh water lake that has survived without being glaciated or have dried up at some point.

  • Cevin Q

    Here is the link to an article about the Clear Lake coreing project.

  • Steve Garcia

    Jonny and CevinQ – Thanks for the info!

    The flea-cold-Y. pestis connection makes global warming sound like a good idea. Even though we have such good medical capacity now, the fleas and Y. pestis don’t know that, and if things turned cold enough it sounds like it could try to take off again someday if the climate goes colder.

    The Medieval Warm Period was ending just about the time you mention, CevinQ. And the Roman Warm Period (even warmer, from what I’ve learned) ended just about the time of the 540 plague. But coming DOWN from a warm period – that is not necessarily exactly COLD. It sounds more like it completely reversed, with no mild period to speak of in between. (Of course, our modern perspective on time then is skewed, and is not the way we look at the passage of time in our present. Now, a couple of decades are seen as a long period, but when we look at the past, we compress multiple centuries and think of that period as being short. It’s like looking at the YD and the 1300 years as being not very long, but it is about the same length of time as from Muhammad to now.)

    One question:

    East equatorial Africa has a hot climate along the coast, but very mild at higher elevations. In fact, Nairobi is said to have nearly the best climate in the entire world. Any “cold” there one would think is a relative thing – even if it was very cold in Europe, simply mild weather would be more likely there. What specific part did the Romans trade with? Do you happen to know?

    If it actually got COLD at all in East Equatorial Africa, one shivers with the idea of what it must have been like in Europe.

  • CevinQ

    The romans traded all the way down to Mozambique and Tanzania.
    I think the temps have to be an average, it does cool down at higher elevations but they don’t have the pronounced seasonality that Europe has.
    I think demographics also had a lot to due with the spread, of plague. Populations were not nearly as dense in Africa and there were fewer urban centers, so if some did catch it they would likely not make it to the next village before they died. This same sort of mechanism has been implicated in the early spread of SIV/HIV in tropical Africa during the first decades of the 20th century. Before the dutch and French and germans built railways into the interior the disease was trapped, same with ebola. It might wipe out whole districs it would eventually run its course, while after transportation was available an infected person was able to infect others before succumbing to such things.