Phantom caldera

The shoreline at Laguna del Maule in Chile has been raised by 75 meter. Credit: Brad Singer.

Guest post by Tallis.

In my previous piece, I went over the process of large caldera forming eruptions and the abundant noise that they would produce but a comment by cbuso5 caught my attention and interest.

I have one slight disagreement. I’m not necessarily sure that the signs before a super eruption would be particularly obvious, at least for a volcano that isn’t being actively monitored. Would there be signs? Definitely, but determining the difference between normal inflation and inflation that precedes a super-eruption may not be particularly easy.

The issue is that we don’t have a long enough time to measure influx into a volcanic system. We only can see the magmatic input into a system that occurred in the last few years, not the 400 years before now. So a volcano that saw Iwo-Jima style magmatic input for 200 years before the magma inflation stopped may still be straddling the threshold between erupting a big eruption and not doing anything even 200 years later. In this type of instance, it may not take a large injection of magma to create a big eruption.

Another point worth making is that Rhyolitic eruptions seem to erupt extremely fast and violently. We know the violent part, but the “fast” part is what I think would catch a lot of people off guard. Chaiten was a great example of this, as it erupted extremely quick due to the explosive nature of the rhyolitic magma and the deep chamber. From what we know about the Taupo Volcanic Zone calderas, they also seem to exhibit this trait, which is frightening considering how large and explosive they are. The Tarawera eruption seemingly occurred in a flash from historical accounts.

The more pressure there is that is held back, the more likely it is that you get an eruption that accelerates extremely quick. Adding to this, it should be noted that in crystallized form, magma won’t exert tons of upward pressure. But if the magma reheats or is rapidly depressurized, that can change in a flash.

This comment is something I had casually responded to but my interest in the idea of a large subtle eruption has greatly interested me. Thanks, cbus05. This point has been strengthened by some volcanic systems. No system is the same; every large eruption will produce an abundant amount of noise and the coming event would be obvious at a certain point in the process leading to the eruption, but there doesn’t have to be a lot of time once that point is reached. The build up phase of a large caldera system will have abundant noise but not necessarily a long warning. The uplift, seismic activity and geothermal convection do not have to reflect the size of a future eruption.

Laguna del Maule is the fastest known uplifting volcanic system on the planet at the moment, it possesses a large magma chamber and is experiencing an intrusion capable of destabilizing the chamber. This very dangerous volcano has some of the most benign seismic activity you will ever see in a system of such potential. Micro seismicity is the only thing of note that you will see from this system in terms of earthquakes.

The system has not reached the point where it is causing earthquake activity through the stressing of regional faults. This system has had periods of unrest like this frequently in the past 100,000 years. Any rocks holding back the intrusion has been rendered inert. The pathway to the system has cleared like Agung recently. The only type of noise you might see from the system now are harmonic tremors and small micro quakes.

A system that has been multiple intrusive periods will not produce consistent seismic activity, that is not to say that there would not earthquake swarms but the swarm doesn’t have to accurately reflect the unrest within the system. This situation is frequent within volcanoes like Laguna del Maule, and makes the fact that there might be significant intrusion at a system unknown to the surrounding area is scary to say the least but what about the other signs?

The Volcano doesn’t need to be uplifting at a particularly rapid rate for the entirety of the unrest, past cyclical periods of intense uplift could easily be missed from an unknown volcano that seems to be rising 20-30 mm a year. In fact, it is so easy improperly gauge the past that some eruptions that have taken place just a few hundred years ago were not properly scaled until recently. Past unrest would be even harder to pinpoint conclusively and once another severe period of uplift start, most would think that volcano is just now starting to prepare for dinner when the chicken has already cooked, and on its way to the table.

A slow intrusion of mafic magma could destabilize the system over a period of 1000s of years while causing minor uplift, and while looking benign in terms of activity.. Frequent intrusive episodes can maintain a certain level of heat in the chamber. The separation of the eruptible magma from the useless mush in the system could be some of the quietest experiences you could ever see. When the severe uplift kicks in it could just be a few months before the main eruption.

It is important to remember that not all volcanoes prepare for a large eruption the same way, when considering the size of a future eruption, the history of the volcano should be established. The uplift for a VEI 8 eruption doesn’t have to be extreme for most of the build up phase. If the chamber is molten enough, the volcano would just need a smaller intrusion to get things moving

The accumulation of eruptible magma would likely cause an increase in heat within the system but that also doesn’t have to represent the danger of the system. Ironically the cooling of the magma chamber could also represent the geothermal activity. Geothermal activity is not a usual sign of a large eruption and would be taken seriously be no one who is in a position of authority, The Los Humeros system is an example of this, despite the heat within the system being possibly caused by magmatic intrusion and the large chamber and scary history, it still is not being monitored properly.

The heat may not even be transmitted to the surface properly, and even then there might not even be any good ways of measuring the activity without a large backing. Rhyolite is the coolest magma, in a way this magma may not even have the ability to generate abundant heat.

It is kind of scary to think about how little warning you could have before an event in excess of 1×10^22 J event. An eruption like this will still create an abundant noise at some point but that point might be at the end of the activity, a few weeks or months before the eruption. It is even possible that something like this already happening with certain systems like Campi Flegrei or Aso but that would be more speculative and unlikely. For monitored volcanoes, this idea could be used to monitor more subtle details and perhaps find out what other signs a volcano makes before erupting.


76 thoughts on “Phantom caldera

  1. I like the MKS quantified figure for a large eruption of 10ZJ (10 zotta joules).

    Should be used more often ….

  2. Good post tallis. I think the main point I wanted to make was that our monitoring of volcanic systems is based on a human time scale.

    For a supposed supervolcanic system to build up to a potential supervolcanic eruption, it probably takes hundreds of thousands of years for the requisite magma to accumulate prior to said eruption. Or in a taupo like scenario, it would take a considerable amount of time to melt the surrounding country rock prior to an eruption.

    In either scenario, we are assuming that our monitoring of volcanic activity over the past 10-20 years can tell us whether a volcanic system is on a tipping point or not based on very short-term seismic signals. If we had a far longer sample set of data going back 10,000 years on our volcanoes, this would likely be a lot easier, but we just don’t have that. And we don’t really know what behemoth volcanoes may have been straddling the border of erupting something very large, or staying quiet over the past 200 years.

    • Naturally when a large eruption does take place sometime in the future, since no one remembers the shear power of a VEI 7. I am sure whatever volcano that doesn’t take place in Japan will not get the proper preparation or monitoring before it erupts.(Not to demean the the scientists)

  3. Does anyone have a comment on the proximity of the Ridgecrest earthquake to Long Valley caldera. I know the earthquake itself is not thought to be caused by magma movement. This is one of those large calderas where the eruptive history suggests a short time span between normal background activity and a major eruption.

    • Long Valley is maybe 100 miles north of Ridgecrest. Same valley but not next door.

  4. Nice article, Tallis!.
    I would suggest that rapid magma rejuvenation needs to be added to the discussion, since this process would result in a sudden pressure increase without necessarily having a large magmatic intrusion taking place. Like a pending earthquake that’s ready to snap from all it’s stored energy and is just waiting for one final “trigger” event, same goes for many volcanoes where the magma/crystal mush is already in place but thermally stable…hence seismically silent. However, add just a few degrees of additional heat from below and the sudden increase in pressure from the chemical reactions within the mush can trigger rapid inflation and seismicity,..;.perhaps leading to an eruption….all without needing a sudden/massive influx of magma at depth.

    • That would be the greatest soda prank of all time. I can already imagine the youtube titles
      GOD SHAKES VOLCANO! (GONE WRONG) Suscribe to Satan!

    • M 5.4 – 227km W of Bandon, Oregon
      2019-07-17 14:59:59 (UTC) 43.477°N 127.165°W 10.0 km depth

      Tsunami Information Statement Number 1
      NWS National Tsunami Warning Center Palmer AK
      803 AM PDT Wed Jul 17 2019


      * There is NO tsunami danger from this earthquake.

      • With a 2.7 foreshock

        M 2.7 – 220km W of Bandon, Oregon
        2019-07-17 07:08:57 (UTC) 43.370°N 127.098° W10.0 km depth

        • On the Blanco fracture zne but I wonder if there isn’t some shifting of the Juan De Fuca plate going on. makes me kind of nervous…

          • Yeah, that’s getting a little close to home (I work in Portland)…

          • That Cape Blanco Fault is a very active one; does not seem to be linked to Cascadia.

          • No. But the slab that it is shoving sideways like a wedge does.

  5. Swarm of quakes over the past month in the Lake Taupo caldera :

    The pale blue circles all under 15km depth. M2 – M4.
    So, it’s quite normal to have these swarms, comments from Geonet following a similar swarm nearby a year or so ago:
    “The quakes are likely to be related to the long-term tectonic stretching of the Taupo Volcanic Zone. Swarms are often characterised by no one main or large event, with many of the events being about the same size. Currently, there are no indications that the earthquakes are related to volcanic activity.”

    When I see this, my heart still skips a beat.

  6. Very large magma chambers need to be very deep, otherwise the centre can’t hold. They can also not have many earthquakes, for the same reason: once the roof cracks, it is over. And they don’t grow that fast because the magma supply rate is limited, and does not grow with the size of the chamber. So inflation associated with them will be slow and widespread. Large, fast inflation, on the other hand, is associated with shallow and thus small magma chambers. I am not particularly concerned about Maule which most likely will suffer a smallish (VEI4-5) eruption. Iwo Jima may go larger, but in general it is the pre-VEI-7’s that will be hardest to recognize. And of course, large explosive eruptions tend to be rhyolitic and these do not show much inflation. Krakatau may seem an exception as I think it did inflate prior to the events, but it started out as a small eruption which suffered an accident during the eruption.

    Which are the plausible VEI-7’s? You need a significant volume so they should be recognizable. Sinabung?

    • In an article on Dr Erik Klemetti’s site, Dr Clive Openheimer, (latent Dr Who look-a-like) puts his money on Uturuncu. He stated something to the effect that it has a quite large heat influx and is surrounded by other previous Large Calderas. (I refuse to use the media fabricated StuperVolcano term).

      Note: that is not a quote of Dr Openheimer, just how I recall the article.

    • Some of the japanese volcanoes and Alaskan volcanoes jump to the top of my mind personally.

      – Ata caldera system / Kaimondake (Japan). Giant caldera system not too different from Aira / Kikai, but hasn’t erupted in a long time.

      – Aira caldera system. Not sure if it’s ready yet by any means since Sakurajima is at least a slightly effective blowoff valve, but Aira has been inflating despite this. If we were to get a large Sakurajima eruption, there is an outside shot of this depressurizing the deeper Aira caldera system if enough damage is done to the system and water from Kagoshima bay causes problems. Not saying it’s likely, just saying it’s a possibility.

      – Aso. If we’re going Japanese, Aso has to be mentioned here as it hasn’t had any significant eruptions in a very long time, yet clearly is not a dead volcanic system.

      – Many of the alaskan volcanic systems, but more likely some of them that haven’t gone caldera recently. There are a TON of big volcanoes in Alaska. Hard to pick one, and there just isn’t enough information about the inflation or activity levels for some of these. But we know there are quite a few big complex volcanic systems that haven’t erupted for a while, yet seem to have a pretty high input rate based on their size.

      There are of course a lot of other great places to make a pick. So many options in the Andes and southeast asia as well.

      I don’t think Sinabung would be a great candidate, it is very clearly a much younger volcanic system and seems to be behaving as such.

      • my money is on Chiles / Cerro Negro de Mayasquer. It’s reasonably large, hasnt erupted in a very very long time and is apparently receiving quite a bit of fresh magma.

      • Central American volcanoes are often overlooked so much so that InSar data is actually lacking that area. Bad idea since that ilpango is there.

          • Dunno about that. I used to study Olongapo. It was quite proficient at draining your wallet.

          • Nabro and the entire Bidu volcanic complex are vastly underappreciated and understudied.

          • Problem is the extreme remoteness, I wonder if anyone has ever stood on the summit of Nabro, maybe some Afari but probably no outsider.

      • Not so sure about the Alaskan volcanoes; if you look at the record many of the calderas date from late Pleistocene/early Holocene when there was a spasm of caldera-collapse events; possibly the geologically rapid unloading of the regional ice sheet at that time may have been a factor

        • Could be, although the caldera events have continued into more recent times (last 3000 years at least)

    • I am more suspicious of Laguna del Maule, the area around it has several VEI 7 volcanoes, Laguna del Maule being one of them but I ignore the state of its magma chamber, so if it is cold and mostly solidified it would not be a good candidate, its active neighbour Descabezado Grande having produced 3 VEI 7 eruptions in the last 0.8 Ma might do better.

      A wide variety of volcanoes can throw a VEI 7. Tambora was an effusive basaltic andesite volcano before its big blast, every subduction zone probably has a couple of volcanoes building for a VEI 6-7 eruption in a, geologically speaking, short term.

    • Fairly extinct, I think. A new eruption here would be rather unlikely. After all, there is a reason it hasn’t erupted for so long.

      Why can’t NatGeo use kilometers rather than miles? No one gives magma volume in cubic miles.

      • To be fair, they do give a “Great Pyramid of Giza” count as reference. I think what we need is a conversion table between pyramids, dump trucks and fridge sized refrigerators.

        • They missed the olympic sized swimming pools which must rank among the most undefined units in the press. Now how voluminous is the great pyramid.. apparently it is about 100 million cubic feet which doesn’t help much. A cubic feet sounds like a bad medical problem. (The number assumes a smooth shape and ignores the stepped stone structure.) It is also about 20 million cubic royal cubits, or a bit less than 400 million cubic palms. You need a degree in archeo-mathematics to read the national geographics.

      • To be honest. The dump truck size reference was mine.

        Mainly because dodging them is a needed skill in Florida.

  7. So rhyolite explodes instead of melts when you heat it up.

    • Depends on confining pressure and gas content, but in general. That’s the idea.

      The high silica content make it thicker and more difficult to evolve dissolved gasses.

      • Would be interested to learn more about this, why rhyolite is more explosive. How does this develop and from what chemical reaction comes the energy to drive an explosion? Maybe some old post in the VC arcives?

      • That’s actually pretty much all there is to it. In a magma chamber, the minerals crystallize and sediment out. This reduces the mafic element concentration leaving silica rich minerals.

        The more silica rich the magma, the higher the viscosity. Dissolved gases don’t come out of solution as easily. When the magma enters an area where the pressure is low enough, the gasses begin to come out of solution and nucleate into bubbles.

        • So, the rhyolite fraction is enriched in the magma chamber due to lower density and its high viscosity results also in trapped volatiles such as H2O etc. When the pressure decrease these volatiles enter their gas phases, expand and cause explosions. There are actually no true chemical reactions involved? Is the rapid entry into gas phase and expansion an energy onsuming transition? In that case this would lower the ejecta temperature! High temperature would be required for an eruption… Just cehcking if I got the model right…

          • When the magma moves into a temperature/pressure regime (high enough temperature and low enough pressure) where the gas “wants” to exsolve, un-exsolved gas has potential energy. When the magma degasses in the onset of an eruption, this potential energy is converted into kinetic energy. Boom!

            Basalt can usually release this energy gradually, leading to less explosive eruption styles. Rhyolite can’t. It’s like the difference between burning gasoline and triggering TNT. You get some fire either way, but no sudden boom with the first one because the energy can get out gradually.

          • @salting. Correct that there is not an energetic chem reaction involved. Gasses under pressure seeking equilibrium are energetic enough. Water is particularly nasty since it can expand up to around 1700 times its volume. Anything phreatomagmatic has water involed.

        • First of all – sorry to hear that Carl is gone. It would be a huge enrichment to this blog, should he decide to return one day.

          Lurking, may I ask: how do pegmatites form then? I thought it is rhyolite dykes that cooled down? Or is it normal dykes with normal basalt magma that differentiates over time and become rhyolite?

          • I think Pegmatites form in the largest ryholite magma chambers ( batholiths ) where cooling is the very very slowest: this forms very coarse grained granites
            Pegmatite crystal structure with largest crystals.

            Pegmatite can have any composition mafic, felsic but often its granitic pegmatite

          • Jasper is quite correct about long stay times yielding the larger crystals.

          • Thanks for your answers! But in the field you often seem them in the form of a dyke or a layer that obviously intruded into a crack in some granitic host rock. Batholiths I guess are more roundish or solidified-chamber-like, no?

          • The really coarse dike pegmatites often have a hydrothermal component and/or exsolved pockets of fluids, in which case you can get some amazingly large crystals with less time required for growth, (growing out
            of saturated solution rather than crystallizing directly out of magma) but this process still requires a long time for the initial differentiation to happen. Mafic pegmatites are usually derived directly from magma and don’t usually form pockets, even so, you can get some pretty large crystals especially the feldspars like labradorite.

          • @TJElJohnny

            Don’t be so certain about Carl’s absence. I’m pretty sure that if something in Iceland goes weird… we have a chance of seeing him decloak here. As far as volcanophiles go, Carl has got it bad.

          • @cabrageo – great explanation that connected quite a few loose ends in my understanding of the process, thank you!

            @Lurking – that sounds promising. Somehow hoping for weird things to happen in Iceland now!

          • Just to make it even easier, think of growing a crystal of salt, if you tried to make it out of molten salt you would need a much higher starting temperature, and are more likely to end up with a fine-grained mass as the temperature passed the solidus, whereas growing out of solution happens at a much lower temperature, but allows larger crystals to form. Superheated water under pressure is an excellent solvent, and will move silica and other materials around surprisingly easily.

  8. UK attempts to re-create Doggerland 🙂

    Huge sand dune designed to prevent major gas terminal falling into sea

    In a UK first, 10,000 cubic metres of sand is being pumped every hour to create a 6km (3.7 mile) sand barrier to prevent Bacton gas terminal, which supplies a third of the UK’s gas, from tumbling over the edge of a cliff into the sea.

    It’s hoped it will also save the coastline, as well as the villages nearby.

    • I trust someone told them that sand dunes have a habit of moving…
      (And to be frank, that video reminds me of King Canute’s efforts.)

  9. IMO:
    “An earthquake of Magnitude 3.3 was detected in Torfajökull glacier at 14:15 today. An earthquake swarm has been in the area since last night with some smaller earthquakes having been detected. All of them have been measured to be of Magnitude 2.0 or smaller. The swarm is still ongoing and not all earthquakes have been manually checked, so more earthquakes are likely to appear on the seismic map soon.”

    before Greip and Grim..?

    • Green star there now, very shallow just as the weaker quakes. There have been several swarms before in recent months but then in the northwestern parts

  10. Here’s some video of Ubinas (local news show, it seems)

Comments are closed.