A Super-pooping Supervolcano?

Image from Wikimedia commons.

Image from Wikimedia commons.

Though I walk through the valley of the shadow of death, I shall fear no evil

It is said that within great beauty lies grave danger and few places are more beautiful than Lake Atitlán. Poets have filled cheap notebooks with particularly flowery hyperbole in their attempts to catch the beauty of the caldera lake and the encircling volcanoes roaring out of the lake mist, catching fire as the morning sun hits them turning the green into imaginary gold and fiery red. It is truly a place of dreams, and as with every dream there is a spot of darkness hinting at a lack of a sign: Beware! Here be Dragons.

I rarely write about supervolcanoes. To date I have only covered two real supervolcanoes. The first was the perfectly balanced volcanic system of Tondano and the second was the rather comatose Yellowstone. I have though on occasion mentioned that some volcanoes have better PR agents than others. Where Yellowstone in its dying shroud sprouts a scientific volcanic paper a month, the rest of them will be happy to merit one per year for all of the known supervolcanoes.

It is sad that things are like this, but it is in the end hard cold political facts running science. Yellowstone will get you money to write a pointless paper and it is all about the number of papers, not the quality or merit of research that runs academia in the twenty-first century.

In the end it might not matter that much, the main reason I do not write about supervolcanoes is that they are such a boring bunch. After all, they very rarely do anything interesting, and when they do they either suffer a side vent eruption like Tondano, calmly inflate like Uturunku, or do a bradyseism and goes back to sleep. Or they do nothing and get papers written.

So let us get back to the matter at hand, the rather interesting Lake Atitlán.

A brief background

Who put a small roof in the lake? Photograph by Dr Carmen Morataya.

Who put a small roof in the lake? Photograph by Dr Carmen Morataya.

Lake Atitlán as we see it today is the result of activity after the 300 cubic kilometer Atitlán III eruption that happened 84 000 years ago. In the wreckage of the caldera collapse, the volcano started a new cycle towards yet another major explosive event, by the rapid serial construction of 3 major volcanic edifices, one after another and each of them built more rapidly than the previous.

The oldest of the 3 volcanoes born after the Los Chocoyos caldera event is the 3020 meter high Volcán San Pedro. It was built during 44 000 years of eruptive activity ending 40 000 years ago. After that volcanic activity started at the Volcán Tolimán which in the next 30 000 years grow into a 3 158 meter high edifice.

On the flank of Tolimán is the failed volcano of Cerro de Oro that might have erupted 5 000 years ago. Instead the new main volcano was born on the rim of the old caldera instead of within it as Volcán Atitlán started erupting 10 000 years ago. This volcano grew very rapidly and matured into a classic 3 535 meter strato volcano.

In historic times Volcán Atitlán is the only erupting volcano in or near the Atitlán Caldera. The volcano suffers mostly from small to medium sized explosive eruptions. In the 384 years between 1469 and 1853 the volcano suffered 12 eruptions ranging from VEI-2 to VEI-3, but after 1853 the volcano has gone into a dormant state without any obvious signs of a pending eruption.

The area is highly seismic and in 1976 the area was struck by a M7.5 earthquake that killed 26 000 people. The earthquake most likely fractured the bottom of the lake and within a month the lake water level dropped two meters. It though rapidly filled up, most likely due to later seismicity closing the fractures.

The Lake

We all want lifeguards in the water, but lifeguard-towers? Photograph by Dr Carmen Morataya.

We all want lifeguards in the water, but lifeguard-towers? Photograph by Dr Carmen Morataya. Notice that the water level has under-mined the stone wall.

Lake Atitlán is more than 300 meters deep and is covered with a 300 meter deep layer of sediments. The sedimental layer is filled with prodigious amounts of methane that effectively hinder analysis of the actual caldera bottom 600 meters below the surface.

Sedimentation occurs with a normal rate of 0.5 centimeters per year, and that accounts for half of the 20 meter uplift of the water level in the last 2 000 years. We know the rate of uplift from C14-dating of artifacts found in a sunken Mayan village.

The lake is suffering from extensive deep hydrothermal activity releasing large amounts of energy (290mW/h inside the lake and 230mW/h outside), this creates an ideal environment for the silt producing cyanobacteria inhabiting the lower two thirds of the lake (the source of the methane in the sediment).

To compound the problem the lake is endorheic, lacking any outlet of surface water, instead release happens through seepage into two river systems through the caldera rock. This traps the water at depth and makes it very stagnant and a perfect breeding ground for highly toxic cyanobacteria.

The Caldera today

Who pooped in the water? Photograph from Prensa Libre.

Who pooped in the water? Photograph from Prensa Libre.

By now the caldera has filled in with about one third due to the construction of the three volcanoes, and another third has been filled in by silting and resurgent pistoning of the caldera floor due to magmatic intrusions.

Under the caldera floor there is believed to be a large reservoir of magma with an unknown amount of eruptible material. The amount of hydrothermal energy seems to signal that the amount of hot material is either very large, or quite close to the surface. The lack of doming though seems to indicate that the main reservoir is at depth since the resurgence is uniform over the area.

During the last 2 000 years there is evidence of episodes with highly increased silting due to hydrothermal energy increase. This most likely are signals of magma moving upwards into more shallow repositories.

During the last 15 months the surface of the lake has risen uniformly more than 200 centimeters, something that has caused great problems for the local Mayan residents.

The dangers of Atitlán Caldera

There are four principal dangers from the current volcanic system. I will briefly expound upon them in the order of possible risk for anyone nearby.

The largest danger is obviously seismic. Another large earthquake may happen here at any time. The second largest risk is of course that Volcán Atitlán suffers a renewal of volcanic activity and erupts. The eruption would not be that large, but evacuation of nearby population might be problematic.

The second least likely risk is due to the unique sedimental setup and the large amount of deep water cyanobacteria. With time the amount of methane trapped in the sediment will be higher than is possible to contain within the confines of the sediment. As the lake gets closer to the threshold of containment the likelihood of an earthquake setting off a chain reaction increases and as that happens the bottom sediment will explode out of the lake bringing with it a deadly mixture of cyanobacteria, SO2 and methane poisoning everyone within the caldera. For those who are unfamiliar with these phenomena I recommend googling the Lake Nyos disaster.

The least likely risk is that a large fast explosive eruption (VEI-6) will empty out the magma reservoir under the caldera floor at such a speed that the roof of the reservoir collapses and a supereruption occurs.


The rare volcanic water-palm as seen in its natural habitat. Photograph by Dr Carmen Morataya.

The rare volcanic water-palm as seen in its natural habitat. Photograph by Dr Carmen Morataya.

There are many possible ways to explain what is currently happening with the water level in Atitlán Caldera. There are though three that are more likely than the others.

Seismic activity could have pushed together the natural pathways for the seepage out of the lake into the rivers. There is though no evidence that the rivers fed by the seepage have suffered a decrease in water reception.

Then there could be increased upwelling of hydrothermal water into the lake. This would lift the level of the cyanobacteria boundary upwards and there would be cyanobacterial (algae) blooming. We know for a fact that there have been large problems with such blooms during the last 15 months.

The third cause could also be piston-like inflation of the caldera floor due to magmatic intrusion either at depth, or in a shallower reservoir under the lake itself. The most likely thing in the end is a combination of a shallow emplacement causing a dramatic increase in hydrothermal activity.

Locally the belief is that the increase in surface cyanobacteria is due to release of waste water and other nutrients into the lake. This is highly unlikely to be the main cause; the waste water flow into the river has been roughly the same for decades. And in the end this would have caused normal algae bloom of garden variety green algae, not cyanobacteria. Why now? Well, for starters waste water contain too little sulphur to feed the bacteria, also you need something that cooks off the oxygen, and explaining that with sunlight warming the water does just not cut it. It is much more likely that an increase in hydrothermal activity has elevated the temperature of the lake and deposited the necessary sulphur to feed the cyanobacteria.

Most likely this is all caused by a deep intrusion of magma at depth, of which a small part has intruded in a shallow reservoir under the lake. This in turn caused increased turbulence of the water at depth as the levels of hydrothermal activity sky-rocketed. In the end there is an increased risk of eruptions occurring in Volcán Atitlán or from a new vent. I do though not believe that this will cause any large eruption within the foreseeable future. After all supervolcanoes are well known to inflate at prodigious rates and very rapidly without erupting, or erupting through one of the side vents that functions as safety valves.

Sorry all, no supereruption seems to be around the corner even though there is a bit of pause for thought when something on this scale shows signs of unrest.



359 thoughts on “A Super-pooping Supervolcano?

      • No, it warns airplanes that the maximum ash elevation might be up to a maximum height of 7.6km. Not the same at all as the ash column being that high. It is a misconception that the VAAC warnings has anything at all to do with reallity. 🙂

        • Well, the Mastin et al document was written with that in mind. Sparse, poor or unreliable data. Either way, the mass output values that I calculate from the VAAC reports should be taken as a maximum estimated value.

          Note: If my statement seems disjointed, I blame my phone which kept talking into my ear.

        • Er …. it says “VA PLUME LAST OBS TO FL250 OBS AT 14/1000Z ON WEBCAM.” Thought “OBS” was “observed” – but could be wrong.

          • but observed by webcam is slightly unreliable,
            EYJO was mesured by Theodolite!
            then topmost referance on “webcam” can be FL250 (“Flight Level 250” – 25,000 feet / QNH 299.92 in / 15 Deg C / or QNH 1013.25 Hp at sea level) Anything hotter, Cooler or different presure QNH then there is errors from that altitude)

    • And this still having “fits” (Esjufjöll / Öræfajökull)

      And Laki “railway line” has no trains coming, so far as is known 😉

      I think we can blame it all on “and”!

    • She has not ruminated in over a week, neither has she been logged on to facebook.
      So we hope she is well and will be with us soon.

  1. The lava dome at Sinabung has turned into a lavatongue. The old dome was really hard and slow moving, this is less viscous lava now moving a bit more easily.
    Thanks to KarenZ for finding the picture.

    • This one is really peaking my interest. The lava is yet another step less blocky and seems to be at tad more fluid. I know it is not runny lava so do not harp about that. My point is that the lava is changing towards a more “fluid” state by every passing day.
      Remember that up untill 3 days ago it constructed a solid magma plug (dome) that is now turning into a bendable lava tongue.
      Image and video hosting by TinyPic

    • Look Carl, this is absolutely normal in an eruption of this type – lava domes can grow into short (sometimes even longer) tongues without any change in the composition or overall fluidity of the lava, all you need is some change in the rate the magma is extruded. In this moment, the Sinabung dome looks exactly like some of the most classical lava domes that are known – those of Merapi. There is no way an eruption like this will turn from a typical andesitic or dacitic, dome-building eruption into some more Etna-type eruption (not to speek of Hawaiian-type). Volcanic eruptions hardly behave according to our man-made classification scheme – a dome is a dome and can only be a dome and thus it categorically must not evolve into some stubby flow-like feature; there are all sorts of transitions and overlaps. It’s something I have learned in now more than 4 decades of following volcanoes and eruptions and doing my literature researches. For now, Sinabung looks as much as a photocopy of a typical (that is, not 2010-like) eruption of Merapi as any eruption can look. And Merapi, after all, is the prototype of a dome-building volcano with each dome-building eruption producing pyroclastic flows, sometimes even quite large and deadly ones, like in 1930.

        • It’s a totally classic lava dome, which is slowly expanding down a very steep slope, where it starts to disintegrate into pyroclastic flows. We’re so much used to images of lava domes growing within a crater without producing pyroclastic flows – like Mount St Helens, or more recently Kirishima – that a dome like Sinabung’s seems something totally new, but it isn’t. We’ve seen it all before, over and over again, and this eruption has nothing exceptional when put into a historical perspective.

          Let’s try and see some other examples, old and new …

          Montagne Pelée, 1930

          Redoubt, 1990

          Redoubt, 2009

          Unzen, 1991-1995

          Soufrière Hills, 1997

          Merapi, 1836 (!!!)

          If enough lava is extruded in a short time, the dome may spread out into a flow (or as named by Lucas, a “coulée”), as happened at Merapi in 1931.

          However, if the rate of lava extrusion is too high, the lava dome and “coulée” may collapse wholesale – as has often happened at Soufrière Hills between 1996 and 2010, and in 1930 at Merapi.

          If the magma supply rate is high and there is a lot of gas in the magma, it will explode, as happened, still at Merapi, in 2010.

          Dragon released: initially sent to dungeon for too many links

          • I am apparantly not getting across regardless of how I try to explain what I am talking about so I give up entirely.

            You are beating me over the head with pretty images of apples when I was talking about a pear. I have though recognised what the problem is, volcanology is lacking a term and I have noticed that sometimes terms are needed to explain what you are trying to tell.

          • Interesting discussion. So Boris, are you saying that it is extremely unlikely, given all these precedents, that Sinabung will develop into an explosive eruption given its present behavior? I think all that Carl is trying to get at is that juvenile magma might be more gas rich when it reaches the surface and that what is erupting now has had a chance to degas. Is this conceivable?

            • .. i.e. the stuff erupting now is the product of the intrusion that occurred a hundred years ago and in that sense is not “juvenile” although technically it is juvenile because it is only now being exposed at the surface.

            • The hypothesis of the eruption becoming more explosive is certainly feasible, but that would mean that there would be no more lava extrusion at all (as we have seen at Merapi in 2010, where rapid dome growth was followed by the major, purely explosive phase of the eruption). I don’t know how likely such a scenario is, but certainly more plausible than the eruption producing more fluid lava 🙂

            • Meh!
              Now you are misreading me again Boris 🙂

              I am not saying that the lava will be fluid like basalt, just that it seems a tinybit less ultra-sollid. Remember that in the beginning it was basically crumly rock, and now it seems to be a bit more “bendy”. The initial lava could never have bent like a tongue.
              Borrowing a term from material sciences, the lava now has a higher degree of tensile strength compared to in the beginning of the eruption. So forget that I ever used the word “fluid”.

            • ok here we are … let’s say, the conduit is getting more lubrificated in some sense, magma might be getting through faster, and at a higher rate. All this might contribute to the dome looking smoother and being possibly more plastically deformable (which probably is the best term to describe what you Carl are presumably meaning 😉 ) — in any case, if we take similar eruptions from the past as example, those at andesitic/dacitic volcanoes that have not erupted for a long time, such as Unzen 1991-1995 or Soufrière Hills 1995-2010, Sinabung may go on erupting for many years and the phases of strongest activity are yet to come. A crucial factor in evaluating how this eruption might evolve lies in the past of the volcano, which seems to be little studied. Nevertheless there’s an article that has a very simple geological map of the volcano, where you see numerous thick, broad lava flows – the so-called “coulée” type – extending radially down the flanks. So after all it seems that many eruptions in the past have maybe initially produced lava domes, which then spread downslope as thick, viscous lava flows, similar to those seen since the 1920s at Santiaguito in Guatemala and also those recently emitted at Kizimen in Kamchatka.

              The Sinabung paper with the map can be downloaded from the following link (registration is necessary but then download is free)


            • Yeah!
              Finally I have made myself understood, and yes, plasticity is a good term for it.

  2. there have been a few misses in the US of A, is the global positioning system gone haywire, just a thought, I think it was in Florida they recalibrated the runways not so long ago.
    It is 32.3 deg at 7pm, I still can’t open windows etc., but two days down and three more to go, was only 38 deg today

      • I might go exploring some extinct ones once there is some frost around, I don’t fancy snakes, there are plenty around now with the dry weather, like Brown snakes they are not to be messed with, red bellied Black usually go away if you give them a chance

    • Misses in the USA?
      I do not think that the GPS system has gone haywire. There have been a few minute dips that might or might not have been caused by solar flares, but no GPS worldwide failure yet.
      But, the GPS system is going to fail within a not to distant future due to lack of satellite replacement. The US military stopped replacing them years ago so most satellites are running on overtime. NASA have shot up two (last time I checked) but that is not enough.
      Therefore many developers of GPSes are starting to build multi system GPSes using the ESA-system and the Russian GLONASS to be able to give high enough accuraccy. Problem is just that neither of those systems are complete either…
      I am happy I know how to use a sextant, we are soon back to using those since the LORAN-system, Lighthouses and almost every other navigation aid has been discontinued. Sooner or later we will have super-tankers cruising down Santa Monica Boulevard if this continues.

        • I still have a DECCA system around.
          In the end it would make sense to return to a radio based system. With todays technology we could quite easilly set up a system containing 11 equidistanlly placed longwave transmitter, 22 medium wave and about 100 HF-band transmitters. By doing it tripple band you get the accuracy where it is needed, and if you grid every Digital Radiostation into the system you get basically the same precision as with todays GPS system.
          The cost would in the end be the same as 1 satellite per ten years.
          Problem with the old systems was that they used a hell of a lot of energy transmitting since the receivers back then was rather shitty in the signal-gainstage. Also, the signal was basically just a location signal. Today you can send a universal timecode that is the same at all masts, that way you would get vastly improved distance-coordinates for the masts. And so on and so forth…
          So what would the precision be? 50m at all spots on the planet including the poles (where no GPS works), good enough for most ship navigational purposes. 1cm to 1m along all continental coast and about 1mm for a continental receiver that has 11 HF-masts. One could also tap into the ping of cell-towers to achieve that, after all the ping is sent out from every mast on the planet for free and the location of every mast is known…

      • Could be just that magnetic north has gone walkabout again. It has after all changed direction from its easterly wandering to more rapidly southerly in recent months – sorry – can’t remember where I read the numbers…

      • Last summer i was talking to a student pilot -he was an native kid from Alaska-his mom taught
        native studies here at the local university- He was saying the same things Carl was.But he wasn’t
        too concerned as he said: “If anything craps on the GPS system I’ve flown all over Alaska in Dad’s
        and Grandpap’s lap in the family Cessna 185, so If i can get a good look at the ground, i don’t need a map,gps or anything else, heck Grandpap taught me how to navigate by the stars.”
        I’m of the same mind myself..His dad is dead-but Grandpap’s still around. I would love to meet him..
        (Cheaper to learn to fly in the lower 48 BTW.)

        • Problem is when you are not in your own back-waters. I normaly sail into harbours and lanes I have never been in. And even when using charts and pilot guides it can be a fairly hair-raising experience if you are going into a dark port at night that is narrow.
          My point being, navigation is possible without GPS if you really know what you are doing, but few people have that level of skill. I do have that skill, and still sometimes I hear “bump” under the keel (guess why I enter harbours really slowly…).

            • Silliest bump was in a 30 fot Van der Stadt competition boat. It had a 2m draught and the keel was razor sharp and stood for about 1.8 meters of the draught.
              We sailed hard and fast at around 10 knots when we felt a really bad “bump”. Since we sailed in a 11 meter deep well marked channel we where really stupefied at what on earth we could have hit. To our amazement we saw two halves of a large salmon emerge out of the wake-water.

  3. Hello everyone!
    Just getting home after a wonderful stay in Austria/Italy. I’ve been reading all the post, but very little time for commenting.
    Great work everyone!
    Rio, 28 ° C, but I was very lucky with the mild winter in Europe – 21 days of magnificent sunshine!!!!
    And thrilled about Ars Eletronica museum in Linz and meeting Spica in person – what a wonderful person!
    Concerned about Sinabung, although I agree that that is gong to be all it will be doing (p-flows) for now. Hard times ahead for the evacuated people, whom, my heart goes to.
    Best, Renato Rio

    • Welcome back home to VC Renato!
      Here the winter has finaly come with a glorious minus 20C, snow and sunshine.

      • Hehe! My friends in Austria are trying to convince me to have a trip to Umea° – the Capital of Culture of 2014 – but you are not encouraging me with such news…
        Big waves from Rio!

        • You are very welcome to come to Umeå, of course with a benevolent guide! I have to work hard at guideing after Spica 🙂
          Soon the weather will turn and become better again!

          • I am sure you will do a great job at guiding too!
            Time to work hard and be able to go back more often – especially if I don’t get much distracted by the wonderful discussion on volcanoes… 😉

            • I do think there is a serioust risk of volcanically related discussions if you come here 🙂

  4. Last ditch attempt to explain what I am talking about when I say that the magma of Sinabung is changing…

    We are used to bimodal lavas from a lot of volcanoes, the most famous paper about bimodal lavas is Carmichaels paper on the lavas of Thingmuli.

    If I have understood Boris explanation of Etnas behaviour it is that before a paroxysm there is a lid of old volatile deprived magma that acts like a cork. Under it is a layer of younger Volatile Rick Magma™, as the volatiles release it pushes the old magma aside and partially up through the vent, the magmas do not mix so you get two stages of the eruption.
    Any mistake here is mine, I am using it as a model, so even if I am wrong the model stands for a bimodal eruption. I am after all not talking about Etna. We could use Hekla as a model instead, after all it has a explosive first phase with one lava ejected, and a sharp boundary to a runny lava (the change take place within a minute).
    We could say that it is a “digital” eruption, consisting of two distinct stages 0 (explosive) or 1 (lava-flow).

    Now over to Sinabung. What I have been trying to get across is that the lava extruded is changing very slowly. The initial extruded lava in december was very stiff, It extruded as a plug that could not bend at all, it fractured directly into ash, semingly without any large amount of glow or blocking. For the first ten to fifteen days the lava did not change at all (at least not that I could see). It looked exactly the same.
    But during the last ten days or so it has changed into a lava that is slightly more flexible. It is either hotter, or marginally of different composition or a combination of there of.
    We could therefore say that Sinabung is “analog” instead of “digital”. It started with the same 0 (plug/dome), but instead of going to 1 (lava-flow) it went to 0.1, 0.2 and so on… It is still producing PFs, that is not a question, it has just changed a bit from the coldest most brittle type of PFs to a hotter PF type.

    If one take a look at the earliest timelaps photos they lack any glowing nice red, now they are filled with it.

    I hope I have actually succeeded with explaining what I am talking about this time around. Otherwise I have a problem since I can not tear out my hair since I am bald as a bat 🙂

    Caveat = I know that bats are not bald at all, just so that no batologist start sending me pictures of hairy bats.

    Edit: Trademarked due to the comment below 😉

    • Hi Holger, do you mean this one?

      The parallels to Rangitoto are striking:

      Covered in bush now but the entire island is made of pahoehoe flows interspersed with A’a flow fronts. There is also a sandy bay connecting it to a neighboring island.

    • It such a cute little island that I really want to adopt it. If I ever go sailing that way I think I will build a hut on the island.

  5. Jobs I do not wish to have.
    In Mexico City there is a diver named Julio. His job is to dive through the shit of people in the sewers to unclog the poop. It is the impressively most discusting job I have ever seen or heard about. Just the image of him re-emerging out covered in human manure was rather revolting.
    So, if life seems like it is filled with poo, think about Julio and you will understand how little poo there is in your life (I hope).

  6. I hadn’t realized that the 2010 eruption of Merapi started out with dome building prior to turning into a more explosive eruption. I had really assumed most volcanoes that go through prolonged periods of dome building & destruction don’t end up with explosive phases of the eruption later-on.

    I guess you learn something new every day 🙂 .

    • I still think that in the end Sinabung will turn into something like its next door neighbour who has explosive eruptions. After all they should have roughly the same sourced magma, it is after all only 8km between them.

      Edit: Better clarify myself before I get thumped again for being unclear…
      I mean that the deep source of the magma should either be sharing the same creation zone, or the same basic composition. So, as time goes by there should be more and more similarities between the volcanoes. I am not in any way saying that they share the same magma plumbing system or the same magma reservoir.

      • I’ll be grabbing screencaps as this is ongoing. The weird thing? Right as this happened, the song “pompeii” by Bastille came started playing on my playlist (on shuffle).

      • I REALLY hope nobody is in the area near the webcam right now. I would be getting out of there fast..

        I just posted a bunch of links to pics, but I think they all went to the dungeon.

          • Really hard to tell. I wish there were another webcam further back where you could see a broader picture. Everything is just a gigantic ash cloud here. I would guess it’s not a dome collapse. My reasoning? The dome has been collapsing for a while here with some decent sized pyroclastic flows, but nothing major. There isn’t enough time for a large enough dome to build overnight for this to happen.

            Also, I caught glimpses of a vertical column coming from the vent, which isn’t indicative of a dome collapse. I think this is fresh, more volatile rich magma erupting.

            • I disagree. For the simple reason that an eruption column from the main vent would go upwards first. The very fact that the webcam view is obscured indicates to me that this is another partial dome collapse. I think there is plenty enough dome still left to generate this amount of ash.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s