Typing eruptions

Guest post from Tallis

Various types of explosive eruptions have been prevalent across history, after all not all volcanoes are the same. However some volcanoes have something in common with another whether it be a magma type, location, setup, age, or size. Allow to define what I mean by “Model”. My strongest science is not volcanology but meteorology. I’ve been on the receiving end of way too many hurricanes and droughts and even got hit by a tornado (a weak one). When I and other Weather people discuss future events, we usually compare the setup to past events in order to get an idea of what could or couldn’t happen. We have preset, unspoken ideas based on basic principles like CAPE, wind shear, and moisture. This is the basis for my volcanic models, variables will go into these models to get an idea for possible outcomes with certain volcanoes.

These models are based on past eruptions that have specialized, extreme or various impacts; have had similar eruptions in the past; and are well studied or observed. These eruptions are not black sheep but they are not the norm as such they are not well prepared for or expected. I have presented 3 more solid models all of which are based on real eruptions. These models are all extremely dangerous in the sense that even with good preparation these eruptions could still cause a great deal of death and destruction.

Here are some volcanic eruptions that I did not model. Keep in mind that this not a list of devastating eruptions and some of the deadliest eruptions don’t have models. I would like to say that I am not a scientist just a person who loves volcanoes. (So don’t make a hyperbolic article based on this, Daily Express.)

I. Nevado del Ruiz eruption of 1985

Armero overrun by the lahars from Nevado del Ruiz

This eruption didn’t make it simply because the actual eruption was far too simple and small. The tragedy could’ve been largely mitigated if the government had taken better precautions. As I’ve said these models could cause a great deal of death and destruction with good preparations. This is not to demean the gravity of the eruption and I do recognize the death it caused.

II. Mt Agung eruption of 1963

This eruption almost qualifies, as it was not what you’d call a black sheep but it wasn’t normal either. The reason that it didn’t make it is because, quite simply, that the eruption can’t reliably be scaled up. If I were to have a model based on the eruption, it would be meaningless for a higher end VEI 6 as well as a VEI 7. All of these models can be scaled up to a low end VEI 7 mainly because I want these models applicable to bigger and smaller eruptions.

III. 2011 eruption of Grimsvotn

Basaltic plinian eruptions are not black sheep but they are rare. However there is not really anything special about the effects of an eruption like this. A valid source of interest for scientists but not for disaster preparation.

Model Alpha

Primary concerns: Tsunamis, Pyroclastic flows and surges.

Specialized concern: Shock waves, lateral blasts, landslides

Location: Volcanic Islands of reasonable size and elevation above sea level

Based on The 1883 eruption of Krakatoa

Potential example: Thera eruption

The Thera eruption would seem to be a likely eruption


This model needs little introduction as it follows the same setup of the most infamous eruption of the world. In order to properly understand the model we must understand the spectacular violence of the Krakatoa eruption. One explosion during the eruption was equivalent to 200 megatons of TNT and at a distance of 160 km, produced a pressure spike of more than 1 psi. That is more than 3 times the range of the Tsar bomb. The material ejected in that one explosion must have surpassed the speed of sound and according to this rather questionable source{1}, the material moved more than twice the speed of sound. Given this dubious prospect let’s say that the eruption released 100 megatons of kinetic energy and we find the explosion released more then 1.3 trillion kg of material and the total erupted mass is around 7.5 cubic km DRE. [2] This means that the total erupted mass was around 18.7 trillion kg, the explosion released less than 1/10 of the total material.

Several theories have been placed on why the volcanoes was so violent most of which involved sea water. Phreatomagmatic eruption are relatively common but the catastrophic nature of Krakatoa is special. The reaction was quick and it held a massive amount of material. Other recorded phreatomagmatic events in the past couldn’t compare.


Most agree that the volcano had developed a plug before this explosion that the previous explosion had established but I don’t believe that is completely true because the system still had the energy to produce one last massive explosion after the third. I believe the previous explosions plugged the vents but the system had deeper stronger plug than that. The third explosion had resulted in the destruction of the plug for the vents and the severe weakening of the deeper plug setting the stage for the landslide. I believe this is the way it went as I don’t think a plug established on the day of the climax could withstand the energy and pressure required to make a 200 megaton explosion.

How would the water make such a massive explosion? The magma for Krakatoa was intermediate as such it was just the right level of viscosity and temperature for the water to get ‘stuck’ in magma. After the second explosion, a degree of water flooded in to the magma chamber and the initial reaction was held back by the deeper plug. The result was a steam logged, hyper volatile magma. When the deep plug failed to hold back the magma, there wasn’t enough open vents to slowly release the energy which resulted in the explosion.

There could be a multitude of other ways the water could react with the magma to create violent explosions, but there is one hypothesis could challenge the water theories. The violence of the explosion could be caused by the sudden mixing of mafic magma and felsic magma and the total mixture only needs to be less 10% of the erupted material to create a violent explosion. In fact I personally think that the material ejected in this eruption moved faster then my “Source” says, so it’s possible that the ejecta is even less then my calculations


This model has one of the most far reaching direct effects you can see especially from moderate sized eruption, as I had brought up before the shock wave alone could cause damage in a radius greater than 160 km. The tsunamis, driven by collapsing eruptions columns, could travel hundreds of kilometers worse still is the fact that there could be multiple 100 ft waves. The most dangerous part of this model is how deceptive it can be. The volcanic island could look uninteresting and the fact that there so many of these islands would mean that there could be a great deal of candidates for this type of eruption. However there is one especially dangerous candidate for this model.


Sant Angelo D’Ischia, a resurgent dome from a flanking vent.

This volcano is one of the scariest sleeping monsters within geology; Carl and I have already brought this volcano up in our own articles so this volcano is certainly no stranger to attention. This volcano has uplifted 800 meters in the past 30,000 years dwarfing every major known resurgent caldera, even Yellowstone and Toba. The volcanic island fits the model rather well and more importantly, the system already has a great deal amount of pressure and a little water would make a huge problem. I am going to get a little speculative. Using some rather crude math we know if this volcano were to follow the model then a lower end VEI 7 eruption could produce a 2000 megaton explosion. Such an explosion would cause damage at a radius exceeding 240 km just through the shock wave. Pyroclastic flows would easily travel 70 km with Mega-tsunamis that would exceed 100 m in height. The explosions could be smaller or bigger but regardless it would be horrible.

However, the magma type for Ischia may not be well suited for my postulated cause for this extreme model and still there could be many more unknown variables for the Krakatoa eruption that Ischia just can’t reproduce. Even if this volcano were to somehow erupt soon, the way this model would work the volcano would still produce abundant warning has high level eruptions would weaken the plug over months giving ample time for preparation before the climax.

This volcano easily has the potential to make a massive eruption, no one has really gathered enough information for us to give a reasonable estimation of a worst cast scenario. This volcano isn’t like Campi Flegrei or Laguna del Maule in the sense that it is not restless nor has it experience unrest recently. This volcano isn’t a phantom caldera either, the plug would have to be broken down over months of eruptions and the activity that would lead to an eruption would be as clear as a glass of water. Unfortunately, if this volcano were to erupt following this model even slightly it would cause a great deal of death.

Model Beta

Primary concerns: Tsunamis, Ashfall, Long term Lahars

Specialized concerns: Super pyroclastic flows, High altitude injection

(Probable) Location: Lake filled caldera with highly viscous magma

Based on Haetepe eruption


This model is based on much less well known but still an extremely violent eruption. The eruption had unusually high ejection rates even for its size. This eruption was a low end VEI 7 but it had ejection rates at the climax comparable to the postulated ejection rates of a VEI 8 eruption. Ejection rates greater than 1.0×109 kg per second happened with this eruption. This eruption is probably one of the few known eruptions to likely have an eruptive column within the lower area of the mesosphere.

This model’s calling card would be incredibly energetic pyroclastic flows born from the massive column collapse. The pyroclastic flows in this eruption covered an area of 20,000 km2 which is an area larger then most metropolitan areas and overran mountains in valleys. As such the flows probably could reach greater distances on a flat terrain.

With a range of more 30 km in even more modest VEI 6 eruptions this model is EXTREMELY dangerous to unsuspecting cities.. The reason for the violence produced by this eruption is still not really studied even though it would seem to be a special case.


I believe what caused the astounding violence of the eruption was a mixture of various conditions, the most obvious being that the eruption was rhyolite in make up, and rhyolite is very explosive. I believe water played a role but not directly like the Krakatoa eruption. I believe the build up to the eruption caused a great deal of steam explosions as well as Vulcanian eruptions before the main phase. That would weakened the inherently fragile nature of the rock above Taupo. The uplift and seismic activity that would have preceded the eruption could’ve increased the frequency of water related eruptions. The strong VEI 4-5 explosions preceding the climax would’ve greatly decreased the power of the system to hold back the energy. I believe that there was some type of external variable that triggered a runaway release of the energy and this could be a powerful volcanic explosion or a major tectonic earthquake

Unlike the Krakatoa eruption, I believe the energy was always there and it was the time-frame of the release that matters in this scenario.

In order to get an idea of my hypothesis, let’s imagine a high pressure container, a politician, wondering why there is concern with the system, decides to stab the container. The less volatile material comes out in reasonable pace all but after that material is depleted, the volatile material starts to come out triggering an explosion which ruptures the container leading to a great deal of material being released in a short time frame.

This hypothesis is a bit more simple then then former but do not mistake this model’s potential. Unfortunately this model has a higher probability of being a Phantom caldera. Rhyolitic eruptions can be quick, and there doesn’t have to be a strong plug, no one may know that a volcano has this potential without proper study, and if some countries can’t make adequate preparations for a VEI 3 eruption then an eruption like this is likely to be handled very poorly.

Past candidate

The town of Kodiak covered in ash during the 1912 Novarupta eruption. (Wikipedia)

The past candidate happened really recently, it was not as big as Haetape but just as infamous. Her name? Novarupta. I already have an idea of what you are thinking, how does Novarupta compare in violence to Taupo? The answer is simple, because Novarupta was a beast for its size. Despite being just a low end VEI 6 the pyroclastic flows remained hot for 15 years in one of coldest region on the earth. Pinatubo despite being in same class didn’t bring feats anywhere close to that. Novarupta had ejection rates up to 2.0 X 108 kg a second[3]! For reference that is comparable to the eruption of Tambora[4] and this just proves that this model, unlike the former, is not relegated to a specific location.


This type of eruption isn’t what you’d call common but not exceedingly rare either. There could actually be a deal more candidates for this model in history that haven’t garnered research. With some volcanoes, million and sometimes tens of millions live within 100 km of a dangerous caldera. The fact that violent eruptions like this haven’t threatened populated regions in 1000s of years would mean that this model may never be considered for a caldera in severe unrest. Worse still that the activity that volcano undergoes may not accurately portray the potential the system has.

If this type of eruption were to happen in an area of frequent rainfall. The lahars would last a span of years or decades and if the flows have a range 30 km or more than you can imagine how far the lahars could go. If the column were to collapse over a lake then you can expect a tsunami as tall as the lake is deep. There is one potential candidate for this model and thankfully it hasn’t experienced any unrest recently.


Coatepeque Caldera. sourced from wikipedia

The Coatepeque Caldera is one of those volcanoes that hardly anyone outside of volcanology really knows but this volcano is dangerous as it is surrounded at a range of 5 km with 616,310 people and at a range of 100 km, 6,486,011 people. There has been basaltic flows but not a great deal by any means besides this there is not a lot of study dedicated to this volcano, so much so that InSar data is lacking.[5} By no means am I saying that this volcano is an imminent threat but this is a terrible position. The recent mafic volcanism shows that this system is still healthy and we’d want an unhealthy system.

If this volcano were to follow the model as a low end VEI 6 then extensive evacuations would have to be made to avoid death. If a VEI 7 eruption were to take place under this model then there would be millions of deaths unless the best preparations are made and with the area’s history of being ignored and underestimated I wouldn’t have much optimism in government response. As I’ve said there has been no unrest recorded and even with a lack of InSar data, the chances of this happening in the near future are not something to lose sleep over.

Model Gamma

Primary concerns: Ashfall, pyroclastic flows, landslides

Specialized concerns: Lateral blast, Massive lahars, Debris flow

Location: Glacier covered stratovolcano

Based on Mt St Helens eruption of 1980

I don’t really think this model needs any introduction or discussion so I’ll just leave it at that and say that Mt Rainier is the most dangerous candidate. So I’ll see you next month. I quietly out walk of my room to get some dinner but my intuition drags me back to the laptop.

This model is the most frequent of the three and the one that even none volcano wise people are familiar with as it is based on the most infamous eruption of all time. The glacier covered stratovolcano has magma forced in it, forming a dome that is not the product of an eruption, the dome collapses leading to a landslide and lateral blast.

Mt Rainier is arguably the most dangerous volcano in North America, and this volcano has already had eruptions that follows this model. I legitimately have no words for this volcano, I mean what could I say that hasn’t already been said? This system is the most dangerous system for lahars, if it were to erupt again its lahars could threaten more then 88,000 people. Once again this volcano is the most dangerous volcano…for lahars but what about the other calling card?

The lateral blast isn’t really the first thing that pops into one’s head concerning this model, usually, it’s landslides or mudflows but if one were to isolate those risks and only acknowledge the lateral blast; you realize that Mt Rainier is not as bad as another glacier covered volcano in North America.

Pico de Orizaba is one of the most dangerous volcanoes in North America in my opinion, and the perfect rival for Mt Rainier. If one was to look at just the volcanoes one would confused. Pico de Orizaba is not much bigger then Mt Rainier nor does it have a much scarier history so what gives?

This is how many people live within 30 km of Mt Rainier: 3,187

This is how many people live within 30 km of Pico de Orizaba: 759,747

For reference, the lateral blast for the eruption for Mt St Helens killed everything within 30 km in a single direction. Not only that Pico de Orizaba also has a history of producing large eruptions akin to the scale of St Helens, and you could only imagine an eruption of that size in this area. Pico de Orizaba has frequent smaller eruptions and has history has shown us, a volcano doesn’t have to be large to be deadly. The Trans Mexican volcanic belt has a plethora of dangerous candidates for St Helens style lateral blasts most of which would threaten more people than Rainier.


77 thoughts on “Typing eruptions

  1. Interesting way to look at Volcanoes…… Wish i knew more/ i could comment more. Thanks for the article and will be reading the other comments. Best!motsfo; sitting in 40F degree weather which is very warm for Alaska. Thanks Japan for sending us Your leftover typhoons.

  2. Thank you.
    Also, thank you for mentioning Ischia.
    Looks so scenic, so pretty, just a harmless old volcanic plug like Scotland’s Edinburgh pair. But it isn’t.

    A B-Movie mega-monster slowly rising from the deep, it has ‘wept’ lava etc past the plug in several historical incidents. Getting most folk off will depend on the local authorities taking sufficient fright before those war-drums begin their harmonic tremor …

    But, if the sea gets in, if Ischia goes fully phreatomagmatic, ouch. The air-blast would trash Naples. Tsunami probably wallops entire Tyrrhenian Sea coast, plus shore-lines from Cannes to Tunis. Given they’re not accustomed to significant tides or storm surges, this would be very, very bad…

    What a series of non-trivial tsunamis do to the facing toe of Stromboli’s NW slump will be interesting, though preferably watched from orbit…

    FWIW, just found this really big open-access article…
    Multiple natural hazards at volcanic islands: a review for the Ischia volcano (Italy)

    • Sorry, did you just say that there have been emissions of lava adjacent too ischia in historic (~2000yrs) times?

      That’s very scary indeed given its location and historic movement thus far.

      • The paper states
        “Since 2.9 ka, about 34 effusive and explosive eruptions took place”
        which is about one per century but activity is intermittent. The last lava flow was in 1302.

        • The last eruptions seem very small though, there was a VEI 3 in ~150 AD but I can’t find anything more remarkable than that. It would still do quite some damage since it is a very heavily populated island with 60000 people

  3. Grotesquely OT; Re: “The Black Swan”

    The term “Black Swan” is used from time to time here on VolcanoCafe. To clarify things for the less than frequent traveler here, a short explanation.

    The Black Swan was the title of a 2007 book by Nassim Nicholas Taleb. The term, as defined by Taleb, is for an event that meets 3 requirements.

    1) It is statistically improbable. So much so, that it’s chances for occurring are wrongfully assumed to be zero.

    2) It is profound in it’s effects when it occurs.

    3) It is logically explained away after the fact. “If only we had considered {whatever}”

    In his later book, “Anti Fragile” Taleb takes issue with the Intelligent yet Idiot aspect of the adoption of the term “Black Swan” usually by those appearing on financial talk programs as a buzzword to make themselves appear more in tune with what they are talking about. In general, “Black Swan” events as they use the term, are not Black Swans. Remember part one of the definition. NO ONE expects it. Truth be told, there can be someone who can see the problem coming down the pike. Take for example, Kiyoo Mogi. Mogi is the scientist who came up with what is known as the Mogi Model. Using GPS displacement data, you can determine the size of a magma intrusion into an edifice by careful examination of the displacement data on the surface. He saw an issue with the displacement data along the Japanese coast and came to the conclusion that there was potentially a very bad thing going to occur there. Technically, his seeing the issue made the 2011 Tōhoku earthquake and tsunami NOT a black swan. But since no one listened to him, that made it effectively a Black Swan. It also fit into the bad Sci-Fi meme of a scientist not being listened to when he sees something bad soon to occur. {Life imitates art?}

    Anyway, I am wholly unqualified to talk about this on a mathematical level, though I did successfully pass Quality Control Inspector training prior to a yard period for a ship I was stationed on. They touched quite heavily on Objective Quality Evidence and the work of Dr. W. Edwards Deming. Deming took his statistical methods to Japan and fostered a paradigm shift in Auto Production. That is one of the reasons that they now have have a much improved reputation as compared to earlier vehicles.

    As used on VC, “Black Swan” carries two separate meanings. One is as intended by Taleb. The other is sort of a reverse Black Swan. In both cases, the term means that the event alluded to is totally out of character for the way a system typically responds or acts. My personal example of that, is Katla. Katla’s eruptions are so frequent, that it should have already gone off by now. In reality, pretty much ALL volcanoes have one sort of anthropomorphic trait. They hate statistics. In other words, you can not reliably quote some statistic about how often one goes off and have any measure of certainty that it will happen. Like earthquakes, what governs their behavior is a chaotic process. Chaotic processes can only really be described on long term measures.

    “In general, it goes off this often” doesn’t mean that something is overdue. In essence, they will go off when they are ready to go off. The gravitational effects of the Sun-Earth-Moon can possibly cause an event to go off a second of so early or late. The gravitational forces are that incredibly small compared to the geological forces at play.

    • I’ve actually looked at the numbers. The effect, if any, is buried in statistical noise.

      Many thanks to Jack@Finland for pointing that out. He stated that anything not exceeding 2 SD above Gaussian noise probably isn’t worth looking at. My signal was maybe 0.5 SD.

      Note: SD refers to “Sigma” or standard deviation. It is one of the metrics used in looking at data with a “Normal Distribution” curve AKA Guassian Curve. This is important because that curve is used a lot in looking at the behavior of natural systems. It is also one of the things that Taleb railed against in his Black Swan book. Per Taleb, it does not faithfully measure probability on the ends of the curve.

      Example, the 2008 Financial crisis was principally caused by a formula for estimating risk called a Gaussian Cupola with certain financial instruments. “Mortgage Backed Securities.” David X. Li’s formula was a method of rolling in several Gaussian risk curves into a single index that was widely adopted by the Financial manages.

      “For five years, Li’s formula, known as a Gaussian copula function, looked like an unambiguously positive breakthrough, a piece of financial technology that allowed hugely complex risks to be modeled with more ease and accuracy than ever before. With his brilliant spark of mathematical legerdemain, Li made it possible for traders to sell vast quantities of new securities, expanding financial markets to unimaginable levels.

      His method was adopted by everybody from bond investors and Wall Street banks to ratings agencies and regulators. And it became so deeply entrenched—and was making people so much money—that warnings about its limitations were largely ignored.”

      The Formula that Killed Wallstreet.

      The fun bit? Li’s disappeared back to China before everything blew up. Poling around a bit, it appears that he works for AIG now. That seems appropriate. His formula was based off of actuarial science and survival functions.

      His own basic premiss about his formula was that it works great, except when it doesn’t. So, you can’t really blame him for the market tanking back then. The key is that the financial guru’s had no idea what made the thing work so they could not see where it was going off track when it slipped the rails.

      From a Wikipedia Article on Li: “According to Nassim Nicholas Taleb, “People got very excited about the Gaussian copula because of its mathematical elegance, but the thing never worked. Co-association between securities is not measurable using correlation”

      • That’s why I hate when someone uses the term overdue I just hate that. That just shows how little they know about volcano and other geological process. Anyways I hope katla give me a nice option for the future.

        • Its like when people observe a roulette wheel and say that No5 hasn’t come up for 200 spins so there is a much higher chance of it coming up next. Its the problem that people have in their heads that ‘random’ means ‘even’ and why stars are randomly distributed in pretty constellations that look like structure….

          • The distribution of stars is not entirely random – the seven sisters really are siblings (and not seven). Translated to volcanoes: the fact that two of the three largest eruptions of the past millennium happened within sight of each other is perhaps not entirely a coincidence. But otherwise, yes, on human time scales volcanoes are random events.

    • The quake / tsunami that wrecked the Japanese nuclear plant has been claimed as an archetypal ‘Black Swan’. True, the tsunami was rather bigger than the ‘Richters’ suggested, but there’s a catch…

      IIRC, plant staff frequented a traditional restaurant at a trad fishing port along the coast. ‘Tis said, beside that establishment, there was a well-kept, carefully tended memorial to the victims of a historic tsunami, with a ‘tide line’ that was significantly higher than the nuclear plant’s sea-wall…

      Corollary to ‘Wisdom of Crowds’ is common ‘Stupidity of Committees’, which often have a nasty habit of ignoring outliers. Unless that’s Richard Feynman with his chilled, so non-bouncy piece of SRB O-ring…

      • For a time. I had the idea that the seawall was designed under a cost-benefit scenario that only allowed for 85% of the possible tsunamis. It turned out I was wrong. Albert found that the seawall was built to handle the largest tsunami seen in the Pacific basin. Sound reasoning, but inadequate to handle the statistical distribution of what was possible.

        • Nuclear power station require water for cooling so they are build near water but obviously not in water. Fukushima was build 11 meters above sea level, which was thought to be safe because the 1960 earthquake had caused a tsunami 3 meters high in the locality, and that was about the largest earthquake physically possible. The water pumps were build 4 meters above sea. Later the pumps were sealed against flooding since it was realized that these were the weak spot. The 2011 tsunami was 15 meters high. It destroyed the pumps, and also the back-up diesel generators which I think were in the same building. There was battery back-up power in the main building but these too were taken out by the flooding. With no pumps and no time to achieve cold shutdown the reactors melted.

          In the long run, the biggest risk isn’t from this. It is from sea level rise which will bring all power stations closer to sea level. Therefore in the future the black swans will become grey. Damaging events will become more likely everywhere. Except of course in places that think that global heating and sea level rise does not apply to them – they are the very biggest risks.

          • Yes, and why nobody see this I am at a loss to explain. Its my opinion that sea water rises will be significantly faster than everyone expects and we should already have detailed plans of how to withstand a 5m rise in sea levels.

            One day these will be needed but sadly many will take 30 years to build.

          • “There was battery back-up power in the main building but these too were taken out by the flooding.”

            Early reports from NHK stated that the operators circumvented the automatic SCRAM process that was triggered by the quake. It’s likely that they were trying to maintain power to the cooling systems though the tsunami had not arrived yet.

          • The external power supply was lost in the earthquake. They started up the diesel generators. While the turbines were out, they routed the steam circuit through the condensers to maintain cooling (no idea how). That was the situation when the tsunami arrived, 40 minutes later.

          • The tsunami events in Indonesia (2004) and Jaoan (2011) were not so different in terms of wave height and power. So, Japan should have expected that such tsunamis could happen anywhere along the Pacific ring of fire, including Japan itself.

            The Atlantic has even seen a massive tsunami from Portugal in 1755, so if such event happens outside of a proper subduction zone, one should expect such events also at the stereotypical subduction zone in Japan.

            Tsunamis also travel across the ocean. For example, the tsunami in 1755 also caused great destruction off the coast of Brazil. As well as the 2004 tsunami caused destruction off the coast of India and east Africa.

            Those are countries which never expect a tsunami, yet they can happen.

            Its not really a black swan event. Its a grey swan event.

          • In hindsight, yes. The 3 meters came from local measurements. Of course tsunamis can reach far higher than that, but the locality of Fukushima lacks the geography for this. They thought. By 2004 it was realized that there was an issue here. That is why the pumps were sealed. The company got a reprieve from the government about further flood proofing. A report was ready that would have forced the company to take more measures, but this was only due to come out a few months after the 2011 tsunami hit.

            The bottom line: companies do not like to take actions against low probability events. For the company, it is not cost effective. The most damage that can happen is that the company goes out of business, so the highest cost equals the value of the company. For society, the costs can be very much higher. That is why risk management has to come from the government. But governments find it hard to go against major companies. The US is at the forefront of this. Take the mining industry. They are legally allowed to use bankruptcy as a way to get out of clean-up costs. The mines put all liabilities on one sub-company, let that go bust, and walks away with the money and no liability.

            And now we have the problem that companies are beginning to take over the government. That happened in South Africa, may happen in the UK and in my personal opinion has already happened in the US as well.

          • “Black or Grey” determination is in effect, a logical effort to explain away the event before it happens. I’m okay with the idea, it serves to break the event down into manageable chunks.

        • Also because nuclear energy brings big risks, I would dramatically limit the number of nuclear power stations to one or two per country, and focus instead in improving the efficiency of our appliances and saving power needs. Our civilizations needs a fair degree of improved electrical efficiency.

          For example, compare how much energy does a fancy smartphone uses compare to a plasma screen: the smartphone uses far less energy and does plenty more.

          LED lights use far less than conventional lamps.

          Things like electric cars should be shared by sharing schemes rather than personal ownership (I currently do this and it works perfectly), and promoting good grids of rail public transport

          These third examples show how much energy we have been wasting.
          We could probably reduce our power grid by a third or more, and cut the number of nuclear and coal power stations this way.

          • Depending on the size of the country, having just one or two npp’s might lead to big losses in transmission. But apart from that you’re right. It takes specialist skills to run them safely, and its easier to have that in fewer places.
            Fukushima type reactors have a special mini steam turbine inside the containment that uses the steam from residual heat to push water into the core.

    • More often, we have what are known as “gray rhinos” as opposed to black swans. Gray Rhinos are big things that are relatively obvious, but get overlooked or ignored until they start causing lots of damage.

      While one could reasonably call the financial crisis a black swan, a lot of it was more gray rhinos within a relatively obvious real estate bubble. I tend to believe personally that there are a few noteworthy gray rhinos getting ignored right now around the world in various areas (not necessarily volcanology), but I’ll spare discussion of those since this is a volcano blog.

      There is a great book that I think is actually better from a statistical methodology of analyzing “black swans”, called the “(mis)behavior of markets” by famous fractal mathemetician Benoit Mandelbrot.

      In short, gaussian bell curves are often not particularly representative of realities and “black swans”

      With regard to volcanoes, I think the big issue is that sample sizes are just far too small to form any reasonable boundary on what can or can’t be predicted. The second issue is that the inputs into the system are subject to change.

      Take for example Katla (as mentioned). It has been very regular, perhaps one of the few volcanoes on earth you could sample a data set for 10000 years and have any degree of a confidence interval based on that data. But the above only works so long as the inputs into the Katla volcanic system remain the same. If deep magmatic input were to surge south from Vatnajokull and more towards Katla, we would start to see different activity at Katla. Or alternatively, if the Godabunga cryptodome were to overtake Katla as the primary spot of intrusion, that could change things significantly. If you were to model that based on statistics, the bell probability would be off the charts based on what we know.

      The issue is that our knowledge of the inputs of the system is flawed, and a lot of bell-curve standard deviation models can’t account for dynamic or changing systems.

      • I like the Rhino analogy. Hugely powerful and generally not an issue until they run amok.

  4. Excellent article, Tallis. That’s a very interesting way of looking at eruptions.
    Was the last eruption at Nova the flank eruption-caldera deepening Bardabunga-alike one, or was there another? But that was the Valley of a Ten Thousand Smokes thought wasn’t it?

    Speaking of Black Swans; one of the last albums the late and very great Bert Jansch released was called exactly that.

    Thanks, now. All the best.

      • A lava dome filled Nova’s crater as the last event of the 1912 eruption.

        • Although arguably the last -or equal last, the relative timings are uncertain- event was the extrusion, and subsequent partial destruction, of a dome on the floor of the Katmai caldera, the remmains long since submerged

    • Whilst trying to understand Tallis’s angle on eruption types you throw out Bert Jansch references ?
      Rather topical in the household right now, as I’m recording a song with a strong Jansch/Renbourn/John Martyn influence…As my daughter heads out to commence recording of her debut album, the headline of which is a song she wrote after hearing Light Flight.

      Ischia is one I’ve read about (mostly here), but I really hadn’t given it too much thought. I’m particularly curious about the potential impact of any tsunami generated from Ischia upon the flanks of Stromboli.

      How credible is that? Can anyone say?

  5. This is the most recent Sentinel image of Agung, from two days ago. The crater has been partly filled with lava over the past year or so. No evidence for any hot spots. The crater is some 800 meter wide, I think – someone may have a better number. The ejecta blanket covers some 3 square kilometer.

    And here is the same image showing a wider area. Agung is surprisingly tiny compared to the two flanking calderas.

    • Impressed at the number of sub-calderas and craters in the larger image, I can´t begin to imagine what the plumbing system looks like,

  6. Mauna Loa is having a tantrum. The shaking has now been going for a few days, and today there were already five M~2 quakes at the summit. This is ALEP, on the north slope

    Now this is one volcano which is strongly overdue. May it remain so forever.

    • “May it remain so forever” No please no, it has been almost 20 years of unrest and slow inflation already! this feels like an endless wait, just give us a summit eruption or something!

      It is not Mauna Loa’s fault though, this is a great example of why the term overdue shouldn’t be used so lightly as it is a very simplistic view and it tends to make it wrong.

      When one looks at the eruption frequency over time the current dormancy since 1984 is by far the longest on record so one may think it is overdue so it should erupt soon or do a really big eruption. Problem is that things are far more complex, Kilauea and Mauna Loa have an inverse relationship in supply that arises from a shared supply and a rift feedback mechanism on the activity of the volcanoes so that when one takes this into account the low activity of Mauna Loa during the last decades is expectable as the volcano magma-starves from the hyperactivity of Kilauea. And for all we know the eruption could still be years away (or not), but certainly not overdue rather right on schedule.

  7. A small swarm in Greip , but much larger in the magnitude . Would these still be magma quakes or both tectonic and magnetic in nature??

    • Exciting too 🙂

      0.3 17.9km

      0.6 15.6km

      0.5 17.2km

      0.5 17.7km

      1.6 19km

      1.2 18.3km

      1.5 18.9km

      1.3 10.9km

      1.4 16.1km

      I noticed BB has a star. Excuse if someone has mentioned it.

      At IMO:
      Saturday 26.10.2019 17:14:14 64.644 -17.365 1.3 km 3.5 99.0 7.8 km E of Bárðarbunga

        • Greip needs a drilling project and seismic blasting to image its magma system

        • Why is there a negative gravity density annomanly in Grimsvötn?
          This is a mystery there is a magma chamber there… but it dont showup on that method of survey… why?

        • It just occurred to me that maybe the plume center is actually located between Bardarbunga, Greip and Grimsvotn, and more accurately somewhere between . Which would explain the weird configuration of all those volcanic system swarms radiating away from this location (and also that of Hamarinn).

          For instance, Grimsvotn fissure swarm goes only southwest, and Hamarinn too. Whilst Bardarbunga goes more northwards. And then there is weird Greip in there.

          If the plume is pressing under Gjalp and Greip it could explain what is happening.

          Whatever is under there, it seems that we are quickly heading towards another significant eruption in soon. Somewhere around Vatnajokull.

          Those deep quakes in Greip mean a lot of magmatic movement at depths of under 20km.

        • About 30 quakes registered by IMO in past two days. A lot considering the depth.

  8. perhaps drilling will be not useful… who knows? Perhaps its going to reach the surface (i.e. to erupt) one day- lol

  9. In other parts of the world:

    The Japan Meteorological Agency (JMA) has upgraded Volcano Alert Level for Kuchinoerabujima from 2 to 3 at 15:15 UTC on October 27, 2019 (00:15 JST, October 28). Kuchinoerabujima is a volcano island in Satsunan Islands, Japan, about 120 km (75 miles) SSW of Kagoshima city and about 1 100 km (685 miles) WSW of Tokyo. A series of eruptions with pyroclastic flows were recorded in December 2018 and January 2019.

  10. There has been 2 Magnitude 6 earthquakes in the past two days is it possible that these are foreshocks to a large earthquake?

    • I assume you mean Mindao? The in-land region has not had earthquakes of this size for almost a century. It is probably capable of M7 but not much larger. The main events are near or off the coast.

    • Not quite, as this quake is similar in size to the previous ones, not larger. It is listed as M6.5, and the previous ones were M6.4, 5.8 and 6.6. But clearly the swarm isn’t over yet. In hindsight, it started with an M5 in July which could qualify as a foreshock. There is an enhanced risk of an M7 event for 1 to a fews days after this. The quake swarm has been progressing to the north east. Two years ago there was a similar swarm about 150 km north, but these quakes were around M5.5

      • Hi. new on site so see if it work 🙂 Was thinking if all this “Banging” quakes in philippines can desturbe Apo volcano.? I know it havent eroupt in near times. And not much known about it.

        You’re welcome. First comments are normally held by the system for approval. Future comments should appear without delay – admin

        • The volcano should be ok. It is fairly far from the earthquake swarm, and these earthquakes are too small to trouble any magma chamber.

          This is a recent Sentinel image of Apo. It is very green! That suggest there has been volcanic eruptions since a long time.

      • It was posted at 6.8 and quickly adjusted down to 6.5 (hence the choice of words). Everything else you write stands of course.

        Any chance old Mt. Apo could wake up?

        • As you said, on USGS the quake was initially listed as 6.8 before settling to 6.5 some time later. The 6.8 initial estimate was also in the PTWC report.

          Preliminary (PTWC)
          Origin Time: 30 Oct 2019 3:11 pm HST
          Magnitude: 6.8 Mwp
          (reviewed by PTWC)

    • Yo, dude. What is you talking about? (For clarification, I am not Tallis. Just his annoying sister and I don’t understand where this came from lol)

      • Well, suddenly this whole site is nearly inactive. I’ve never seen it this quiet in at least five entire years. So I figured someone important must have died to make everyone else so solemn and/or depressed …

        • It goes in waves. I promise the site will wake up if something like Grímsvötn or Mauna Loa decides to do its thing. Personally, I’m just overloaded with work and other stuff going on. I saw a new article just popped up about Greip, heading over there…

Comments are closed.