130 thoughts on “Millennium Volcanoes

  1. Wonderful food for thought and, I guess, a lot of research projects indicated! Thank you, Albert!

  2. … and the syncronicity monkey sneaks up and whacks me on the back of the head. Last night, while trying to go to sleep, my wife had her little tablet PC out and was asking me questions about news tidbits she was reading. The one that caught my attention was bubonic plague. It seems that a teen in Laramie County Wyoming had died from it recently. For those that don’t know, Yersinia pestis is a flea borne disease (mainly) and was probably the cause of the Jusinian Plague and later the Black Death episodes in human history. Some of the events are also possibly linked to weather changes from volcanic events as the population of the carrier animals grew and fell from food availability. (grain crops).

    • Justinian’s name got attached the Justinian Plague because he was running Byzantium at the time. It interfered with his goal of putting Rome back together and in part, the economic activity of Byzantium helped spread the disease as the carriers (rats) stowed away in cargo.

      • It is well known that the Justician plague (black plague) was helped by the famines/crop loses, caused by the 536 eruption that triggered worldwide climate disruption, possibly significantly more than Tambora with its year without a summer.

        The Spanish flu, another deadly pandemic, was also probably helped by WWI and its consequences.

        These two events seem to show that a fragile population is linked to the start of a pandemic (just like it happens at an individual level, when I sleep less, eat badly, catch some cold, and then get a cold or a flu)

        However diseases (and pandemics) can spread and kill possibly without a causal need for a weakening of a population first. In most cases pandemic just happen.

        The Black plague in 1348 was one such case. Despite that it came at a time of increasing cold climate (the great famine of 1311-1318 is a reminder) and also the 100 year war which had just started, I am inclined to say that it just happen.

        • And that plays right into the idea of multiple coincidental catastrophes that a population could normally handle singly, becoming far too close to an extinction level event that wipes out a species when they coincide.

          It could be argued that the idea applies equally well to companies and governments. When the Roman Empire “Fell,” no one really noticed it for a while. Even so, it was “too big to fail” but it did a pretty good job of doing a face-plant.

  3. Thank you Albert !

    The DRE of Samalas/Rinjani was 80km3?!?

    That is a truly mind blowing figure, one can only wonder what will happen to our civilization when we get another of this level…

    GL Edit: Added the superscript tags to your “3”.

    And what will happen? Easy, someone will figure a way to tax the survivors.

  4. Awesome article on one of my favorite topics. I’m going to repost this old graphic that Lurking posted since it’s one of my favorite topics of rumination.

    Provided, this is only a Greenland Ice core, so it’s not a comprehensive show of volcanic signatures since it’s only from one pole. That being said, the massive volcanic so2 spikes around 9000-10,000 B.C are mind boggling, and we have no clue where these came from. I have serious doubts these came from Iceland for various reasons, and the fact that they dwarf other enormous eruptions is crazy to me.

    A little while back, I had thought I found one of the primary culprits, but it turns out I was wrong by approximately 2000 years (although I did write up an article that was mostly ready to go before realizing my culprit was wrong).

    • Oh I think it could be Iceland.

      Remember volcanologist know very well that Iceland had massive volcanism around 9000 BC, just after the glaciation, because of the sudden relief of the ice cap over Iceland.

      Some big confirmed VEI5 include Askja, Grimsvotn, Katla (Vedde Ash), the birth of Hekla, possibly and a few dead zone eruptions. Possibly also Krafla and Torfajokull.

      I can share you one thing.
      Almost as fun as looking at tephra layers in ice cores…
      is looking at tephra layers in the Icelandic soil!

      I did even a study some years ago here
      Look at it: https://volcanocafe.wordpress.com/2012/08/01/the-ash-history-of-iceland-in-my-backyard-part-ii/
      It’s a cool read.

      I remember seeing 1 or 2 big thick ash layers when I was reaching deeper in a soil profile towards the start of the Holocene. I knew those two thickest ash layers were probably the largest eruptions in Iceland in the past 10.000 years (probably located dated around 8000 BC) but I am just an amateur so take my study with a grain of salt.

      • Here is my issue with Iceland being the culprit (I touch on this in the lower post). I don’t have any “hard” arguments, but there are a few things that make me doubt it as the primary suspect.

        Laki, an extremely large fissure eruption in Iceland didn’t even produce 1/10th of the So2 that the 9200 B.C. spike produced, and that’s keeping in mind the thought that Laki was extremely so2 rich even for Iceland standards.

        Similarly, we know that the Thjorsa lava flow is the largest Holocene eruption in Iceland, and it occurred too early, at around 6600 B.C to be the culprit. Theistareykjabunga, despite close timing proximity, is also an unlikely culprit, for reasons similar to why Thjorsa wouldn’t be a good target – they simply aren’t large enough. Laki, at 15km^2 was large, but produced 1/10th of the So2 necessary for the spike. It wouldn’t make sense that the original Theistareykjabunga eruption, which was at most, around 25-30km^2 could somehow create 10x the gas as Laki.

        The final argument for iceland is that there were many shield volcanoes erupting post glaciation. It’s true that there was greatly heightened volcanism going on in Iceland at this time period, but it still is unlikely that
        these lava eruptions were all occurring contemporaneously. Remember, So2 doesn’t last very long in the atmosphere, so even if there were a few large eruptions over the course of say – 10 years, they would all register separately on the So2 record, and the so2 data wouldn’t compound unless they occurred at exactly the same time.

        So to sum things up – going off current gas contents of more recent eruptions, it would seem unlikely Icelandic volcanoes were the culprit of the 9200 B.C. spike as there simply wasn’t any eruption large enough, even with their heightened gas content. You would need either a truly anomalous volcano in terms of so2 levels based off the eruption sizes to produce the spike seen in 9200 B.C. That’s not entirely out of the realm of possibility, but I certainly have my doubts about this.

        • I have a better hypothesis for you then.

          About 8900 BC +- 100 years a comet/asteroid impacted/air bursted over North America and set great areas on fire, setting the early Holocene mass extinction and the young dryas 1000 years return to glaciation.

          Anyways, the dating of the Holocene extinction around 8900 BC is remarkably close to your 9200 BC SO2 spike, and also linked to a sudden dramatic climate shift.

          • It’s an interesting hypothesis, although the so2 is specifically volcanic so2 I believe. I don’t think an impact event would create such signatures in the ice core, but who knows.

            • Wasn’t the Silverthrone area active around that time? That whole region was much more glacially active at that time (and in the time since) and alpine glaciers and ice fields do a heck-of-a-job removing/moving evidence of volcanic activities, especially in (still) remote areas like this. Another place to look would be along the Aleutian Islands; maybe a once-island-now-a-seamount could be the culprit.

        • And now to correct a few things here.
          Thjorsá was just the last of the great eruptions of Iceland, and it was far from the largest eruption.
          The largest known was Theistareykjarbunga that erupted 35 +15 cubic kilometers (shield volcano + hraun). Ad to that the Odadhahraun (Askja) at 30-35 cubic kilometers, another 15 at Krafla, 10 to 15 at Fremrinamúr. Then we should not forget the caldera formation of Grimsvötn and an assortment of other known Icelandic eruptions. All of this happened within a couple of hundred years as the ice layer melted away rapidly.

          Iceland was definitly the main culprit. That being said, the same most likely happened in Alaska, the Aleutian chain and Kamchatka. Judging from combined eruptive current pace I would say that 70/30 to 80/20 was from Iceland.

          Or in other words… Pretty much every single active volcano in the region went “boom” at the same time. And with most of them (and the largest) Iceland is really the culprit with the rest just being supporting cast.

          • I can see Hekla being the 4200 bc culprit, and Grimsvotn the 8200 bc culprit based on their gas content, and the fact that these were explosive eruptions that sent gas high up into the atmosphere. I’m still skeptical of Theistareykjarbunga as the source of the 9300 bc spike.

            Going off the records, it would need to be over 3x as So2 rich as Laki to be the primary culprit. Not out of the realm of possibility, but to me, it still seems fairly unlikely.

            “Or in other words… Pretty much every single active volcano in the region went “boom” at the same time.”

            Geologically speaking, the same time may be true. But as I said (and as you know) so2 doesn’t last long in the atmosphere, no longer than 2 years at most. Looking back in the geological record, many of these eruptions may look like they occurred at the same time, but in reality, they were most likely just happening with a higher overall frequency. Sure, some may have occurred at the same time, but even a 5 year gap between eruptions would eliminate the effects of compounding so2.

            I guess, I’m just not fully sold on Iceland being the culprit, but who knows. I feel like Kamchatka and Alaska are just as strong (or stronger) candidates.

    • That was one of my favorite plots. Thank you for remembering it! 😀

      I did one later that ran an antarctic SO2 plot along with the Greenland one. It was very confusing. Stuff that was known to be Southern latitude showed up in Greenland, and stuff known to be Northern Latitude showed up in Antarctica. I attribute it to weather since any other explanation makes me question my sanity as an aging primate somewhere in Florida.

    • No! It is surprising that even so recently, major eruptions could go unnoticed. It may have happened in a sparsely explored region, or it may have been hidden by another event (such as a hurricane). Long Island (PNG) is always a possibility: it had a very large eruption, believed to have been before 1700 because of a report on what the island looked like at that time, but perhaps that has been misinterpreted? I could not find the original report, just a second-hand comment about it.

      • Yes, I dont find it surprising too.
        After all in 2011, the Eritrea/Nabro eruption was found by satellite, wasn’t it?

        And now I remember: this eruption was high in SO2 and gas travelled NW towards Europe… Hmmm, maybe the same location for the 1831 event?

  5. Thank you Albert…Some real food for thought here and lots to keep an eye on! Great reading and certainly it puts a whole new slant on Volcano watching!

  6. This is really a valuable post Albert. A tasty treat for anyone who enjoys volcanism.

    Outstanding to see how good are Icelandic volcanoes are depositing sulphate, but also they are much much nearer Greenland, as you well said.

    Laki 4.6 compared to Tambora 1 (but Tambora deposited in both hemispheres, and being so far away it gives an idea how large in sulphate could have been compared to Laki, but probably not by much), Katla 1918 0.1

    I am sure Bardarbunga 2014 would also give a strong signal in this regard, since it was to gaseous. All rifting eruptions in Iceland are, as magma comes from very deep.

    Also interesting is Hekla 1766 0.35, way more than Hekla 1947 and Katla 1918. That one was well known to be a strong eruption. But it is still 10 times less than Laki, which is already a big number! Of course Hekla 1104 was 1.25 and even a remote chance that possibly deposited in the south hemisphere too, as the remark indicates, but I don’t think so. This shows that Hekla releases much more sulphate than Katla! It is well known that Katla never seemed to have resulted in climatic disruption, while some Hekla eruptions are proven to have done so. And I think this is becaus Katla erupts from a shallow chamber, while Hekla erupts from a very deep chamber, hence much more gas.

    Interestingly Tambora was not that sulphurous. Look at the 1257-1258 eruption, that one was a mammoth, possibly larger than Tambora 1815.

    Also big eruptions indicated in 1600 (a well known, causing possible famines in Russia but this was also during the little ice age) and even let’s realize that while Tambore left no tephra layer in Greenland, the 1600 eruption that came from Peru did! And also did the 1258, from a similar location than Tambora.

    Also big are the eruptions in 1458 (already well known but reports are relatively scarse), 1229 Quilotoa, and 1641 in Phillipines. Wasn’t aware of these last two were so big!

    Veidivotn was surprising low, Only 0.4 in north hemisphere, plus 0.1 in siuth hemisphere supposedly linked to an unknown eruption. Veidivotn erupted as much magma than Laki, but apparently much less sulphate. It is interesyingly, because while Laki was largely efusive, Veidivotn had quite an explosive component, due to contact with groundwater. But this is another surprise nevertheless.

    Oraefajokull 1362, one of largest ash eruptions in Iceland since settlement, but no detectable sulphate, also surprises me, since itwas one of few to leave threat in Greenland. Clearly very ashy but not gaseous. Makes sense as it is away from the hotspot.

    Finally, I have something to add to the list: the 1340 unknown eruption is probably Hekla. The eruption happened 1341 and it was also one of Hekla biggest since settlement, and the ash travelled northwestwards, so am sure any sulphate would be easily deposited in Greenland!

    I could check whether other unknowns came from Iceland, provided a +- 3-5 years error in dating.

  7. And what happened in 1831? Dry fog over Europe, but a signal in Greenland rather weak (at least way smaller than Laki or even other Icelandic eruptions), but then it shows over the Antarctica ice core.

    Obviously not Iceland. And possibly not Alaska nor Kamchatka.

    Location estimated is at 42N but it could be more near the Ecuator. If dry fog came over Europe, but not over Greenland, then it must have reached Europe, from either the southwest subtropical jet stream, indicating it came from the Latin America. Or it came from Africa and affected Europe when in summer the winds from the Sahara push air northwards. Location probably East Africa. Only by a small chance it would mean something like Turkey or elsewhere around the Mediterranean.

    I would say to look first for Latin America. Anything there should have been documented if it happened anytime around 1831 +- 3 years. I am more inclined to say East Africa.

    I checked the Azores and there are no eruptions around that year. I wouldn’t expect the Azores, because there are larger chances of atmospheric sulphate getting to Greenland from there, though not the most common atmospheric path.

    • The dry fogs are an interesting mystery and is in fact what first made me look at the volcanic records (first triggered by a recollection of Diane!). The oldest record of a dry fog I could find is from AD 1252. This is before the huge eruption of 1258 and the cause is uncertain. Other reports are from 1348, 1557, 1733, 1755 (coincident with a Katla eruption), 1764, 1782 (Laki), 1783 (still Laki), 1819 (attributed to fires on the American plains) , 1831, 1834, 1846 (with reports in Scotland of an ‘offensive odour’), 1847. (This list is from the Astronomical register, vol 4-5, 1866, and obviously lacks more recent events.) Some are volcanic, others (probably most) are from peat fires in North America. Telling which is which is a problem, but most are likely from fires. The 1831 event was long lasting (4 months) and affected not only Europe but there were also reports from Africa, which to me makes a volcanic origin (Central America?) more likely.

      • Yesperhaps there are undocumented peat fires which would shock us if they would occur today. I reckon they happen in the arctic caused by methane catching fire.

        Laki happened in 1783 not 1782. Th dry fog report might be innacurate, or this year as well as for 1258.
        Katla 1755: it was largest Katla eruption since settlement. 1km3 eruption. So it might be the cause.

        Interestingly I find 1348 peculiar as it is the start of the black plague, and 1846 in Scotland, as this as during the Irish famine. Two interesting coincidences, which makes us think whether something was happening. And actually in 1845-1846 Hekla had a large ash eruption, so perhaps the culprit? Anyways, from south Iceland to Scotland is just a small distance for sulphur to travel. It’s a perfect temporal match and espacially it makes most sense! So, again Hekla is the culprit I find!

        1819: Tambora occured just 4 years before. Maybe again a dating problem?

        1764: Hekla erupted 2 years later (but ash cloud went northwards). Also Bardarbunga probably erupted this year.

        • For the old records the dates could be off, written records normally have accurate dates, but they may have been based on recollections. But 1819, the dates are not in doubt: this is newspaper material. Iceland is the most likely origin for those that are volcanic. The 1831 one first occured on the African coast – you may be right that Eritrea is a plausible origin.

      • I would doubt central america, Central america was well populated around that time, and we would likely have known of such a large eruption.

        I would guess somewhere in the pacific if I had to bet.

      • ‘offensive odour’ … a polite way of saying it smell like shit or rotten eggs?

        Safety Note: If you have grandkids, and they find an easter egg from the previous easter, don’t be surprised if they screech in terror if it explodes in a puff of yellow dust on them when they find it. That kid will never forget that smell…

      • There is also another culprit that is more likely to attribute to most of the dry fogs.
        Sahara. There is actually a dry fog currently in Guatemala that is in fact a large cloud of saharan dust that has been sucked up by a sandstorm and then blown across the atlantic. This happens quite often, I think it was last year that England got a saharan dust coating if memory serves.
        My wife has published a few nice pictures of the Guatemalan dry fog on her FB page.

        • I had been wondering about that possibility. In the UK, Saharan plumes drop dust and it is quite obvious what it is. They also tend to be short lived here. But back in time, this may not have been understood. A report of dry fog from Africa is unlikely to be dust – they know their dust there. This culprit may contribute elsewhere. Industrial pollution can also not be ruled out – industry is effectively a controlled fire, after all.

          • I think it would be productive to compare Italian and English records of dry fogs. Italians know very well what is Saharan or not. At least some of the dry fogs would probably be possible to eliminate that way.

      • The signal from Eldgar in Greenland is about 2.5-3 times that of Tambora, the second largest over the past 2000 years, behind Laki and just ahead of the 530 event. (The 530 event seems to be a double in the Greenland core: summing the two together makes it just behind Laki. This far back the dating and time resolution may not be as good as the paper claims.) The 870 event is about 1 Tambora. The 530 event, based on sulphate deposition, took place in the northern hemisphere, between 25N and 45N depending on whether it was a single or a double event. As usual, these latitudes should be taken with a healthy amount of doubt. It was possibly similar in size to the 1258 and 1458 events.

  8. I actually grabbed the wrong photo in my earlier post. This one is better for recent time periods. I’m pulling this as an excerpt of a post I had written (was never published).

    For me, the above image is incredibly interesting for quite a few reasons. First, it shows a relevant comparison of recent large eruptions to past large volcanic eruptions. Second, and most important is that we can start assigning corresponding dates to these spikes, allowing us to denote which eruption produced a crapton of So2 in a short amount of time.

    The 9200 B.C. Mystery

    What was most interesting to me from the above So2 chart is the enormous spike around 9200 B.C. This spike is the largest on the entire graph, which points to an eruption that put out more So2 than Kikai, Kurile Lake, Crater Lake, Taupo, and many other enormous eruptions that would be extremely catastrophic if they were to occur in this day and age. Does this signify that it was a larger eruption than the aforementioned eruptions? Not necessarily, since it may be closer to the ice core location, or may simply have been an extremely So2 rich volcano. Nonetheless, this was a huge eruption that is entirely unaccounted for.

    GVP Search For the Mystery Eruption

    Unfortunately, a search through the Global Volcanism Project’s eruptions does not yield any strong candidates outside perhaps Theistareykjabunga in Iceland. Theistareykjabunga would be close to the Ice cores, was a very large effusive eruption, and was likely very So2 rich. With that said Theistareykjabunga, is still an unlikely culprit. If Theistareykabunga could impact the So2 chart this dramatically, it would be very likely that we would see Laki with at least a comparable showing on here, but Laki is nothing more than a blip compared to some of the other large So2 spikes that are visible. Considering Theistareykjabunga’s eruption size wasn’t that much larger than Laki’s , it’s unlikely we would see an So2 output of such an enormous magnitude higher that it would dwarf all other Holocene eruptions.

    Unfortunately, this lead me to a dead end – I found no eruptions in the GVP database within that time period being listed as larger than VEI-5. So like many things, I stored away the time period in memory should I ever come across a volcano that could be a likely suspect within the relevant time period.

    Originally, I had thought I had found the likely suspect with the Fisher Caldera system, as the GVP had a large caldera eruption linked to 9200 years BP, but I confused 9200 years BP with BC, which provided a 2000 year gap in the data.

    • I was thinking about something. That sulphate travels to the Greenland ice core depending on wind. So one factor that you must enter in consideration is that for Icelandic eruptions, unless too big or too long, the wind is a critical factor. Whilst for big worldwide eruptions, the sulphate eventually reaches the arctic.

      Whatever happened between 11000 BC and 5000 BC, was either massive at a global scale, or local, possibly Iceland. It might be that the methodology has also some problem and as we go travel backwards in time, signals get larger. Or wind patterns in the north Atlantic were diferent prior to 5000BC and thus SO2 signals from Iceland got amplified.

      I could assign the peak 5200BC to Hekla 5 eruption, the 8300 BC to Saksunarvatn ash Grimsvotn eruption, and the 10100 BC to Vedde ash Katla eruption. Dating of these big eruptions is within 300 years.
      On Vedde ash: http://www.inqua2011.ch/?a=programme&subnavi=abstract&id=1917&sessionid=1
      Veidivotn eruptions also around 4000BC and 6500BC.

      But sincerely I do not know… There must be a reason for these high peaks…

      • I put my best guesses in here, I feel pretty confident about most of these, but there are quite a few that are big question marks. Most noticeably, you look at an eruption like Tambora’s, and then peer back 4000-8000 years, and there are a LOT of volcanic eruptions that show up with similar size signatures that we know nothing about.

        Given, this is the Greenland ice core, and some volcanoes are more so2 rich than others, so it’s not entirely indicative of eruption sizes, but it’s definitely worth noting at the very minimum.

        • I mostly agree with your identification of the SO2 peaks with past eruptions.

          Well, you can just add Hekla3 eruption to the 5200 BC eruption. Iceland is very near Greenland, so any huge eruption will show very well there. I told you on a previous comment, how Hekla seems to be desproportionate high in sulphate compared to other Icelandic eruptions. Remember 1.25 for Hekla 1104 eruption, while Laki is only four times larger. Hekla3 eruption was several times stronger than the 1104 eruption, at least it shows that way in soil profiles and pumice pieces thrown at great distances from the volcano (more than 60km away). It must have been huge, there is some agreement it was a VEI6.

          Then, add the likelihood of the VEI6 Grimsvotn eruptions around 8300 BC and Katla around 10000 BC.

          I think an Icelandic VEI6 will show as big as a non-Icelandic VEI7 in your ice core data.
          Just as an Icelandic VEI4/VEI5 shows as big as some VEI6 non-Icelandic eruptions.

          • True: Iceland is disproportionally represented in Greenland (as long as the wind blows from the East!) It may be an idea to see which of the large spikes are also seen in Antarctica. An example is the 9200 BC one which I think is also strong there. There was also one around 6000BC. Iceland was I think the main volcanic region which was covered by large ice age glaciers. Alaska and Kamchatka did not have the major ice cover that Europe and Canada had.

          • Yeah, I can see Hekla and Grimsvotn working in these instances. I think one factor that’s relevant in whether Icelandic eruptions show up in the eruption record is whether they’re explosive or not. A VEI-6 (say hekla 3) eruption in Iceland would still likely have less overall tephra output than an eruptions such as the Thorsa lava flow, but the style of eruption blasts more so2 high up into the atmosphere. I believe this allows for the so2 to linger much longer and affect the climate more than it otherwise would (Lurking knows much more about this than I).

            • Only moderately so, and mainly in that I have the idea that COS may be emitted also in the degassing of flood magmas, and Carbonyl Sulfide exists in the atmosphere long enough to get transported to the tropopause at roughly 60°N via the normal atmospheric circulation cells. I have no proof, just an idea. Above the tropopause, COS can be dissociated via UVc light and become part of the Sulphate formation process. 200 and 270 nm light can easily split the normally tightly bound molecule. I don’t have e-fold data for it, but SO2 and Sulphate rates were covered in a previous VC article that I wrote. SO2 takes about 2 months to convert to sulfate, and sulfate takes about 50 months to sediment out of the stratosphere.

    • could an ash sample be retrieved from the ice core, then one could work out where it might come from

      • On occasion, they do that. Other deposits are found all over the place, many of the samples have been chemically linked to specific eruptions.


        Playing with the sample data there, you can even get an idea of what sort of variations occur in single volcanic systems by plotting out the chemical signatures of the different eruptions. I’ve done this using a plotting program that I use quite a bit. On some of the eruptions, you can infer the Sulfur content based on the FeO and TiO2 ratios. The formulas that I have seen are mostly concerned with flood basalts, so I don’t know how well they translate to typical explosive eruptions.

        • keeping an library of sorts for samples of volcanic eruption deposits would help presently not asingend ones in the future

  9. In the scientific literature it is common agreement that Salamas produced a DRE 40 km 3 :
    We have to go back the Quaternary to find a DRE 80 km3 with Mount Aso :
    Here list of known eruptions compiled in the database LaMeve with contributions of Natalia I Deligne 🙂 Click submit to have the list alphabetically or put the name of the volcano :

    • Thank you for the comments! My notes say that the published value of 40km3 for Salamas/Rinjani should be considered a lower limit, based on the analysis that was used to derive it. I would have to go back to the original papers (Oppenheimer 2003 – eruption volume, not DRE-, and Lavigne 2013 – the PNAS paper) to find out why I wrote that down.. I obtained the value of 80km3 by scaling to the most recent published values for the DRE of Tambora (Kandlbauer 2014), given that the ice signal makes Salamas/Rinjani about twice as large. It is indeed a rather uncertain number! If you prefer, 60+-20 km3 would cover the likely range. The climatic effect of this eruption was considerable, certainly much larger than that of Tambora.

    • No,we definitly do not need to go back to the Quaternary to find such an eruption.
      Aniakchak was 100km3 DRE weighing in as a sturdy VEI-7. Interestingly enough it occured in the same year as Theras great eruption and for many years the ash deposits in the ice was misinterpreted as coming from Thera. In reallity the bulk of it was from the far larger Aniakchack eruption.


      I would also like to mention that the base surge limiter of 50km we are using for our decade volcano program of MDE scale is derived from the Aniakchack eruption. Enjoy the read.

  10. In my research use :

    Traufetter, F. , Oerter, H. , Fischer, H. , Weller, R. and Miller, H., 2004, Spatio-temporal variability in volcanic sulphate deposition over the past 2 kyr in snow pits and firn cores from Amundsenisen, Dronning Maud Land, Antarctica , Journal of Glaciology, vol. 50, no. 168, pp. 137-146

    Gao, C., A. Robock, and C. Ammann, 2008, Volcanic forcing of climate over the past 1500 years: An improved ice core-based index for climate models, J. Geophys. Res., 113, D23111, doi:10.1029/2008JD010239

  11. Time for some Icelandic hypothetical doom 🙂


    70000 years ago Torfajokull erupted 16 km3 of rhyolite (VEI6), in a massive caldera ring fracture type of eruption. It deposited a layer of 5.5cm thick of ash in the Norwegian sea!

    This is one of my favourites Icelandic volcanoes. In all its counted Holocene eruptions (about 8), they always come in a peculiar pair. First Bardarbunga injects a large amount of basaltic magma along Veidivotn (by a method that we now know to resemble the dike seen in August 2014). Then (by the same manner the 2014 dike nearly reached Askja), the Veidivotn dike reaches Torfajokull and triggers an eruption there, of rhyolite lava flows and moderate ash eruption. This happened both in 1477 and in 870, and a few other times before settlement. So, one volcano triggers another, repeatedly over time.

    In soil profiles, we observe that the Torfajokull rhyolite ash layer comes first and then the Veidivotn basalt ash layer (respectively light pink colored, and dark colored). Thus the dike might first trigger the rhyolite eruption and then the fissure unzips backwards.

    However the 70000 years BP ring-fracture mammoth eruption is what called my attention.
    Once again, the link https://davemcgarvie.wordpress.com/2012/05/18/torfajokull-waiting-in-the-shadows-7/

  12. This will sound a bit strange, but, is there anywhere on Earth where ice goes back 70 or so million years? If so, could the asteroid impact and the Indian trap formation show up on it? Just wondering. Maybe the bottom of the Antarctic ice?

    Released from Limbo /Lugh

    • Not ice: before the onset of ice ages, there was little or no permanent ice on Earth. But ocean cores can go back much further. These are sediment cores, and can measure conditions in principle as far back as the life time of ocean floor (about 200 million years). There are no annual layers so you can’t count them. Dating has to be done in other ways. They are the best way to measure climate going back to the time of the dinosaurs. Ejecta from Toba are known from ocean cores.

      • Geolurking: isn’t Antartica in continuous glatiation since a couple millions years ago (like 2-5MY), rather than just 400.000 years. Apparently it started getting ice caps around 35MY ago, but permanent ice caps only more recently.

        Albert: go to Mariana trench in the northwest Pacific Ocean. There ocean floor is oldest on Earth. Probably over 100MY old. But the Atlantic is younger than 65MY so I think you you need to go to the Pacific ocean floor to catch the 65MY impact material.

        • Yes, but the ice is flowing so any ice older than cirka 400 000 years old has by now ended up in the water and melted.
          For sediment cores you can also go to central africa. Lake Turkana is a really good place for sediment cores dating back all of human history. Yes, it does not extend as far back as the oldest ocean floor, but it is far more detailed.
          Also, there is old seafloor that is uplifted and visible at surface, we have examples of that spanning the last 350 million years and we do not need to bother with the diving part.

    • The only real book that I liked was Global Tectonics 3rd edition… but it doesn’t really deal with volcanoes. It looks more at the overall process that drives them. Definitely a “go-to” book for things geologic.

      Most of the volcano centered stuff I read are the papers that occasionally get linked here and there. There is a wealth of information in papers.

      Pretty much anything by Sparks if you can get your hands on it. For Canary Island stuff, books by Caracedo is about the best that you are going to find. If you find anything by Perez, best of luck to you. He tends to focus specifically on Perez.

      By far, one of my favorite papers was by Mastin et al. I tend to wind up citing that one more than anything else when it comes to estimating the ejecta rate of the occasional VAAC plume report. Since I mentioned it, Carl made the poignant observation that a VAAC may tend to over report a plume’s height. Their focus is on aircraft safety, not on emperical volcanic activity. So, when you use a VAAC report to start your calculation, remember that the ejecta rate may be estimated a bit high. With enough reports over a long enough time, you can come up with a pretty close VEI estimate for an eruption as it’s pondering along. You have to interpolate the missing data, but if you can cobble together a mass ejection rate on a second by second basis, then project that into the future, summing it you you can nail a pretty good VEI eastimate for an ongoing eruption. You just have to make sure that you account for errors in your model and take it with a grain of salt.

      A multidisciplinary effort to assign realistic source parameters to models of volcanic ash-cloud transport and dispersion during eruptions Mastin et al (2009)

      And, if you are not math focused, the paper is a good read by seeing the ideas that went into the construction of their formula. (which is essentially an update of Spark’s formula).

      • BTW, Global Tectonics can be a bit pricey (mileage may vary). If you can locate it in a library you may find the book quite illuminating.

        On a sad note, the lead author for that book died, and two of his cohorts finished it and sent it for publishing.

  13. Reblogged this on Willow Andreasson's Journey Into The Mysteries of Life and commented:
    It’s been a while since I’ve been able to visit one of my favourite Cafes and so, armed with a mug of coffee, I have just enjoyed reading this meticulous article, written by Albert. I am aware that ice core examinations reveal vast treasure troves of information in relation to past volcanic eruptions. Recommended reading for this peaceful Wednesday morning and yes… it is likely a wise idea to keep an eye on the volcanoes listed on Albert’s spreadsheet. Thank you for sharing this information with us, Albert.

  14. The activity at bardabunga is interesting & intriguing. Are these earthquakes a sign of new magma? I am getting a sense that we are still in new territory in terms of learning what the system is doing.

    • Well, the new “phase” of activity began with the tremor increase at Hamarinn, probably due to hydrothermal activity, since a few days later, a glacial flood began from two cauldrons at Hamarinn.
      The flood subsided, but the tremor is still ongoing, unlike at Grimsvotn where the tremor stopped after the flood.
      The current tremor is also evident on a larger area than when the flood happened from Grimsvotn.
      Just look at HUS, HAM and VON stations. Iceland is under a ridge at the moment, so it will be wind-noise free for a day more or two. It could be that the tremor will slowly decrease and normalize. Or it will continue and go into another phase.

      • We just have to look at the well-documented Icelandic history of eruptions, to guessestimate what happens next.

        I say we will see a significant increase with eruptions in Vatnajokull, but also elsewhere in Iceland. This marks a frequency of eruptions every 2-3 years instead of the usual 5-8 years. Also larger eruptions will also occur, if historical patterns continue. By large I mean stuff like the Myvatn fires of the 1720s, the Oraefajokull eruptions, the 1875 Askja episode, etc.. Usually the big ones occur when the hotspot activity peaks in this cycle of 130 years.

        Also ritfing occurs widely, both north and south of Vatnajokull. We should expect this now. Not everywhere but in several spots. Back in the 18th century peak, rifting occurred between Katla, Laki, Bardarbunga, Holuhraun, Kverfjoll and Krafla. Also larger quakes along the SISZ and the Hengill region. In the late 19th century peal, rifting occurred between part of the Veidivotn rift, Bardarbunga, Thordarhyma, Holuhraun, Askja and all the way north almost to Krafla.

        So I expect most of the central part of Iceland to rift one way or another, and also a few locations well outside of Vatnajokull, like along Reykjanes and also large quakes along SISZ: This is what we could expect.

  15. Click to access KamchatkaBindeman2010.pdf

    This is one of the best papers I’ve read on Kamchatkan volcanoes. One of the interesting points it finds is that the majority of the large eruptions in Kamchatka occured not during interglacial periods, but instead during glacial maximums. This is rather surprising in some ways, but has some reasoning in the thought that heavy glaciation promotes extra pressure build-up, allowing for fewer intermediate eruptions, and more large singular eruptions.

    Ont interesting excerpt on the Karymsky volcanic region.

    “Several large calderas are associated with the Karymsky Caldera Center (Fig. 2), which has overall dimensions of 55× 65 km, and a total volume of exposed erupted material (younger than 3 Ma) of 1700 km3”

    That puts the Karymsky volcanic center on par with some of the larger volcanoes in the world in terms of overall output over a long-term geological period.

    Another interesting excerpt touches on how there is a direct correlation between larger explosive eruptions and crustal thickness in island arc settings. This is interesting to me since you get large calderas in regions with exceptionally thin crust as well (New Zealand, Southern Japan, etc).

    “A recent compilation by Hughes and Mahood (2008) show that Kamchatka has the highest density of calderas larger than 5 km/1000 km of island arc, for any volcanic arc, and this is likely due to a combination of high arc perpendicular convergence rate (7–8 cm/yr) and the presence of thick continental crust.”

  16. Great article, Albert 🙂

    For those still watching the car park at Hakone:


    From a drone flight to assess where to put monitoring equipment, scientists found that sulphur has been emitted and three hot spring facilities have suffered damage. The article does not say what cause the damage – increased emissions, earthquakes or deformation.

  17. I am lifting down an answer to CBUS in regards of the SO2-spike of 9300BC so it would not drown.

    Two things that makes me rather unsold on it being Kamchatka, Kuriles and Alaska is the culprit.
    1. The emitted lava is subduction induced and is as such less rich in SO2. Typically in the order of 3 to 10 times less gassy than the fresh MORB/Plume derived Icelandic magma with the plume magma being more gaseous than the MORB-derived magma.
    2. There are 4 likely candidates for Iceland, and none known for the period in Alaska, Kurilles or Kamchatka.

    The ones known for Iceland are Skjaldbreidur (Prestahnukur, Langjökull), the largest holocene formation of a shield volcano on the planet. The amount of lava erupted larger than the amount erupted at Theistareykjarbunga since the shield is far larger. Eruption occured 7 550BC +/-500 years. Rough estimate gives 70 cubic kilometers for the shield and the lava hraun.
    Theistareykjarbunga, 35 cubic kilometers of shield formation and a 15 cubic kilometer lava hraun. Uncertain dating at 9500BC.
    Odadhahraun (Askja), 35 cubic kilometers. The largest holocene lava flow on earth. 8 910BC +/-200 years.
    Grimsvötn (Saksunarvatn tephra). 10 to 15 cubic kilometre DRE, amount of emplaced lava is unknown, probably surpassing the amount of tephra. 8 230BC +/-150 years.

    In regards of the dates cited above. You can drive a truck through them.

    Now, what would we need in the AKK-area? Well, a fairly stupendous unknown caldera. To release the same amount of sulphates as from Theistareykjarbunga and Skjaldbreidur we would need a caldera that formed in a 210 to 350 cubic km eruption.
    For Odadhahraun we would need something in the order of 175 to 350 cubic km. And for Grimsvötn we would need 300 to 500 cubic kilometers ejected DRE.

    Now, back to the ice cores. AKK would have ejected prodigious amounts of ash during such a large eruption, and those amounts are just not there in the ice cores (not in those amounts at least). I think we could fairly reliably point to one of the above culprits.
    Now, let us look at the dating of my cited 4 icelandic large eruptions (and a few other tidbits).
    Skjaldbreidur is first of all not the one due to having happened at the wrong time. Also, it happened over an extended time period, about 100 to 150 years, so the SO2 was diluted by time.
    Grimsvötn is easily detected by the ash it left. So, that is not the culprit.
    Now we are left with Odadhahraun and Theistareykjarbunga. The initial 15 cubic kilometer hraun of Theistareykjarbunga was released at the same speed as Laki, but was not large enough to form enough SO2, and the shield formation that followed took a couple of decades. So, no fun with Theistareykjarbunga.
    Odadhahraun, 35 cubic kilometers is an estimate made based on surface area covered. It does not take into account infilled lakes, valleys, rivers and other general low points, as such it was probably larger. It is at the lower time scale of being a likely culprit, but on the other hand it was erupted inside 1 year. My bet? Odadhahraun is the far likeliest source for that SO2-spike and I will keap to it unless someone can find a large unknown caldera in AKK that occured during that timeframe.

      • It could have been from the tropics for all we know, but I’m sort of assuming a northern source as that would provide a more direct deposition in the Greenland ice core.

        as for the akk region, the key here is “known” large eruptions. Remember that we just within the last 3 years found the likely source of the 1258 eruption, and that was in a well studied region. If you push the time back 10,000 years, things are obviously a lot murkier.

        Another thing to note regarding older eruptions is that glaciation is extremely effective at hiding past large eruptions, and much of the world was glaciated at this time. Not only does tephra not stick well when it lands on top of a thawing glacier, normal erosional processes will shave off significant chunks of ignimbrite deposits, making eruptions appear smaller than they may have actually been.

        Albert mentions grimsvotn, but could it have had two massive eruptions in this close of a time period? I realize it’s prolific, but I would be slightly skeptical at least. Does anybody have a link or source to information on grimsvotn ‘s early holocene large eruptions?

    • 4 of the 6 spikes from before 8000BC which Cbus identifies coincide with ash deposits, and all are reported to be icelandic. The one at 11000BC is Hekla, 10200 is Vedde, and both 9300BC iand 8300BC are identified as Grimsvotn (Saksurnarvatn)

      • I like both your ideas Albert and Carl ones. I also gave my bets. But I guess we aren’t sure until some ash is detected in that layer and the identity attributed.

        I am sure that Vedde and Saksurnarvatn feature there. The shield eruptions well, I think the sulphate may be concentrated in the first few years of the eruption within the ice core. Carl, you orgot Trolladyngja, the tallest shield volcano in Iceland (Bardarbunga system supposedly, its near Kistufell). Btw, Odadahraun is linked both to eruptions of Askja and Bardarbunga system (spots like Trolladyngja and Holuhraun).And, importantly most of Odadahraun actually came from a few big shield volcanoes north of Askja, called Kollotadynja, Ketildyngja and Kerlingardynja, whether they are part of Askja or independent system is part o controversy.

        All of these ages; yes you guess it, early Holocene. 6 shields eruptions: Trollandynja, Skjaldbreidur, Theistareykabunga, Kollotadynja, Ketildyngja and Kerlingardynja. Plus Vedde and Saksurnarvatn eruptions, these which left thick ash layers in Iceland, the thickest in thousand years.

          • 😀 Hehehe!

            Yes I forgot about Irpsitdynja. I visited it last September by airplane and even comtemplatd the illegal act of going there on foot, but it was too dangerous.

            Now I am very eager to return this August to Iceland and go there hiking to its top, if possible. I am mostly thinking doing this with an Icelandic hiking group, just in case you think I’m too crazy.

            I must confess I didn’t expect the Holuhraun eruption to be a sort of crosshybrid in between a Laki-style fissure and Icelandic shield volcano style. It wasn’t that much of a fissure, as quicky the fissure reduced to a crater row of quickly just a single vent, just like a shield volcano would do. So I dreamt for an eruption lasting up to 100 years like Sjaldbreidur. But it didn’t happen.

            No I think those shields weren’t just built in a single eruption but rather several.

            Last shield eruption in Iceland was circa 1000 AC, in the west part of Langjokull. Check it out. Beautiful shield there, vikings were around but I never heard of a proper written record about it. Its also a remote place.

            I guess both Irpsitdynja and Carlvatn formed for a while but they never turned out to be the great geological features we dreamt of.

            I still think one day in the far future, there is going to be a rift valley lake at Holuhraun.

      • “11000BC is Hekla”

        Albert, is Hekla that old?
        I thought Hekla was much younger (within Holocene), but there must have been a first eruption of Hekla.

        I have sometimes suspected that Tjorsáhraun was the birth of Hekla, as the Bardarbunga fissure radiated southwestwards, but this was a gut feeling I had once, and obviously a wild and bold claim. But you know every central volcano in Iceland starts as a crater row of a fissure swarm that might well have originated from another central volcano. What happened in 2014 shows a lot more is possible than what we think.

        Carl, any idea of Hekla first eruption?

        • The oldest ash layer in Greenland attributed to Hekla is 13000BC. This is from a paper by Mortensen (2005). Thye say that Hekla is the only producer of intermediate magma in Iceland, and the main one for rhyolite. But in Iceland itself there is evidence for a pre-holocene Hekla eruption and so it is interesting that there is evidence from Green;and that Hekla existed before the Holocene. But there are also some indications that Katla may have contributed to this ash.

          Also from this paper: “the eruption of Oræfajokull in AD 1362. is
          the largest of the historic rhyolitic eruptions in Iceland. This
          eruption deposited abundant ash particles in Greenland ice
          cores, but it has only formed a small chemical sulphate signal
          —possibly because the eruption was partly subglacial.”

  18. For the “missing” volcanoes of 1838 and 1808/9, were they big ash producers, big sulphate producers, or both?

      • Flood basalt is a big SO2 producer. If the cuplrits are in the tropics, how quickly can lava flows get covered by vegetation?

  19. Oh this is incredibly sad and criminal!

    I just happened today to join for the first time volcanocafe in facebook (i was in a frenzy suscribing to new groups around) and then see the post of Carl and was shocked by it. I can’t imagine all the good posts, and all the good discussions, being lost!!!

    We should back-up the website!!!

    It interests, from a philosophical perspective to know how in hell would want to do such a thing. Would it be just a hacker of someone directly targetting us. No reason for that.

    Anyways I am excited to join the facebook group at least..

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s