Eruptions to come

Let’s start with a question. Which country do you think has the most frequent volcanic eruptions? Before you read on (or peek below for the answer), take a minute to think about it. You can probably guess that Australia is not a front runner. In fact, only two or three countries readily come to mind. Iceland is up there, with an eruption on average every 3 to 4 years. It is hard to beat. Indonesia is also plausible, which its plethora of active volcanoes. It is responsible for some of the worst eruption of the past millennium: Tambora and Rinjani account for two of the four largest eruptions over this period. But which country is the worst?

Merapi. One to avoid

The previous post triggered some discussion on what next year would bring, in terms of the next VEI-4 eruption. After all, the predictions that were made for 2018 turned out to be pretty insightful. Of course, those predictions weren’t ours – these were professionals sticking out their necks. Can VC do as well?

Let’s start by looking back. What do we know about such eruptions of the recent past? Where were they? The 21st century is still young, but there have been some memorable eruptions. Wikipedia helpfully contains a list of major eruptions. The list defines ‘major’ as VEI-4 or higher; this deliberately excludes effusive eruptions. With the benefit of hindsight, this seems a missed opportunity: how can the volcanic history of the 21st century not include Holuhraun, Pu’u’O’o or Leilani? (And should Pu’u’O’o and Leilani count as one eruption? According to the standard definition, yes. Leilani was very much it own event and I would happily count it as separate, but it doesn’t qualify.)

Another point of contention is when to start. We celebrated the turn of the millennium at the start of Jan 1, 2000. But in fact, the old century didn’t end until 31 December that year. Our time keeping suffered a millennium bug: there never was a year zero, and therefore the century begins at year 1. I will therefore liberate the rebel within and discount eruptions of the year 2000 as so-last-century.

So what was this century like, so far? The first decade was decidedly busy: there were 11 VEI-4 eruptions in the period 2001-2010. The most destructive eruption was Merapi in 2010, and the most expensive Eyjafjallajökull, also 2010. (This was also the world’s first carbon-neutral eruption, as you may recall.) The decadal leader board for 2001-2010 runs 1. PNG (3 eruptions), 2. Indonesia and the USA (2 each), and several countries with 1 including Iceland. So how did you do?

Rabaul, PNG

The second (and current) decade still has 2 years to run (according to my time keeping). In terms of explosive eruptions, it has been quiet. There have been five VEI4+ so far, including Puyehue, which in 2011 suffered the largest explosive eruption so far this century. I’ll add Holuhraun and Kilauea (for now). Over the 21th century, this gives the following order:

  • Chile (3)
  • Iceland (3)
  • Indonesia (3)
  • USA (3)
  • PNG (2)
  • Russia (2)
  • Ecuador (1)
  • Eritrea (1)

Does the appearance of Chile surprise you? Perhaps it shouldn’t. Going back another century, there have been 31 different volcanoes erupting in Chile since 1900. It has about the same frequency of volcanic eruptions as Iceland; admittedly these eruptions come from many more sites and a 7 times larger footprint than Iceland. Cerro Hudson in 1991 reached VEI-5 and in fact was very close to VEI-6, at a DRE of 2.7 km3. And this was the second Chilean eruption of this size in the 20th century: Cerro Azul in 1932 was also a VEI-5 to 6.

But looking harder, it is surprising after all. The list of post-1900 eruptions shows that Indonesia contains more than 60 separate volcanoes that have erupted since 1900, twice as many as Chile. Indonesia has an eruption frequency that is well above Iceland and Chile. Nine or ten Indonesian eruptions reached VEI-4 since 1900, and one of these (Agung) reached VEI-5. Over the same time, Chile had five eruptions of this size. The numbers indicate that Indonesia is twice as volcanically active as Chile. That is still respectable. (One may also query whether some eruptions may have been missed in the inhospitable deserts of Northern Chile or its icy southern regions.)

So how does the 21th century compare so far? The 20th century had up to 65 VEI-4 eruptions (for a few of these it is questionable whether they reached this high). Over 18 years the 21st century has shown 16. (Now I am not counting the two effusive eruptions as the VEI scale does not really apply to them). Scaling from the 20th century, a total of 11-12 would have been expected. So this has indeed been a fairly busy period. In fact, the current decade is pretty much on schedule, whilst the period 2001-2010 was quite over-active.

Can we use this to improve our statistics of nations? To do this, I have combined the lists of the 20th and 21st century, but removed the doubtful cases and also left effusive rift eruptions out. That left me with 75 VEI-4+ eruptions since 1901. They are distributed over the decade as follows:

  • 1901-1910: 5
  • 1911-1920: 9
  • 1921-1930: 3
  • 1931-1940: 6
  • 1941-1950: 4
  • 1951-1960: 6
  • 1961-1970: 6
  • 1971-1980: 5
  • 1981-1990: 9
  • 1991-2000: 6
  • 2001-2010: 11
  • 2011-2018: 5

In this list, two decades stand out: the 1980’s and the 2000’s. In contrast, the current decade is not exceptional. (But this discounts the two effusive eruptions so it is not the full story!) Overall, we can expect a VEI-4+ eruption about once every 1.5 years, or 6 per decade. There was one exceptionally long period without such eruptions, from 1933 to 1945. This is likely a political gap: the world was too busy with nationalism to be concerned with volcanology. (There was an eruption from Rabaul within this period, but it was not documented well enough to qualify). Apart from this, the longest wait was 5 years: this occurred twice.

So going back to the original question, which countries stand out? The list below is the answer. Chile is now number 5, and Iceland number 6. Top of the list is Russia, Indonesia is second and third is the USA. Does this surprise you? I must admit that when I picked countries for the next VEI-4, I missed three of the countries in the top four. Ouch.

  1. Russia: 16
  2. Indonesia: 12
  3. USA: 11
  4. PNG: 8
  5. Chile: 7
  6. Iceland: 5
  7. Ecuador, Guatemala: 3
  8. France, Japan, Philippines, Mexico: 2
  9. Eritrea, Vanuatu: 1

This needs a bit of explanation. The numbers for Russia contain 6 eruptions within the chain of the Kuril islands, possible the biggest concentration of events in a small area. The ownership of these islands is disputed, as they are also claimed by Japan. The US numbers are boosted by eruptions in the Northern Mariana islands. Ecuador includes the Galapagos, and France is responsible for all VEI-4 eruptions in the Caribbean.

In the comments on the previous post, a number of ideas were floated for which volcanoes could cause a VEI-4+ eruption within the next year (or two). The following volcanoes were mentioned:

Iceland: Grimsvotn, Thordarhyna, Oraefajokull, Katla, Askja
PNG: Manam, Rabaul
USA: Kilauea, Augustine
Indonesia: Merapi, Sinabung
Guatemala: Fuego
Chile: Nevados de Chillan, Puyehuhe-Cordon Caulle
Mexico: Colima, Popocatepetl
Japan: Sakurajima
Russia: Shiveluch, Avachinsky, Kliuchevskoi
Ecuador: Cotopaxi

The top-8 of the country list are all represented. On the other hand, France, Vanuatu and Eritrea are not. The most notable absence is perhaps the Philippines, home of the largest eruption of the past 130 years but not seen as a likely contributor over the next two years.

So after this, what do I think myself? My opinions have changed. Based on the numbers, Russia is the top candidate, and especially the region of the Kuril islands. But Iceland should not be ignored, especially since Oraefajokull is building up to activity and it has a VEI-4 history.

And which particular volcano? That is a very long shot. But it is worth noting that VEI-4 eruptions can be a repeat offence: such an explosion is not strong enough to fully disarm a volcano, and a repeat occurrence is possible. In the list of eruptions several volcanoes indeed appear more than once: Avachinsky, Mount Lolobau, Kelud, Augustine, Rabaul. Therefore, It is worth keeping an eye on those volcanoes who have done this before.

And here I’ll end and leave the conclusion to the commenters. Did you change your mind? Where do you think the next event will be? Or do you think that the 50% chance of no VEI-4 eruption in the next 1.5 years will come up trumps, and we will have two years of volcanic solitude? Or do you think this is meaningless if the large rift eruptions are not included? DO you think the wikipedia lists are wrong or incomplete? (And yes, I am aware that the VEI’s listed in the wikipedia list for the 20th century are on the high side.) We are looking forward to your thoughts!

Urup island, Kurils. Not likely to explode again, but it is a sign of eruptions past.

Albert, December 2018

120 thoughts on “Eruptions to come

  1. Urup island island photo is awesome! Never seen nor heard of it before. Looks like a mini Devil’s Tower in the middle of the ocean.

  2. Big effusive eruptions should absolutely be added, but only if they exceed maybe 50 m3/s for longer than a day, meaning leilani is an easy pass but pu’u o’o and erta ales new eruption are not. An alternative definition could be an eruption rate maximum of over 1000 m3/s, which was achieved by wolf and hekla pretty easily (wolf actually exceeded this by an order of magnitude on the first day of eruption).

    Just yesterday actually, I looked at the exact Wikipedia list and was very tempted to add in leilani and holuhraun. It includes nyiragongo 2002, a comparitively tiny eruption which was only notable because of its fatality list. Also ontake 2013 was nowhere near even a VEI 2, in fact technically it wasn’t a real eruption.

    If I could add to that list and I really wish I could, I would add every effusive eruption since 2000 to exceed 0.1 km3 in less than a month – hekla 2000, sierra negra 2005, piton de la fournaise 2007, alu dalafilla 2008, nyamuragira 2010 and 2011-2012, tolbachik 2012-2013, bardarbunga 2014-2015, wolf 2015, kilauea 2018 and sierra negra 2018, with special mention to kilauea for also erupting about equivalent to 0.2 km3 every year this century, equivalent to a tephra volume of 1.2 km3 every 2 years, or a small VEI 6 every 5 years…

    I would also like to add that wolf 2015 and tolbachik 2012-2013 are both listed as actual VEI 4s by GVP, their eruptions were pretty violent at the start, the tephra from both of them reached 15 km high. The size of an eruption should be judged equally by intensity rather than only volume, a long lived eruption like pu’u o’o or the Iceland shields will have enormous volume but very low intensity, while even a VEI 2 might be stratospheric if it is a single explosion.

    • We really need something to replace the simple VEI rating system, which needs all manner of contortions to accommodate primarily effusive eruptions, because the VEI was originally proposed as a way to improve on Lamb’s Dust Veil Index – both the DVI and VEI as constituted were looking at the injection of volcanic dust into the upper atmosphere -and, therefore, primarily concerned with violent EXPLOSIVE events (hence the name), Maybe a scale based on the total amount of thermal energy released could be devised, although the VEI would still be useful as a measure of violence in a particular eruption, its original function, after all

      • Thermal energy of the magma, added to the explosive power if there is any, this still means VEI 7-8 calderas are among the biggest, but it also fairly places large flood basalt eruptions where they should be, squarely at the top with a long way to second…

        1991 is sometimes noted as being a particularly notable year because of it’s almost double VEI 6, but individually both holuhraun and leilani involved much more total energy than pinatubo and hudson put together, both were equivalent to a VEI 6 in volume converted to tephra, and a mid sized VEI 7 on thermal energy.
        It is through both these methods that I have come to be so humbled by effusive volcanism, but especially kilauea, even before this year the only year since 1983 that pu’u o’o was actually beaten in total yearly lava output was when holuhraun was active (year of June 2014-June 2015), every year of pu’u o’o was almost a VEI 5 in tephra volume equivalent (0.6-0.9 km3 tephra, 0.2-0.3 DRE). If all of pu’u o’o was translated to tephra volume it would be 25 km3 of tephra, large VEI 6 almost 7. If all of kilaueas historical activity was put into one eruption and translated to tephra, it would be about 50-70 km3, a solid VEI 7, and that is only in the past 300 years. Not many volcanoes could potentially do a VEI 7 every 300 years…

        Basically to put it into the VEI scale, since 1983 kilauea has had at least 30 VEI 4s and 1 VEI 6, and since 1790, it has had a VEI 5, followed by an actual explosive VEI 4+ less than 2 years later, then 10-15more of those in the next 30 years, then equivalent of VEI 5 every gear of crater fill, then a borderline VEI 5, still managing about VEI 4 per year in tephra equivalent until 1924, only being below this for 38 years, and then after 1952 up to 1960 is 4 VEI 4s, VEI 4 in 1965, VEI 4 every year after 1968 up to 1974, VEI 4 in 1977, then the about 30 almost VEI 5s of pu’u o’o, then the VEI 6 of this year, and only months later there is signs of recovery. This also does not include intrusions which were frequent 1840-1952 and 1975-1983 and some of them were pretty big.
        Doing it this way really puts things into the bigger picture, and does it fairly. Suddenly the ‘big’ eruptions don’t seem so impressive anymore if they take out the entire volcano too, especially when kilauea plays with large eruptions and throws VEI 5s for fun.

        Just a thought, kilaueas new caldera is as wide as pinatubos caldera, maybe not entirely a coincidence…

        • Another way would be to go off average eruption rate, some effusive eruptions will generate little tephra but erupt at a rate similar to a plinian eruption, skaftar fires after its initial explosive stage would have still been going at that rate but the vent would be wide enough to allow effusive eruption, like a way bigger version of holuhraun or fissure 8. Flood basalts could likely exceed the eruption rate of anything short of a large scale ignimbrite sheet event (which is kind of like the silicic equivalent I guess).

  3. No thinking needed. Indonesia or Iceland. Maybe somewhere in Africa since they have that whole rift valley thing going on… but as noted in a segment of “Whose Line”, Africa is a continent, not a country.

  4. How is our old Bolivian pal, Uturuncu doing? I would never consider it a candidate for the very near future because it is likely to be very many years before it becomes dangerous. But looking at these volcanoes (thank you for the article Albert!) put me in mind of it.

  5. Chile gets my vote. All the mega-thrust quakes since 1960 have been moving a lot material/crust under the entire coast….and it’s more then reasonable that many Chilean volcanoes will be/are feeling the increased pressure/magma feed.

    • from eastern Java to New Guniea this subduction zone has had a lot of earthquakes over the last few years and is due for a major eruption iin one of the volcanoes in this area

  6. Ref the smiling volcano in the article. Years ago, a good friend of mine related a story about winding up in a town with a lot of time to kill. He wandered into a movie theater and saw some movie that had a volcano/mountain in it with an odd song playing along with it. The movie was Dunderklumpen! and he was suitably freaked out about it. (though not at badly as the old black and white movie “Freaks” which featured an EXTREMLY disturbing imagery such as armless and legless man with a knife held in his teeth crawling through the mud under a wagon perusing an antagonist.) Anyway, the still image above is from a Pixar short named Lava. It is inspired by the Lōʻihi Seamount and the island of Hawaii. Not a bad short at all.

    • I remember Dunderklumpen as quite enjoyable, but didn’t recollect any volcano! Volcanoes in Sweden are rather thin on the ground.

      • Apart from the giant “super-volcano” under Lake Dellen. The one Carl wrote about on 1 April 2014. Caused a bit of a fuss, that volcano.

        • Yes, I know. That was the one that caused the Bardarbunga eruption. Hell of a long dike.

      • Well, it may have just been a mountain. The combination of it and the song is what stuck in his mind.

    • It would be nice if they had this level of data from other subduction zones. It’s like seeing one giant green squirrel with buck teeth and assuming they are all like that.

      Yeah, I know Scrat isn’t green, but he is one of my favorite animated characters. He epitomizes the plight of most people in dealing with life. And while Scrat is just a fictionalized creature, Cronopio dentiacutus was an actual thing. It’s a pretty decent morphological analogue of Scrat… though I imagine that it’s behavior was closer to squirrels and shrews rather than Wile E Coyote.

    • Yeah some suggestions of a landslide- no earthquake reported. It was having a paroxysm so it is plausible. If this is the case it’s quite a reminder of its past! (Cue the tabloids bleating on about how it is going to repeat it all over again).

      • I got to agree with the tabloids though. With an active volcanic island closeby one has to take events like this into account. Even if its not on the same scale as the big one, smaller events like this will happen again.

    • This is telling and as noted by mjf, quite plausible

      “It says the possible cause of the tsunami were undersea landslides after the Krakatoa volcano erupted.”

      Anak Krakatoa is a fast growing volcano. High growth rate = steep deposition slopes. Steep slopes = Unstable material prone to mass wasting events.

      Wikipedia places the summit at 813 meters and notes; “Eventually a fourth island named Anak Krakatau broke water in August 1930”

      That yields a growth rate of about 9.24 meters per year, or 2.53 cm/day on average, mostly from tephra deposition.


    Not new (obviously) but I just saw this video taken when kilaueas eruption was starting to ramp up. Start of fissure 21, not every day you see strombolian eruptions out of a patch of trees in an inhabited area… It is this sort of novelty of a volcano in your back yard (literally) that made the eruption so interesting, even if it was to be expected.
    This was at the same time fissure 17 a few km away started going crazy and larger flows started from 16, 18 and 20.

  8. It is a disaster. This has been a terrible year. I have not seen evidence that there was an eruption: there was no earthquake (but USGS is of course closed thanks to the shutdown so their lists may not be complete). This was a collapse of the slope, I think, but not explosive. Anyer was the same town that was completely destroyed in 1883. This was obviously not as severe: the wave height is reported at 4 meters. But it seems to have arrived as a wall of water, not the slower build up seen in some tsunami.

    • There was a 5.0 with a depth of 5km in the Sunda Strait about 35 minutes before the tsunami (if I have got the time differences correct) some distance from Krakatoa. Probably not the cause of the tsumani, itself, (the area gets a lot of earthquakes that do not cause reported tsunamis) but may have contributed to an underwater landslip.

      • Perhaps the ash column collapsed? Dumping a lot of ash at once into the sea?

      • “Oystein Lund Andersen, Norwegian volcano photographer, Anyer Beach in West Java

        I was on the beach. I was alone, my family were sleeping in a room.

        I was trying to photograph the erupting Krakatau volcano.

        Earlier in the evening, there was quite heavy eruption activity. But just prior to the waves hitting the beach, there was no activity at all. It was just dark out there.

        And suddenly I saw this wave coming, and I had to run.”

        — Video posted downthread

        • That was the first smaller wave which he could outrun. The second was larger and hit when he had reached the temporary safety of his hotel(?) and was waking his family to get to safety..

  9. One thing I have learnt from here, especially after this year, is that if any volcano with a history of erupting big shows activity, watch it. Another is that it generally takes a long time to prepare for an eruption of large proportions, but that the event itself is very fast. The 3rd is that effusive volcanism is diabolically underestimated, and that any basaltic volcano with a high magma column should be watched very closely, especially if it is inhabited. In just the last 20 years no less than 5 basaltic volcanoes have undergone a deep draining event, all but one were caused by large eruptions, with 2 of them being eruptions of global significance. Ambrym might have very recently made that 6 but that has yet to be determined.

    Maybe the most important thing though, is the sheer unpredictability of the most predictable volcanoes… Every volcano is unpredictable, but some are more watched than others. A volcano in Alaska might be forever surprising, there is little monitoring equipment there so an eruption might be noticed only after it has already started. But kilauea is seen by millions every day one way or another, and was considered to be in the most stable situation a volcano can be, a fully open conduit feeding the magma direct to the surface, a literal hole in the earths crust. It had worked for 30 years, might have lasted hundreds of years like this before, and pu’u o’o had already survived several attempts at destroying it before, yet just as an expected part of the cycle began, it went off character, the lava lake overflowed vigorously at 10 times the eruption rate of any point in the decade before, and within a week it threw 30 years of clockwork eruptions out the window in a single eventful night, and went to hell 18 days later. When it was done 4 months later the worlds perception of volcanism was changed forever, as was kilaueas reputation, kilauea is not just a ‘safe’ volcano with little slow eruptions sitting under its massive neighbor anymore.

    Now 2018, the year of lava, is nearly over, 2019 might not see much lava, the least lava in a long time in fact, but with so many awakening long dormant volcanoes, it is sure to be interesting. And who knows, with the two most powerful hotspots currently surging, anything can happen really.

    • None I know off in the way are used to. Though the volcano is close enough for lots of amatour videos to exist.

  10. All the volcanic build up recently fits the theory of deep solar minimums triggering (higher) volcanic activity than in (solar) active periods. The mechanism seems to be connected to increased cosmic radiation penetrating our atmosphere with groundpentetrating energies due to a highly reduced magnetic shielding during deep minimums. We are currently lower than Dalton…

    But hey! Dismiss it and stay notoriously conservative untill thoroughly observed sometime through 2021-22. Then we (if correct) won’t be able to dicuss it online anyways if we see a decent VEI-6. At least for some time. To me the theory is obvious, logical and… in the provingphase.

    Darwin has confirmed Krakatau to FL550 using Himawari-8 as source of info. Will be interesting to see if this will be a lasting event. Plenty of videoes of the initial blast online.

    Stay safe. And Merry Christmas to everyone!

      • Would you care to elaborate on that? I stick to published theories. With CERN stating there are (potenially) millions of unexplored energenic particles with the potential to penetrate and even go through our planet. Not me. Svensmark and others links cosmic radiation to the same, not me. Bartol meassures that cosmic radiation INcreases during solar minimums. And that it is close to breaking space age records. Not me.

        And the papers stating there IS a statistical significant higher rate of mid VEI5 -> eruptions during deeper solar minimums are plentyful accessible online. And particles +2,5ppm will block both sunlight and satellitecommunications. I don’t make this up. Check with the European Space Agency.

        So. Exactly what makes this THEORY psuedoscience? That you don’t like it? That it is not consensus?

        If we didn’t stay dynamic in exploring new links and theories we would be a poor bunch of sods.

        • I regretfully have to agree with mjf here. The theory that cosmic rays are involved with volcanic eruptions is based on two studies. One is an old one by Stothers who found a weak periodicity in eruptions over the past few hundred year. The second is an investigation of 11 japanese volcanoes. Both papers are flawed in their math, obviously so for the japanese volcanoes, and a more subtle issue in the Stothers paper. And if the math error isn’t enough, the claim that a 5% variation in cosmic rays that penetrate no deeper than 10 meters can set off a big volcanic explosion should raise warning signs. Not all papers that are published are correct. In these cases you look for three things: (1) is the paper correct; (2) have other studies reproduced their findings; (3) does it make predictions which have been confirmed. These two papers fail on all accounts. But I am aware that this one has fooled more people.

          You are right that scientists should be open minded, and we normally are. But a claim that is based on flawed statistics, has no plausible physics behind it and has not been confirmed requires more than an open mind. I think you see the scientific process in action here, and this claim has not survived it.

          Is it pseudo science? That is an emotional term, and this argument was in fact triggered by real papers. I would say that the Stothers result was a valid discussion. The study of the japanese volcanoes was so poor that it should never have been published in a reputable journal and it does fall short of our standards. It was in fact a decent journal which published it, and I wonder whether the publicity value overruled the weakness of the science. It sometimes happens.

          I suggests you apply the three rules above and judge for yourself.


    Øystein L. Andersen
    ‏ @OysteinLAnderse
    7h7 hours ago

    Just completed a hi-speed video of yesterdays (22th December) eruptive activity at #Krakatau volcano. A new lava-flow can be seen descending to left of the island. Filmed 47km away from it, only a few hours prior to the #tsunami hitting the coast of Java. @id_magma @infoBMKG

    • Checkout the video to the left of the cone. There is a significant amount of gas / steam or ash rising from something just out of the frame of the video.

  12. This is the map of affected area. Anak Krakatau is the white island in the centre of the three islands (which together made up old Krakatoa). The tsunami hit Lampung, at the end of a bay which pointed directly at Krakatau. Sitting duck comes to mind: everything seems designed to focus the water here. On the other side, Anyer and Carita are southwest of Cilegon, where the coast is closest to Krakatoa. The three surrounding islands would have given some shielding here. Lampung was not shielded in any way. From the map, it is not clear on which side of Anak Karkatau the collapse occured. If on the west side, most of the tsunami may have been directed towards the open (Indian) ocean. If on the southeast side, both affected areas could have been directly in the firing line. I would also expect significant damage ot the three surrounding islands (unoccupied, I believe) and to the island of Sebesi.

    The M5 earthquake was between Karaktoa and Anyer and does not appear connected.

    • A satellite image taken last september. Click fo full resolution. The flank collapse is where the cloud is in the image. In the top right is part of the coast that was hit by the tsunami.

  13. By the sound of it, Ujung Kulon National Park may be hit hard, the only remaining habitat of the critically endangered Javan Rhino. If only a few of them have died it may spell disaster for the species.

  14. Latest update dated a few minutes ago

    DTG: 20181223/1400Z
    PSN: S0606 E10525
    ADVISORY NR: 2018/371
    TO N
    OBS VA DTG: 23/1400Z
    OBS VA CLD: SFC/FL250 S0608 E10532 – S0554 E10527 – S0559
    E10455 – S0544 E10445 – S0523 E10534 – S0544 E10539 MOV N
    10KT SFC/FL550 S0609 E10531 – S0629 E10514 – S0718 E10329 –
    S0846 E10145 – S0806 E10034 – S0736 E10100 – S0611 E10326 –
    S0554 E10527 MOV SW 30KT
    FCST VA CLD +6 HR: 23/2000Z SFC/FL250 S0609 E10532 – S0554
    E10527 – S0559 E10459 – S0536 E10440 – S0506 E10531 – S0544
    E10541 SFC/FL550 S0609 E10532 – S0725 E10438 – S0914 E10146
    – S0837 E10030 – S0757 E10026 – S0615 E10330 – S0554 E10527
    FCST VA CLD +12 HR: 24/0200Z SFC/FL250 S0608 E10532 – S0554
    E10528 – S0602 E10501 – S0531 E10429 – S0454 E10540 – S0550
    E10545 SFC/FL550 S0608 E10532 – S0729 E10440 – S0943 E10138
    – S0841 E10004 – S0754 E10006 – S0645 E10229 – S0554 E10528
    FCST VA CLD +18 HR: 24/0800Z SFC/FL250 S0608 E10531 – S0554
    E10528 – S0557 E10454 – S0531 E10429 – S0454 E10540 – S0551
    E10545 SFC/FL550 S0608 E10531 – S0731 E10439 – S0942 E10138
    – S0841 E10004 – S0755 E10006 – S0605 E10322 – S0554 E10528
    NXT ADVISORY: NO LATER THAN 20181223/2000Z

  15. Karangetang raised to orange alert due to heavy seismic activity.

  16. Based on that VAAC, and what archival data of them I could find, it’s been maintaining the FL550 plume ever since it started.

    Remembering this is a marine environment, and water interaction should be factored into any analysis.
    For now, I get a VEI-4, It will not reach VEI-5 unless it maintains this plume more than 50 hours. For Anak Krakatau, FL550 is about 5517.40m³/s.

    Important. The source document for that calculation (Mastin et al) notes that it is accurate to within a factor of 4. In other words, it’s just a ball-park figure and could be quite different than what the officially tallied ejecta calculation turns out to be.

    I don’t have the current tropopause height, but based on the Annual averages, it’s getting pretty close to it for that latitude.

    • A quick question on the FL-scale; volcanodiscovery states that FL550 is (of course) 55,000 feet. But. Also states this equals to 15 km. height.

      When using standard calculation from feet to metres i get 16,764 km. The FL-scale also gives this height in meters/km.

      Which is the correct?

      All the best from Norway 😉

      • The Fl scale comes from airplanes. It stands for ‘flight level’. It is measured in 100 feet above sea level, so Fl550 is 55,000 feet which is a bit over 15 km. I think in this case it is meant as an upper limit: it won’t be any higher but airplanes flying lower than this are at risk of encountering ash. However, it also states that in his case no ash has been detected so there is some confusion.

        • Krakatau was in the middle of a big paroxysm when this happened, so there must have been quite a bit of ash. Maybe not at higher altitudes. Would it be possible that if the event happened so quickly, the magma got in contact with lots of magma and produced surtseyan wet ash almost immediately and therefore the ash didn’t rise to very high altitudes? Kind of was the same at bogoslof, the upper few kilometers of the eruption plumes were mainly steam with only minute amounts of ash.

  17. This was published by volcanodiscovery, two sentinel radar images. As they point out, it leaves little doubt about what has happened. The southwest slope where the lava had bene flowing for months slid into the sea, taking half the crater with it. That will have left an unstable slope and more (smaller) collapses are possible. You can see the waves on the lower image, presumably from on-going slides. They go to the southwest, suggesting that Anyer was hit by waves reflecting off Sertung Island. The main collapse must have been quite sudden to displace so much water.

    It is possible that the collapse started at sea level where lava flows can build up very unstable flanks. This is similar to Kilauea, although there the mountain slope itself is more stable so only the coast lava plain is at risk.

    Note that the eastern tip is also gone, perhaps a victim of the tsunami.

    • I guess the question now is how stable is the remaining edifice? & is sea water likely to get into the magma chamber?

  18. This situation is very concerning with phreatomagmatic interactions ongoing along with unstable slopes both above and below sea level. This situation needs to be very closely monitored to avoid further loss of life.

  19. To be completely honest, I have no idea about the sulphur load of this event. Indonesian volcanoes tend to be on the high side of SO2 loads, but with greater water interaction, both at the surface and due to tropical humidity, the SO2 has a greater leaching/conversation to sulphate rate than high latitudes. Given that, the stratospheric loading, if any, will convert to sulphate in about 2 months. In about 50 months any sulphate generated from this event will return to pre eruption levels.

    • I guess the big thing to look for, are ameseth skies near sunset.

  20. Downside, it had an extensive, rather unexpected flank collapse.
    Upside, it does not seem to have followed that with a St Helens’ lateral blast…
    Against that, even without an accompanying quake, such a blast might have been noticed…

      • It was having an ongoing Stombolian eruption in the run up to when the tsunami was spawned –>

    • I dont know. The flank that collapsed pointed away from any inhabited areas and the collapse itself could have occured underwater.
      I think anyone in hearshot of the blast may have mistook it for the normal rumbling of the volcano or would have been much to close to be able to report it in.

      I think a Lateral eruption occured on Anak. Though not on the scale of Helens.

      • The St Helens lateral eruption was triggered by the flank collapse. The internal pressure was of course behind the flank collapse but the immediate cause was an earthquake, and the sudden loss of weight allowed the explosion. Krakatau was already erupting and so there was much less pressure inside. Apparently this allowed the side to collapse without a serious explosion to follow (a minor explosion cannot be excluded). Note that an explosion would not have been heard 10 km away: there is a distance range which acts as a sound shadow . That was also the case in St Helens where people within eye sight heard nothing.

        • If the collapse cut off the summit right away then you are suddenly removing the uppermost part of the magma column and the reservoir is overpressurized by moments, an intensification in the eruption rate, bigger or smaller, should in that case be expected. The interaction with water should also result in increased explosivity. I assume that the 16 km tall plume formed following the flank collapse, can that be confirmed? And I also find more likely that the plume was mostly steam from the surtseyan activity as others have already suggested.

        • But what about the original Krakatoa eruption explosion that was heard 3,000 miles away?

          • The explosion was heard nearby, not 10s of kilometers away, but again a few 100 km away, loud enough to keep people awake. It is actually the same with lightning: you can’t hear thunder from a storm at the horizon but do hear it closer and further away. The more distance places hear it by reflection of sound waves off the stratosphere. It is discussed somewhere in the St Helens post.


  21. I’m a bit surprised the event at Krakatau isnt being talked about much here. :p

      • I could point at Kilauea which is hardly eurocentric and had the most posts this year. This year, there have been 8 posts on Europe and over 40 on other things or locations, as far as Ceres. Volcanoes in the western world tend to be better documented and we do rely on what information is readily available or published in scientific papers.

        As for Krakatau, we tend to wait until enough is known, as VC aims to be informative rather than sensational. The comment section is recommended for on-going events: the commenters do a fantastic job.

  22. Looking at the Sentinel radar image, it seems Sertung Island, west of Anak Krakatau, may have lost a peninsula. Comparing old and new, the northeastern tip of the island seems to have disappeared.Radar can see different things than photos and this needs confirming.

    • The upper part of Sertung Island is made up of a thick (c 60m?) deposit of mostly unwelded pyroclastic flow material from 1883, maybe it would have little resistance to the impact of a forceful tsunami

    • The tip may be covered in water from the tsunami wave at the time of the 23 December photo – you can see the ripples of a wave in the lower left of the image.

      Be interesting to see what shape the islands are when this settles and we can get daylight images.

  23. It looks almost certain that the Krakatoa eruption caused the tsunami, probably by a submarine landslide.

    What are the odds of a new VEI6-7 event in Krakatoa?

      • It’s had two already; separated in time by, possibly, some 20,000 years, so give it a few millennia to cultivate a nicely gassy silicic soup then stand well clear.

        • The last caldera collapse before 1883 was circa 416 CE. In between these huge eruptions, three new volcanoes emerged before merging to form the Krakatau island that eventually blew up.

          • The 416 CE event is a popular myth, I’m afraid. It is one of those things that everyone knows but never happened. The area around Krakatoa shows no deposits dated to that period. Clearly there was a large caldera already before 1883, and the previous Krakatoa island had grown uo inside it. But that caldera must be much older.

            The data of 416 comes from a javanese legend but it does not name krakatoa. The corroborating evidence which it would need has not been found.

    • My feeling is that the tsunami was triggered by a sudden slumping of the flank, producing a landslide which carried away much of the lava flow cover on the upper flank as well as most of the summit cinder cone. Remembering that the lower flanks are likely mostly tephra at the angle of repose, so repeated lava flows on the upper part of the island will have been making it increasingly ‘top heavy’ and structurally unstable. Perhaps that nearby earthquake was the final factor to set off the slump? So, you could argue that the (ongoing for some time, I believe) eruption was not itself responsible.

  24. I think extremely low. As I understand it a VEI 6-7 in subduction arch volcanoes is a once in its lifetime event, or it takes a lot of time to get back to being able to produce one.

    • Some of the Sunda arc volcanoes seem to go caldera multiple times. Batur for example.

  25. 6.5 Tonga area

    116 PM HST SUN DEC 23 2018





    ORIGIN TIME – 0109 PM HST 23 DEC 2018



  26. NASA tv is currently still running Apollo 8 50 year anniversary programming although not on all platforms. NASA employees are apparently able to continue to work unfunded until 26th December due to a legal loophole which means the Furlough notice is legally served on Wednesday due to Monday and Tuesday being Federal holidays (Saturday and Sunday don’t count) and there wasn’t enough time to serve it on Friday.

    After the 26th NASA employees are not permitted to work voluntarily so only a few critical personnel remain.

    Insight on Mars has now deployed the seismometer

    NASA InSight
    ‏Verified account @NASAInSight
    Dec 20

    In case you missed it, here’s one for the history books: for the first time ever, I’ve placed a seismometer on the surface of #Mars! Once it’s all set up, I can start listening for marsquakes. More:

    • How long now? It started yesterday, and I saw you say that a VEI 5 is about two days at current rate.

      I think kilauea will still win Erik’s annual Pliny competition, it took up a third of the years headlines among many other things, but it has been a real race to the top right at the very end, and in both cases were from well known volcanoes breaking their long term trends.

      Pretty incredible century so far, there have been as many major eruptions as there have been years, and now one more for good luck.

      • I would certainly be surprised if it makes it to 5, I dont even think it will reach VEI 4. Take into account that cumulonimbus clouds reach this height or nearly this height without the need of volcanic processes in that part of the world. And this plume is mostly steam…

  27. I would expect that 55K cloud has relatively little ash in it, skewing the VEI score to a large degree.

  28. with the small stratocone reomved .. seawater can easly enter the conduit and thats why we are getting surtseyan eruptions.. at these rates the island will quickly revuild itself

  29. To put the current event into perspective, I took the satellite image Albert posted and marked the outline of the 1883 event in red and (approximately) the current collapse in yellow.

    This is a significant event but, unlike what people in social media think, there is still a long way to go to make it a repeat of 1883.

    • I’d suspect that the tsunami damage could have been considerably worse if it was the northern side of the island that collapsed.

    • I may have gone a bit wild with the red paint. At least the two top parts (islands Verlaten and Lang ) are remnants from an older caldera. The 1883 caldera stretches approximately from the bottom red marking up to Anak Krakatau and then a bit further west under water, to the south of Verlaten island.

      Point is still valid though – 1883 was a much bigger event than the current one.

      • The1883 event was larger because, it is thought, that sea water got into the magma chamber.

  30. How strong is the roof of the magma chamber? I read recently the 1883 eruption was very explosive because the roof fell in and seawater got into the magma chamber.

    Anak Krakatau is a very new island, less than 100 years old.

    What happens when a caldera-forming event interacts with water?

    Viewed from the air, Santorini and the islands formerly known as Krakatau bear a certain resemblance.

  31. … and still going;

    DTG: 20181224/0841Z
    PSN: S0606 E10525
    ADVISORY NR: 2018/377

    At this time, it is probably very safe to conclude that most of the plume we are seeing is made up of water vapor. (see “Damon Hynes” at 4:46 above) If so, that would mean that this plot is probably an over estimate of current and projected VEI. The solid line indicates projected current VEI based on Darwin VAAC reports, the dotted line is if you carry that FL550 rate out into the future. (again, FL550 is 5517.40 m³/s for Anak’s 813 meter summit, which is probably gone now. For a summit of 400 meters the DRE for FL550 would be 6134.78 m³/s)

    Remember, that intense tropical thunder storms can easily get to the troposphere with their anvil tops where they flare out into a high cirrus outflow. The idea that this could be mostly moisture is not a far fetched idea. One aspect that I haven’t put much thought into, is that that plume could contain a lot of HCL from the seawater-magma interaction.

    • I find some conflicting figures about the height of Anak Krakatau before the current collapse, but I think that 813m is the height of 1883 Krakatau, before the big eruption. One figure puts Anak Krakatau at 400m in 2017. Other figures are 6.8m yearly growth since 1950s, which also puts it a bit above 400m.

  32. So far in my lifetime, and assuming everyone elses’, there’s only been one colassal VEI 6 eruption (Pinatubo in 1991). I remember it well, the brilliant sunsets, almost a blood red and a thin brilliant red line at sunset too. I’m interested in the colassal and mega colossal VEI-6-8, becauee or their magnitude, historical significance, effects on our species, and their relative rarity compared to lower grade eruptions. Before Pinatubo, the last VEI 6 was Cerro Azul in 1932 or Katmai in 1912? So it’s safe to assume we will probably have more than one VEI 6 eruptions in this century. Where? No idea. I’d bet in Indonesia or maybe Pacific Islands, but that’s just pure guesswork on my part. I’m not sure many in our modern world can fathom how bad a VEI 7 eruption would be. It would cause major climatic effects and disrupt air travel and due to the increasing population clusters in certain areas depending on where the eruption occurred it could have catastrophic local effects not to mention the long term effects. I’m not sure when we will see another VEI 7. To my knowledge, before Tambora, assuming the mystery 1808 eruption wasn’t VEI 7, the prior one was Samalas in 1257, a gap of hundreds of years. Either way, whenever the next VEI 6 or 7 occurs, it will probably be at a volcano long dormant, almost forgotten. Volcanoes that haven’t erupted in hundreds or thousands of years tend to go off violently when they do. So the next colossal eruption will probably happen at a volcano with few past eruptions.

    rescued from the depths of akismet

    • The 1453 eruption was probably also VEI-7. The source volcano is not known but the climate effect was in the range of such eruptions. We estimate that there is about one such eruption every 300 years.

      • Oh yeah I forgot about that one! Wasn’t that Kuwae? Is it possible that maybe that eruption along with the mystery 1808 one occurred on islands and just completely blew themselves apart leaving no or little trace, therefore making them harder to trace? I’d bet the next colossal eruption will be at a volcano with no historical record of eruption, that has very infrequent violent eruptions. It will probably be one so obscure that will rumble to life and have a cataclysmic eruption. Pinatubo was dormant for hundreds of years and Tambora thousands. So the next VEI 7 will probably be a couple hundred years from now. But we will probably see a Pinatubo or Krakatoa size eruption this century.

Comments are closed.