El Salvador eruption in A.D. 536 triggered extreme global cooling

Would have filled the skies with ash and dust for more than a year.

El Salvador’s Lake Ilopango – an ancient crater lake boasting 100-meter-tall cliffs – may have been the site of one of the most horrific natural disasters in the world. It may also have caused the extreme climate cooling and crop failures of A.D. 535-536, reports Robert A. Dull of the University of Texas at Austin.

Ilopango volcano - NASA
Ilopango volcano – NASA

The massive event, the second-largest volcanic eruption in the last 200,000 years, would have instantly killed up to 100,000 people, displaced up to 400,000 more and filled the skies with ash and dust for more than a year. “This event was much bigger than we ever thought,” Dull said.

Such an eruption would explain the extreme global cooling of A.D. 535-536, an 18-month period of cloudy skies, crop failures and famines described in both Roman and Chinese historical accounts. “It’s very well established this event took place,” Dull said. “The question has been the cause.”

Sulfate levels in ice cores from Greenland and Antarctica dated to A.D. 536 indicate that a massive tropical volcanic eruption had ejected ash around the planet. However, the location of such a tropical volcano that could have produced a sufficient volume of ash to block the sun remained unknown.

Over the years, researchers have discovered huge tephra deposits as far away as Nicaragua and Honduras, as well as in offshore deposits.

The wider geographic distribution of deposits indicates that the eruption produced a huge  volume of ash and debris, which Dull now calculates at 84 cubic km (20 cubic miles), more than 64 times as much as the 1980 eruption of Mount St. Helens.

Twenty cubic miles!

“An eruption that big has to leave a big hole in the ground,” Dull said, and the dimensions of Ilopango’s caldera, at 11 kilometers by 17 kilometers, fit the bill. Such an eruption would have been larger than the 1816 eruption of Tambora, which caused the “year without a summer.”

(According to Wikipedia, the Tambora eruption occurred in 1815. It created an ejecta volume of 160 km3 (38 cu mi), much larger than Ilopango. During the following year crops failed and livestock died in much of the Northern Hemisphere, resulting in the worst famine of the 19th century.)

The most recent eruption at Ilopango took place in 1879-1880, during which a lava dome formed creating the Islas Quemadas in the central part of the lake


Thanks to J.H. Walker for this link

19 thoughts on “El Salvador eruption in A.D. 536 triggered extreme global cooling”

  1. The El Salvador eruption may have triggered extreme global cooling in AD536.

    David Keys, 1999 (CATASTROPHY. An investigation into the origins of the modern world) proposed a volcanic explosion in the present Sunda Strait, based on ancient Chinese and Javanese manuscripts. The evidence in ice cores fits both locations. The presented theses w.r.t. the sudden onset of global cooling are similar.

    • The paper which suggested this SA site as one of the largest eruptions to fit the time line plot, the original paper suggested two others of a large size in various parts of the globe.
      The point which isn’t emphasised is the Sun’s Trefoil orbit around the Solar System BarryCentre during Solar Minimums, imparts gravitation effects on the Terrestrial planets and enhances the likely hood of an eruption happening much sooner and in company with others on a worldwide basis.
      I suggest that most Solar Minimums, if not all, are accompanied by Volcanic Rifting events Shield volcano events, or plate slippage earthquakes, some very , very large. This modern Solar Minimum has another 30 years to run, before the Sun returns to the next Solar Warm Period with stable ordered orbits and much reduced volcanic activity.
      For example Krakatoa erupted in 535 during the same time frame, it has also erupted in other Solar Minimum periods named for large MIA events for example, Shallow AMP events 850, 950, Oort – 1050, Oort end period -1150, Wolf -1320, and Sporer – 1530, Gleissberg -1883.
      So have many other volcanos one at a time throughout history, fortunately they all haven’t gone off all at once.
      The other eruption tagged in this time frame is https://en.wikipedia.org/wiki/Rabaul_caldera
      Researcher dare not suggest other reasons for the Warming or Cooling periods the earth experiences other than Human emitted CO2, otherwise they are not published, don’t get research grants and don’t eat. The government funded gravy train.
      Government fraud, begat’s greater fraud, until the states currency is worthless, we then have 1929 and 2008 financial crashes. Not one banker has gone to Jail for this financial cockup. The Carbon Credit fraud will be far worse and destroy the Wests banking system and its economy as it unwinds.

      • JimBob,

        Rabaul had been tagged as a possible candidate for the AD 536 climatic events (Stothers and Rampino 1983; Stothers 1984) but the date of its actual eruption has always been uncertain until very recently. Previously a total of five samples of charcoal recovered from Rabauls pyroclastics formation had been radiocarbon dated with ages of 1280+/-81, 1390+/-55, 1430+/-90, 1505+/-90 years BP (Heming 1974) and 1360+/-55 years BP (Nairn et al. 1989), where BP = before the year 1950.

        Three more radiocarbon dates had also been obtained from palaeosoil samples below the pyroclastic formations with ages of 830+/-70, 1600+/-70 years BP (Nairn et al 1995) and 1450+/-60 years BP (Hemming (1974).

        It is apparent that the charcoal dating is much more consistent than the dating of palaeosoil, and a chi square test on the charcoal results showed that the ages were statistically indistinguishable with a calculated weighted mean of 1380+/-34 years BP (McKee et al. 2011).

        Unfortunaetly, it is not a simple task to simply just subtract the radiocarbon date from the year 1950 to get the true calender date since the amount of radiocarbon within the atmosphere varies over time, and so one must use the radiocarbon calibration curve to convert radiocarbon ages into true calander ages. These calibration curves for the last 10,000 years are created by successively radiocarbon dating precisely dated tree rings and comparing the radiocarbon age with the known calendar age of the tree rings. One can then in principal take any radiocarbon age, and determine the most probable date for that radiocarbon measurement.

        When this is done for the weighted mean age of 1380+/-34 years BP, we obtain a probable calibrated year of eruption of Rabaul of 634-670 (with 68.2% confidence) and 600-687 (95.4% confidence). Clearly, Rabaul did not cause the 536 events.

        However, in July 2015, a paper was published which reports the use of Bayesian wiggle-match radiocarbon dating on a sample of charcoalized tree log removed from the the Rabaul Pyroclastics Formation (McKee et al. 2015). The idea behind this method is simple. Just like how the initial radiocarbon calibration curves created, if one takes a series of radiocarbon measurements from a series of rings from the outside to the inside of the log, and “wiggle” these ages on the calibration curve to achieve a “best-fit”. This can produce extremely accurate results.

        McKee et al performed this analysis on the tree rings from a log from the rim of the Rabaul caldera, and found that the log was most likely killed and buried by the pyrocalstic material somewhere between AD 667 and AD 699. When we now look to the ice core records (Sigl et al 2013; Sigl et al. 2015), we find that there are only two significantly large tropical/equatorial eruption signals in this time period; 681/2 and 684. Both dates coincides with years where frost damage occurs in bristlecone pines in North America in AD 681 and AD 684, which have been argued to be good proxies for large explosive volcanic eruptions (LaMarche and Hirshboeck 1984, Salzer and Hughes 2007, Baillie and McAneney 2015).


        (Baillie MGL and McAneney J (2015) Tree ring effects and ice core acidities clarify the volcanic record of the first millenium. Clim Past 11:105–114. doi:10.5194/cp-11-105-2015

        Heming RF (1974) Geology and petrology of Rabaul Caldera, Papua New Guinea. Geol Soc Am Bull 85(8):1253–1264

        LaMarche VC Jr, and Hirshboeck KK (1984) Frost rings in trees as records of major volcanic eruptions. Nature 307:121–126

        McKee CO et al. (2011) A remarkable pulse of large scale volcanism on New Britain Island, Papua New Guinea. Bull Volcanol 73:27–37.

        McKee CO et al. (2015) A revised age of AD 667–699 for the latest major eruption at Rabaul, Bull Volcanol 77: 65.

        Nairn IA et al. (1989) Rabaul Caldera, Papua New Guinea 1: 25,000 reconnaissance geological map and eruption history. New Zealand Geological Survey, Department of Scientific and Industrial Research.

        Nairn IA et al (1995) Geology and eruptive history of the Rabaul Caldera area, Papua New Guinea. J Volcanol Geotherm Res 69:255–284.

        Salzer MW and Hughes MK (2007) Bristlecone pine tree rings and volcanic eruptions over the last 5000 years. Quat Res 67:57–68

        Sigl et al (2013) A new bipolar ice core record of volcanism from WAIS Divide and NEEM and implications for climate forcing of the last 2000 years. J Geophys Res-Atmos 118, 1151-1169, doi:Doi 10.1029/2012jd018603.

        Sigl et al (2015). Timing and climate forcing of volcanic eruptions for the past 2,500 years. Nature DOI: 10.1038/nature14565.

        Stothers RB (1984) Mystery cloud of AD 536. Nature 307:344–345

        Stothers RB and Rampino MR (1983) Volcanic eruptions in the Mediterranean before AD 630 from written and archaeological sources. J Geophys Res 88(B8):6357–6371

    • In the other story on this page – Geologist exposes climate change hoax and with a video – the geologist mentions Krakatoa and Indonesia at 535 AD. Combined with Salvador as a one/two punch, triggering cooling from aerosols on both sides of the earth in sequence? Wiki says there was a Krakatoa event in 535 AD. Even if not as big as the 1883 blast, would it’s chemical output have a large impact on the atmosphere? Just wondering.

  2. Snowy winter in Moscow:
    From March 3, 2016 the increase of fresh snow in the city was 245 cm. This is more than 160% of the seasonal norm.
    And there is still almost a month of winter weather.
    So this winter will be one of the most snow in the annals of the Moscow weather.
    “Global Warming or Climate Change … I know.”

  3. The year 536 is significant. Looking at european history it was the start of famine, plague, death and political disruption. People who relied on farming were decimated. It even gave rise to the king arthur legend. King arthur from wales saved his people by raiding the scots by stealing their cattle. Hunters and those who husband livestock survived better than farmers. It was a disaster all through the northern hemisphere.

  4. Ilopango is probably not responsible for the AD 536 cooling event, but it may have been responsible for the following AD 540 cooling event.

    The reason being that the information regarding a large tropical eruption in AD 536 is incorrect, or rather is it based on dating of ice cores.

    In 2008, Larsen et al. published data that showed that there was two large eruptions eruptions dated to 529 and 533 in Greenland ice cores (using the Greenland Ice Core Chronology 2005 timescale (Vinther et al. 2006). These appeared to be Northern Hemisphere and tropical respectively. However, Baillie (2008, 2010) argued that there was a systematic dating error in this suite of ice cores and suggested that moving the ice core dates forward by 7 years in the the period AD 500-700 resulted in a better match of ice core volcanic horizons with severe climatic effects in precisely dated world tree ring chronologies, as well as dates of historically recorded dust veils.

    Then in 2013, Sigl et al. published data two new cores from Greenland and Antarctica (NEEM and WASI Divide respectively). By comparing the bi-hemispheric sulphate signals in these two cores, they showed a large northern Hemisphere eruption dated again to 529 and a tropical eruption in 532/3. However, the two cores were constrained by the GICC05 timesscale from Greenland cores, and Baillie and McAneney (2015) demonstrated that again moving the dates for these ice cores forward by approximately 7 years in the first millennium AD created a better agreement between large eruptions and tree ring effects as well as historically recorded dust veils. Furthermore, Sun et al (2014) had also demonstrated that there may have been a chronological offset in the GICC05 timescale amounting to around seven years before AD 1000.

    Finally Miyake et al (2012, 2013) identified the radiocarbon signature (excess radiocarbon production by extremely energetic cosmic ray events) in Japanese cedar dated to AD 774 and 993. This observation was replicated in global trees Usoskin et al. (2013), and attributed to a massive, or a series of large Solar proton Events. Such events not only produce excess radiocarbon, but also excess beryllium 10 isotopes which precipitate quickly from the atmosphere and onto polar ice. Sigl et al (2015) investigated the beryllium 10 isotope concentration at annual resolution in ice cores and found that the dates for peak 10Be deposition were 767 and 987 in the GICC05 timescale, confirming that an chronological offset of around 7 years did indeed exist in the GICC05 (and all ice cores based on this time scale), before AD 1000.

    Sigl et al (2015) then implemented a better layer counting of the ice cores, and confirmed that the 529 Northern hemisphere eruption was actually AD 536, and that the 533 tropical signal signal was AD 540, and thus were indeed the cause of the AD 536 and AD 540 climatic effects recorded in history and tree rings.

    So the upshot of this is that Ilopango may not have caused the AD 536 event, but it is possible that it caused the AD 540 event (which seems to have been much more severe than the previous vent, possibly due to it occurring so soon after another very large eruption). The trouble is that AD 540 is just on the limit of the radiocarbon date for the Ilopango eruption, and the uncertainty is quite large, so AD 540 may not be the eruption date, and another volcano may be to blame for the AD 540 climatic effects. Ilopango is still in consideration as having erupted around AD 540 if we go by the tentative (and questioning) attribution of the this eruption to this year in Buntgen et al (2016).

    The important take home message of this though is the now recognisable hazard of two large eruptions within a few years of each other. Indeed, Buntgen et al. have argued that the double volcanic punch may have been a trigger for a 125 year long “little ice age”.


    Baillie, M. G. L. Proposed re-dating of the European ice core chronology by seven years prior to the 7th century AD, Geophys. Res. Lett., 35, L15813, (2008).

    Baillie, M. G. L. Volcanoes, ice-cores and tree-rings: one story or two?, Antiquity, 84, 202–215, (2010).

    Baillie and McAneney, Tree ring effects and ice core acidities clarify the volcanic record of the first millennium, Clim. Past, 11, 105-114, (2015).

    Buntgen et al. Cooling and societal change during the Late
    Antique Little Ice Age from 536 to around 660 AD, Nature Geoscience. DOI: 10.1038/NGEO2652 (2016)

    Larsen et al. New ice core evidence for a volcanic cause of the A.D. 536 dust veil, Geophys. Res. Lett., 35,
    L04708, 2008.

    Miyake et al. A signature of cosmic-ray increase in AD 774-775 from tree rings in Japan. Nature 486, 240-242, (2012).

    Miyake et al. Another rapid event in the carbon-14 content of tree rings. Nat Commun 4, doi:Artn 1748 Doi 10.1038/Ncomms2783 (2013).

    Sigl et al. A new bipolar ice core record of volcanism from WAIS Divide and NEEM and implications for climate forcing of the last 2000 years. J Geophys Res-Atmos 118, 1151-1169, doi:Doi 10.1029/2012jd018603 (2013).

    Sigl et al. Timing and climate forcing of volcanic eruptions for the past 2,500 years. Nature DOI: 10.1038/nature14565 (2015).

    Sun et al. Ash from Changbaishan Millennium eruption
    recorded in Greenland ice: implications for determining the eruption’s timing and impact, Geophys. Res. Lett., 41, 694–701, (2014).

    Usoskin et al. The AD775 cosmic event revisited: the Sun is to blame. Astron. Astrophys. 552, (2013).

    Vinther et al. A synchronized dating of three Greenland ice cores throughout the Holocene. J Geophys Res-Atmos 111, (2006).

    • There’s good historical support for some kind of truly catastrophic event in 535 CE. Writers of the period describe a year with a very dimmed sun, or no sun for long periods. That event was immediately followed by the first documented outbreak of bubonic plague in the eastern Mediterranean. No one really knows just precisely what the event was, but it DID occur in 535 CE. Documentary evidence ranges from the Eastern Mediterranean to China. One intriguing but problematic document from Indonesia seems to indicate that Summatra and Java were one island prior to an enormous eruption where the Sunda Strait is.

      Procopius, for instance describes the effects of some really disastrous event in 535-536 CE:

      “… during this year a most dread portent took place. For the sun gave forth its light without brightness … and it seemed exceedingly like the sun in eclipse, for the beams it shed were not clear.”

      Additionally he describes crop failure due to the loss of solar energy and this carries on for years afterward.

  5. I sent you information on the mini iceage in the 6th century AD. a few weeks ago, hope you read it through.

  6. Heavy snow affects travel in the north of England


    Reader’s comment: Remember that complete nut case climate idiot dr. David Viner who said in 2000: Children just aren’t going to know what snow is?
    This is what one of the greatest climate crooks looks like:

  7. Several cars stuck in Buryatia (Russia) by snowdrifts on the track Gusinoozyorsk – Zakamensk .
    In the Republic of Buryatia today, March 4, due to heavy snowfall was forced to close the traffic on the regional road R-440 “Gusinoozyorsk – Petropavlovka – Zakamensk.” This was reported in the press service of the republican traffic police. On the highway traffic jam, the traffic is complicated by strong winds and snow drifts, whose height – 1.5 meters. The Selenga district put on high alert mode.
    According to the press service of the Ministry for the development of transport, energy and road sector, heavy snowfall in the territory of the passage of the route began in the evening on March 3rd. On the eve of the road workers clear the drifts left in Dzhidinsky District. This morning received alarming signals from the Selenga area. To combat the effects of the elements on the way out 11 emergency vehicles. However, to eliminate drifts road builders can not – push the truck from the roadway snow that the wind pushes back. Currently, works are carried out from 3 to 12 km of the route. The situation on the ground monitor the rescuers, they are ready to assist people who find themselves in a snowdrift. What exactly got in the car, “snowdrift”, not reported. When will be able to eliminate the congestion – is unknown, the press service of the Ministry.
    For weather forecasts, March 5 heavy snow and blizzards persist in the region.

  8. instantly killed up to 100,000 people

    way back then I seriously doubt there were anything like that many people in the area or close by to be killed immediately
    mighta been lucky to have that many people total in the areas a considerable way around it really

    • ” Don’t know much about history……..”
      The city of Teotihuacan in Mexico had a population of between 120000 and 150000 at that time. It’s collapse occurred because of this event. The entire area was covered with large population centers of 10’s of thousands of people each.

  9. Yes, there are records of catastrophic cooling and crop failures in Ireland during this part of the 6th Century. Volcanic eruptions in Indonesia also contributed to the disaster.

  10. Estonian Journal of Archaeology, 2014, 18, 1, 3056 Andres Tvauri doi: 10.3176/arch.2014.1.02 THE IMPACT OF THE CLIMATE CATASTROPHE OF 536537 AD IN ESTONIA AND NEIGHBOURING AREAS In 536541 AD a short-term and sudden cooling took place in the northern hemisphere. Archaeological and palynological data reveal that this event caused crop failure and demographic catastrophe in what is today Estonia. It took until at least the end of the 9th century to return to the previous population level. As a result of this crisis, previous power relations, trade networks, and handicraft traditions were disrupted, settlement structures transformed, and the entire world view of people changed. The climate catastrophe may have influenced the development of agriculture in Estonia. One can assume that the predominance of rye cultivation in Estonia in the second half of the first millennium is a result of the climate anomaly of 536537. In Finland, where hunting and fishing were the main livelihoods and thus land cultivation was of marginal importance, the climate catastrophe does not show in the archaeological record as a decrease of archaeological finds as it does in Estonia or eastern Sweden. One can, though, observe disruptions of the previous trade networks due to the events of 536541 and the formation of new ones. These changes had an important role in forming both the unique material culture of the Finnish Merovingian Period and in expanding land cultivation in the area of southern Finland during the next few hundred years. The impact of the catastrophe of 536 was so substantial that it could be considered an important threshold in the Estonian archaeological chronology. It is then that the greatest cultural upheaval since the major changes between the Early and Late Bronze Age are visible. It would be appropriate to set the border date between the Migration Period and Pre-Viking Age approximately in the year 550. Another advantage of this date is that it allows the Estonian Iron Age periodization to be synchronized with those used in the Scandinavian countries and Finland. Andres Tvauri, University of Tartu, 18 Ülikooli St., 50090 Tartu, Estonia; andres.tvauri@ut.ee

    • The migration of the horse tribes which became the Anglo- Saxons of UK and the Germanic tribes started from east of the Urals at this time. The above would indicate the migration path via the Baltic coast would have been much easier due to the reduced populations of local tribes.
      Solar Minimums coupled with major geological events end and start civilisations, this new civilisation normally last two Solar Warm periods, with each warm period bounded by an MIA or 376 years.
      The West’s civilisation by being a warm water, sea based culture based on the edge of a major continent with a semi permanent, warm air stream flow provided by the Westerlies and the Jet Stream, has for the time being stepped out of this cycle of rise and fall, it is however, still liable to be kicked back into the iron age if it ceases to have access to cheap energy and is led onto the population reduction rocks by the siren voices of “Green” dark side.

  11. There appears to be much talk about a large eruption of Krakatoa erupting in AD 535/536, which either caused, or contributed to the climate cooling in AD 536.

    It seems unlikely that Krakatoa is the culprit of the AD 536 cooling, or rather, one cannot make a definitive ascertain that it erupted in AD 535/6. When we look at the most up-to-date ice core evidence, the AD 536 event does not have a very strong equatorial signal. It looks as if the eruption for AD 536 is a Northern Hemisphere eruption, judging by the large concentration of sulphate in the Greenland NEEM ice core, in comparison to Antarctic ice cores. There is a small amount of sulphate in Antarctic ice cores dated to AD 535/6 but it is 12 times less than that of the signal in Greenland (see Sigl et al 2013).

    Now this does not rule out two or more volcanoes erupting in AD 535/6, with perhaps a small equatorial or southern hemisphere eruption, and a larger northern hemisphere eruption, so Krakatoa could have erupted then, but then again, so could any other equatorial volcano.

    The main issue is though, you cannot identify a volcano simply from its sulphate signal. In 1999, David Keys proposed Krakatoa as the cause of the the AD 536 dry fog. One of teh main reasons for this identification was that he was looking for an equatorial volcano to support the emerging ice core evidence (which we now know was dated incorrectly during this time period). To support his hypothesis he uses historical documents, such as reports of two “thunders” south west of Nanjing in AD 535. he claims that this could be from a massive eruption of Krakatoa. However, Keys does not consider that China has quite a number of active volcano fields, two of which are Tengchong or Nainan Dao volcano field, the latter being south west of Nanjing.

    Another source that Keys uses is the Javanese Annals or Book of Kings. There are quite a few issues with these though, the primary one being that the earliest copy dates from the 19th century. So are they recordings of real events, or just made up? this has been a prime debate by scholars. But lets assume that they are real recordings. One argument for them not being made up is the inclusion of the description of what sounds to be quite a volcanic eruption, indeed it is this description that Keys uses to support his AD 535/6 Krakatoa hypothesis.

    Now teh date given in the Javanese Book of Kings for this event is around AD 416. David Keys is happy to drag this 130 years forward, perhaps due to copying errors, since the books were written on palm leaves which do not survive long in the equatorial climate. But this is a crucial point. It was dragged forward to back up the assertion that the 536 climatic event was due to Krakatoa. Since the physical data in ice cores weakens this hypothesis, then could the date of the Book of Kings event be closer to AD 416? In actual fact, if 416 is indeed a copying error, then it could potentially be any date. Looking at Sigl et al 2013 or 2015, and we see that they give an equatorial eruption around 432 in (the corrected) Greenland and Antarctica ice cores. Why cant this eruption be the eruption recorded in the Book of Kings? To really hit the point home, if we look at Sigl et al (2013) yet again, we find another equatorial eruption around AD 574. If David Keys can move the Book of Kings date forward 130 years, why not move it forward another 40 years? Could AD 574 then be the Javanese chronicler’s event?

    The Smithsonian’s Global Volcanism program lists AD 416 eruption of Krakatoa as a historically recorded eruption, no doubt using the Book of Kings. However there is no sulphate signal in Greenland or Antarctica at this date. So either Krakatoa didn’t erupt in 416, or if it did it might not have been a large enough eruption to deliver sulphate to the ice caps of either hemisphere. Indeed, the Smithsonian records its (Volcanic Explosive Index) VEI as 4, which while large and likely locally devastating, it is not that large in the global scheme of things.

    This is the trouble with invoking copying errors though, they loosen the constraint of variables for dating, and hence it is easy to suck it into any event you like.

    The only true way of identifying the date of an eruption in lieu of a true and well documented eyewitness account of the eruption (such as the AD 79 Vesuvius eruption), is through the identification of tephra in the dated stratigraphy of ice cores. Unfortunately, as of writing, while multiple tephra has been found in the ice cores at AD 536, it has not yet been linked to any volcano. Indeed, some avenues of research suggest that Tephra from equatorial eruptions may find it difficult to reach polar regions to be laid down in ice, making the identification of most equatorial eruptions problematic.

    Finally regarding the culprit of the AD 536 cooling. Current speculation links the cooling to a high latitude northern hemisphere eruption, possibly from an Alaskan volcano, and that this eruption occurred around March AD 536. We know this as this is when the European dry fog/dust veil is first reported. When the 20 metre asteroid exploded 23 km above Chelyabinsk in February 2013, satellites were able to map the distribution of its dust plume as it spread through the stratosphere. What they found was that the distribution of the dust was rapid, taking less than a week for it to circumvent the globe in a belt of stratospheric dust. Given this observation, it would seem unlikely that an eruption of Krakatoa in AD 535 could produce a northern hemispheric dust veil so long after the eruption.

    To summarise though, the eruption of Krakatoa in AD 535/6 is not certain, and remains a hypothesis only. This hypothesis was based on old incorrect dating of sulphate in ice cores. The new dating and identification of bi-hemispheric signals weakens this hypothesis, though one cannot rule it out. The documentary evidence that supports the hypothesis is also weak. What is certain though is that if Krakatoa did erupt in AD 535 or 536, its effects were dwarfed by a much larger climatically effective northern hemisphere eruption.

Comments are closed.