Ten volcanoes with super-eruption potential: Part III

Here comes the conclusion to the series, the 3 volcanoes that I considered the likeliest to produce a VEI-8 eruption.

3. Calabozos and neighbours (Chile)

This volcano is located in Chile. It forms part of a little known, little studied, silicic flare-up of the Southern Andes Volcanic Zone. Steepening of the subducting Payenia Slab gave rise to massive amounts of basalt lava flows and shields in the back-arc, while large caldera systems erupted from the arc. Calderas as big as the 20×30 km Varvarco volcano, of which almost nothing is known about. The Payenia flare-up is located to the north of another group of calderas that “died” at around 1.6 Ma, and to the south of a new flare-up area that is just starting to pick-up some strength. This is as if the subduction steepening was propagating northwards sequentially.

The Calabozos system became active at around 2 Ma, creating a 300-500 meter thick plateau of andesitic lava flows, a volcanic field in all likelihood. It’s possible that this reflects the formation of a large zone of molten andesite, similar to the Altiplano-Puna Magma Body, but much smaller. The Calabozos Caldera collapsed into a 26×14 km depression during large eruptions at 0.8, 0.3 and 0.15 Ma, each around 500 km3 in volume, but these should be considered minimum estimates that do not include distal ashfall. At the same time monogenetic vents and 2 stratovolcanoes, Cerro Azul and the imposing Descabezado Grande, erupted andesites over an area to the west. This westward migration would be attributed to the ongoing slab rollback (steepening of subduction).

Only recently the area to the west of Calabozos has shifted from andesitic to rhyodacitic volcanism, seemingly during the Holocene. Apart from some very small volume eroded flows, silicic activity is formed by 2 late prehistoric lava flows that look youthful, and eruptions of 1760 and 1846-1932. This rhyodacite magma body is 5-7 km deep (from thermobarometry studies), which is a similar depth to the magma bodies of most large calderas. It must have coalesced with the immediately adjacent Calabozos, resulting in a 30×25 km magma-mush body that is twice as big as the older Calabozos system which produced previous collapses.

Map of the Calabozos Caldera area. Holocene volcanoes are shown in green (basaltic-andesite) and red (rhyodacites). Created in Google Earth.

It was 1846, an arriero was riding through the pass between Cerro Azul and Descabezado Grande mountains. He noticed nothing strange, but a few hours later, with not the smallest earthquake whatsoever, lava started flowing out effusively from the pass, and would come to erupt 5 km3, inundating nearby valleys. When a local was asked about the name of the new volcano, he said QUIen SAbe PUes, which means: who knows? And hence it was given the name Quizapú. The conduit of Quizapú remained open, and starting in 1907 it was having small gas and ash explosions, but with not enough pressure to overflow. However explosions grew gradually more severe and a gas jet effect in 1932 allowed it to eject another 5 km3 of dense magma in a plinian VEI 5-6 borderline event. Two months later the north flank of Descabezado Grande blew up, literally; it formed a hole a km across (Respiradero Crater). This explosion I suspect could be related to heating of the hydrothermal system following pressure loss of the volcano after the plinian eruption.

The size of the Calabozos Complex is 40×30 km, more than large enough to host a VEI-8. Much of it is probably molten already, as shown by large effusive eruptions like Quizapù or the Mondaca lava flow. Young basaltic-andesite eruptions next to Cerro Azul show that hot magma keeps rising into the system and making it grow further. On the other hand it lacks a ring of vents typical of more mature large calderas like nearby Laguna del Maule and eruptions are not as large as those of other volcanoes considered.

I think this Calabozos has got a lot of potential but it’s not ready just yet. And thankfully, its remoteness means that eruptions are unlikely to harm anyone.

White pumice from 1932 mantles the ground, lava flows from 1846 form a blocky wall and Descabezado Grande towers behind, By Jfbustos, Wikimedia.


2. Serdán-Oriental (Mexico)

Volcanoes evolve over time, and on a much grander scale whole volcanic arcs also transform over time. Among them the Trans-Mexican Volcanic Belt has one the most fascinating stories of them all, and I can only but scratch its surface here. México is the best place in the world to find the processes that lead into major silicic flare-ups taking place right before our eyes.

In the past México was the scenario to one of the greatest episodes of silicic volcanism the Earth has ever seen, the Sierra Madre Occidental silicic igneous province. After 20 million years ago the northern portion of it became the Gulf of California, while the southern became the Trans-Mexican Volcanic Belt and entered a period of flat-slab subduction.

But things have changed, the flat-slab is plunging down into the mantle (rollback), exposing a hydrated crust to the hot upwelling mantle. Extension opens grabens like Tepic-Zacoalco or Chapala. The crust has thickened already to 50 km below the eastern part of the belt. All those geologic processes that are inferred to have caused other active or long dead flare-ups are taking place right now below México.

The dazzling snow-covered stratovolcanoes like Popocatepetl (meaning smoking mountain) or Citlaltépetl (the highest volcano of North America) get most of the attention, and why wouldn’t they? They are prominent so that they dominate the landscape. But that would be missing the most unique volcanoes of Mexico, which are none other than its gigantic andesitic volcanic fields. Anywhere you go along the arc, they are there, volcanoes, visible as small conical hills to voluminous shields that rise a kilometre above the surrounding valleys. It is no wonder that some of the youngest new volcanoes of the world formed here in 1759 and 1943 (Paricutín).

Michoacán volcanic landscape. Photograph from Pah Strosahl.

The volcanic fields reveal that vast portions of the crust are melting. This creates the ideal conditions for large calderas to pop up. Enormous amounts of silica-rich melts are created within these melt zones by diverse differentiation mechanisms and supplied upwards to growing magma reservoirs, the nascent large calderas. But a batholith is not created in a day (nor in a million years) most of them are still immature, however there is one that has gone ahead. The first supereruption of México since the Sierra Madre Occidental might well come from the Serdán-Oriental Volcanic Field, also called Oriental Basin.

This vast volcanic field is big enough to host multiple large calderas, and actually it has got one already, Los Humeros, which collapsed in the 115 km3 DRE, Xaltipán Ignimbrite. A much bigger silicic volcano could be emerging further south, below the basin. A series of 6 large rhyolitic eruptions less than 25,000 years old seem have taken place along a curved line, delineating a melt body that must be, at least, 20×30 km across, putting it in the VEI-8 range.

Most calderas behave like trapdoors where one side of the ring-fault opens up and erupts while the other closes. Taupo for example usually erupts from its east side, while Yellowstone prefers to erupt from the west. Serdán-Oriental shows, perhaps, a nascent trapdoor caldera structure, although it is poorly defined because of the small number of eruptions so far.

Recent rhyolite eruptions within the basin are shown in red. Created in Google Earth.

The last major eruption took place around 20 AD. The ground was lifted up from within by 2 enormous lava domes, this rock carapace started to collapse into debris avalanches together with pyroclastic surges and lahars. This is thought to have caused the pre-hispanic populations, who lived in nearby villages, to flee to the city of Cantona further north. The population of Cantona may have increased from 28,000 to 52,000 people as a result. The volume of rhyolite extruded during the eruption was 10-11 km3 and formed the twin domes of Las Derrumbadas.

Las Derrumbadas, two veritable mountains formed in the blink of an eye. Image from to Bodofzt.

Another way volcanic activity has impacted local population is through obsidian trading which was what the economy of Cantona was mostly based on.

The city of Cantona is built atop recent lava flows from Los Humeros Caldera. A conical lava dome rises further ahead. From Comisión Mexicana de Filmaciones.

Action has just begun in this corner of Mexico. Continuing rollback should provide a powerful melting source for a few million years to come. Rifting could also take place like it did during the Sierra Madre Occidental event. The very big rhyolite magma body of Serdán-Oriental should, if anything, grow even more. How long will it take? It cold be tens of thousands, or hundreds of thousand of years (remember we are talking of geologic timescales here), but I do think it will get the VEI-8 eventually, and it might just be the first of many similarly huge calderas to emerge across México.

And in the meantime beware of maars. Those, often scenic craters, that can show up in a cornfield or the middle of a town, are highly explosive and destructive eruptions. The Oriental Basin has got a number of them, including rare examples of rhyolite maars, like Cerro Pinto, which produced pyroclastic currents reaching 18 km from its 2 craters and must be one of the most violent eruptions of its kind.

Alchichica, a basaltic maar crater in Serdán-Oriental. Maar craters are called Xalapascos in México. Image from Itazabi RL.



1. Okataina (New Zealand)

My number 1 choice brings us back to New Zealand. Okataina volcano is located at the northern end of the Central Taupo Volcanic Zone, which is known for its many calderas. Together with Taupo it has been the other major player during a resurgence of activity that has taken place during the last ~55,000 years. Okataina is located in a rifting arc, spreading at rates of 1-1.5 cm/year, which is somewhat faster than the East African Rift to give a reference. This is really important for how the activity of Okataina functions. Its last eruption in 1886, although much smaller than its usual eruptions, illustrates these processes.

Maori people used to build their villages next to hydrothermal fields. Hot springs are good for cooking food and pools of water provide relaxing baths. They were not aware though of the volcanic powers within the Earth sitting below their villages like a ticking bomb, they did not know their land as the powerful volcano it actually was.

At 12:15 AM on 10 June 1886, minor earthquakes started to be felt and ramp gradually in strength awakening people from their sleep. By 2 AM a black plume was erupting from the top of Tarawera Mountain. Soon after the mountain was torn open by a propagating fissure above a dyke intrusion and fiery fountains of basalt lava shot up to ~500 meters. Henry Roche was viewing the eruption from a high point:

“We then beheld the striking spectacle of a dark, flat-topped mountain more than a mile long, red hot along its crest, and surmounted by a wall of fire 1,500 feet high. Over this hung an immense volume of dense black smoke clouds through which the forked lightning flashed without ceasing.”

Painting by Charles Blomfield of the 1886 eruption from what he was told by witnesses.

The eruption had so far been harmless. The dyke however kept propagating to the southwest and intruded below Lake Rotomahana, an already hot and fractured hydrothermal system, was turned to steam by the hot basalt from below, while new fractures formed as the ground was pulled apart, decompressing the system. At 3:30 AM Lake Rotomahana uncorked in a series of explosions that sent pyroclastic currents over a radius of 6 km destroying a few Maori villages, leaving no survivors in most of them. 120 people died as a result of the unforeseen Rotomahana explosion. Even now with our knowledge and monitoring something like this would be very hard to foresee. Back then many people may not have understood what was happening.

Henry Burt gives a description of the event:

“As soon as Ruawahia stopped sending forth its terrible balls of flame a huge white cloud issued from the cup of Rotomahana, and heavy booming was heard, followed by dense volumes of white compressed steam from Rotomahana. It rose with terrific velocity, and seemed to be going towards Okaro lake… and the appearance it presented was something like a huge boiling cauldron, bubbling in all directions.”

By 6 AM the eruption was over. Smaller steam explosions had opened the Waimangu Valley, today occupied by large hot springs. The ground was fractured all the way to Waiotapu (indicates the subterranean path of the magma intrusion). In total a dyke intrusion of over 25 km long must have intruded along the tectonic rift.

The explosion at Rotomahana excavated a crater complex 2.5×5.5 km across and unfortunately destroyed the beautiful Pink and White terraces, and various other terraces and hot springs that existed here prior to the eruption.

An article I was reading from Ronald F. Keam explains how the Pink and White terraces formed and why they were unique in the world. An eruption 5700 years ago had dammed Tarawera River with rhyolite lava flows which caused a great lake to form east of Mount Tarawera. Several geysers were created during this high-stand, and later, as the river slowly eroded the lava barrier, the lake level gradually fell, geysers found themselves perched on steep slopes above the water. Eventually Rotomahana became separate from the rest of the lake. By this time the geysers were 30 meters above the water level and the waters they discharged precipitated silica to form silica terrace formations unique in world (they are different from the more common travertine terraces that are made from carbonates). The many resulting bluish pools of water were also found at adequate temperatures to have a relaxing warm bath in them.

Lake Rotomahana before its 1886 eruption. The Pink Terraces are visible right below, where 2 persons are standing. The White Terraces are seen across the lake to the left. Photograph taken by D. L. Mundy, 1875.

Today’s lake Rotomahana fills the craters created in the 1886 eruption. Mount Tarawera behind is formed by 1300 AD lava domes, and the fissure from 1886 can be seen running down its face. Photograph from Pseudopanax, Wikimedia.

Known to the Maori as Te Tarata (meaning “the tattooed rock”), the White Terraces, were particularly impressive. The top formed a crater 20-30 meters across filled by a turquoise-blue pool of boiling water and surrounded by high steep walls, probably excavated in the hydrothermal explosions that gave it birth. Te Tarata was a geyser that erupted frequently in diverse ways but sometimes as powerful jets, like this description I encountered:

“By 4 p.m. the action had become vigorous, and dense clouds of steam were given off; and at 5 p.m. the basin was half full, the water violently agitated. Suddenly a deep boom was heard and felt on the upper terrace, like an underground explosion, and the water in the basin was instantly lifted into a huge dome, from out of which shot vertically with enormous velocity a glistening fountain, the top of which was lost to sight in the dense mass of accompanying vapour, the broken waters falling in heavy showers around.”

Te Tarata, the White Terraces. Painting by Charles Blomfield.

However the magma-water interaction that created them would also bring their destruction. People who visited the location after the eruption found that the location of the terraces had become a massive crater. While there are still those who have been searching in hope of discovering any buried or submerged remnant of the terraces, I think that this search is hopeless. Early witnesses to the site agree that the terraces had been in the location that blew up.

The 1886 eruption shows many aspects of how volcanism in the Taupo Volcanic Zone may work: with sudden rifting events, steam-driven explosions and fissure eruptions. But it doesn’t represent Okataina well. Most previous eruptions from Okataina have erupted rhyolite from long fissures along either the Tarawera “Fissure Swarm” (I will borrow the term from Iceland) or the neighbouring Haroharo Fissure Swarm. Nine such eruptions are known in the last 25,000 years, they all start with VEI-5+ plinian or phreatomagmatic eruptions followed by the extrusion of the voluminous rhyolite lava flows and domes of up to 12 km3 for an 8000 years old eruption.

The last major event was the Kaharoa eruption around 1300 AD. It started with a fissure plinian eruption from Mount Tarawera that produced at least 15 km3 of ash together with steam (hydrothermal) explosions through a line to Waiotapu, forming a 25 km long stretch of erupting vents. It was followed by the formation of 3 lava domes that form the flat top of Mount Tarawera.

Other smaller eruptions like the one in 1886 are produced by dykes of basalt, previous events may have been hidden by the larger rhyolite flows and ashes. 2 odd examples are known that are outside the usual fissure swarms. A basaltic eruption 3400 years ago formed the Rotokawau line of craters. Another 10,000 years old eruption created a line of steam explosion craters south of Mount Tarawera at Lake Rerewhakaaitu, this one is an example of a dyke that didn’t breach the surface since the ejecta consists only of old rock.

Fissure system that fed basalt lava fountains in 1886. From Richard Waitt, USGS.

Okataina is large enough to host a VEI-8 caldera and the voluminous rhyolite eruptions reveal the existence of a sizable magma chamber. We know a lot about how the magma was stored over time and can place the moment this magma chamber was created. Interestingly it happened at a time Taupo was also undergoing great changes. Looking further the history of Taupo and Okataina appears to be connected to some degree.

Taupo, Okataina, and the entire group of calderas had been very quiet until around 55,000 or 45,000 years ago depending on the controversial estimates of the Rotoiti eruption at Okataina. The Rotoiti eruption was a caldera collapse with a volume of 370 km3, and no more than a few thousand years later, known from stratigraphy, Taupo started to produce plinian eruptions (and presumably lava flows that were later destroyed in the collapse). Magma from these plinian eruptions had been stored at shallow depths of around 4 km.

Studies of zircon crystals in Taupo magmas reveal that the volcano got a very large influx of melt that rapidly triggered the ~25,300 years old Oruanui VEI-8 eruption, which destroyed the storage of Taupo. Interestingly Okataina formed a shallow 4-5 km magma chamber around that time. An eruption ~25,200 years ago (basically contemporaneous) was the first to tap shallow storage, and initiated the series of voluminous fissure eruptions that continues to present without much change. There are some changes that took place separately in each volcano, like the collapse at Taupo in 200 AD. However the seemingly simultaneous change that both volcanoes experienced around 25,300 years suggests it was something more than a coincidence.

The details of how Taupo and Okataina seem to have received magma at the same time remain speculative. Storage at Okataina is right now very similar to storage at Taupo before Oruanui, could it be a pre-caldera stage? If it goes caldera it could very well be a VEI-8. Perhaps it’s a matter of when the next major regional influx of magma will reach Taupo and Okataina. A 3rd large influx could send Okataina into collapse like it happened with Oruanui, and with Rotoiti before that. It is impossible to know if or when that could happen but the possibility it could take place somewhere in the next several thousand years makes Okataina the volcano with the best cards to produce Earth’s next VEI-8 eruption. New Zealand blows again!

Painting of the Rotomahana Crater after its eruption. Painted by Charles Blomfield.

Information and research

On the Payenia flare-up and plateau basalts:




Okataina and Taupo:






Photographs and description of Lake Rotomahana geysers:


67 thoughts on “Ten volcanoes with super-eruption potential: Part III

  1. Ah, Tawawera. Destroyer of the great terraces, and creator of the truly monstrous Waimangu, the largest geyser to ever have been recorded by humans. 1,500 feet, or roughly 400 meters, tall eruptions. For reference, the tallest active geyser in the world, Yellowstone’s Steamboat Geyser, “only” reaches 380 feet, roughly 100 meters. Said geyser, despite it’s power, didn’t die by destroying its plumbing system, but instead lost a lot of water to a water table shift. The worlds largest hot springs are still in that area, however. I doubt we’ll ever see anything on the level of Waimangu again, but as a passionate geyser gazer, I sincerely hope that we do.

    • Waimangu Geysers where indeed awsome, they where so large that they looked like pheratomagmatic detonations rather than normal geysers. Always dirty with mud and sand and dark, huge eruptions.
      Totaly dwarf all other Geysers today, incredible these old photos are.

      • I think the weirdest part is that they somehow didn’t appear to be doing much damage to the plumbing system. if anything, prior to the water table shift it seemed the eruption were slowly gaining force. I’m genuinely curious as to how that’s possible, honestly.

  2. Well looks like I was right, but also wrong… 🙂
    I probably should have clarified when I considered Taupo as the number 1 candidate I was talking about the TVZ as a whole, very interesting post and rather a scary number 1 because it actually could conceivably erupt right now…

    • And you missed Calabozos by only 40 km. Laguna del Maule is a good candidate for a VEI-7 but not big enough for an 8, but was that a good guess anyway.

  3. While its not the number 1 spot, Serdan Oriental really has something kind of unnerving about it. The idea that a supervolcano isnt actually even a single volcano at all until it erupts, just a cluster of seemingly random massive lava domes that one day destroys itself. It is the same sort of feeling upon when lava first erupted in Leilani, that it was always there but no one really thought about it.

    Probably the biggest factor in this though is that Serdan isnt in the middle of nowhere, its actually pretty densely populated, and all of those people live on top of 1000+ km3 of rhyolite that can throw a Pinatubo suddenly from a random hole in the ground. Or a basalt lava flood apparently, theres a lot of massive flows just to the north that dwarf Holuhraun…

  4. This paper (https://www.frontiersin.org/articles/10.3389/feart.2014.00016/full) suggests the Toba eruption had a total bulk volume (ashfall and PDCs) of 13,200 km³ (5,300 km³ DRE)- 65% of this being ashfall, erupted in 15 hours, with the eruption rate at 5,000,000,000m³/s (a.k.a. a Pinatubo every 2 seconds)! The eruption column was 42 km high, thousands of kilometres wide, formed by multiple vents producing multiple columns merging, as an unimaginably-large fissure eruption. The total SO2 ejection was 1.7-3.5 Pg (petagrams!), 1,000x more than Tambora.

    So, that would make it a VEI 9 event! Is this paper reasonable? It seems…. excessive?!

    • Toba must have been an incredible sight! imagine all the volcanic ligthing.. I wonder how it looked like 😄basicaly a wall of angry grey cauliflower for a few thousands of kilometers.
      Pinatubos PDC s are miniscule in comparison

    • I had the same doubts when talking about Toba. The DRE volume seems perhaps too large for the caldera size, a DRE of 2800 km³ seems more reasonable.

      • If we use a magma density of ~2,300 kg/m3 for the Toba rhyolite along with a DRE of 2,800 km³ that still gives a mass of 6.44*10^15 Kg. That would be a magnitude of ~8.8 ……..much closer to a VEI of 9 than a minimal VEI 8, if we define VEI as the magnitude of the eruption rounded down to the nearest whole number. I find it hard to chase down the original information behind the volume estimate of the ignimbrite for the Toba event. The mass of the ignimbrite would depend on the density as well as the volume. Densely welded ignimbrite can exceed 2000 Kg/m³, while loose, unconsolidated ignimbrite could be more like 1000 Kg/m³, depending on crystal and lithic content. Also, substantial parts of the Toba ignimbrite were deposited in the Indian Ocean and Malacca Strait, adding problems for volume estimates. I wonder if it is still possible that the youngest Toba event could have a mass of 10^16 Kg…..making it technically a VEI 9.

        • I suppose you could call it a VEI 8+ as it seems the eruption was at the high end of VEI 8

    • I am highly sceptical. In their model, almost all the ash falls in the ocean. There is no data to compare with that. For the distal ash, that is very dependent on weather and the seasonal tropical jet. The eruption rate is far higher than what can sustain a column. They mention that, in fact, and propose that they can get around that by using the fissure to spread the eruption out but they don’t actually do those calculations. Extraordinary claims require extraordinary evidence. This falls some way short.

      And a final thought, to coordinate a 15 hour eruption along a 200-km ring fracture requires very fast magma flow. Otherwise one point will erupt and the other side does not know until a day later. The main eruption is believed to have lasted less than 10 days and that seems plausible, but I would need to be convinced that it can be done in 15 hours. Too many holes in the argument that still need filling.

      • One thing that I noticed about the results of that paper is that the estimated mass of the co-ignimbrite ash is greater than the supposed mass of the ignimbrite itself. My impression is that as a fraction of an entire eruption, co-ignimbrite ash tends to make up perhaps 25-50% of the total eruptive mass. The eruptions having the largest fraction of co-ignibrite ash seem to include those eruptions having large ratios of outflow sheets to caldera filling material, and eruptions involving crystal poor magma. This rule of thumb does not include fine grained vitric rich co-plinian ash that can be hard to distinguish from co-ignimbrite ash. If an eruption has more than 50% by mass of fine distal ash, the mass of that ash may be inflated by co-plinian ash. As the younger Toba event does not include a plinian phase, a result where co-ignimbrite ash makes up more than 50% of the total eruptive mass does seem suspect.

        • I had not spotted that. You are probably right that this puts some doubt on the numbers. More research needed!

          • Thank you Albert. As for hard research I have not even been able to find the original work about determining the mass or volume of the ignimbrite itself.

    • Yes, mjf….that does seem excessive. The following argument is not proof that the Costa et. al. paper is wrong, but it is my take about their numbers……..A key result of that paper is that the estimated mass of the co-ignimbrite ash is greater than the supposed mass of the ignimbrite itself. My impression is that as a fraction of an entire eruption, co-ignimbrite ash tends to make up perhaps 25-50% of the total eruptive mass of most ignimbrite forming eruptions. The eruptions having the largest fraction of co-ignimbrite ash seem to include those eruptions having large ratios of outflow sheets to caldera filling material, and eruptions involving crystal poor magma. This rule of thumb does not include eruptions with a large plinian phase. Fine grained vitric rich co-plinian ash can be hard to distinguish from co-ignimbrite ash. If an ignimbrite-forming eruption has more than 50% by mass of fine distal ash-fall, the mass of that ash may be inflated by co-plinian ash. As the younger Toba event does not include a plinian phase, a result where air fall ash fall makes up more than 50% of the total eruptive mass does seem suspect. Hence the calculated mass may be excessive.

  5. I can’t remember if anyone has posted a link to this paper?

    The cascading origin of the 2018 Kīlauea eruption and implications for future forecasting
    M. R. Patrick, B. F. Houghton, K. R. Anderson, M. P. Poland, E. Montgomery-Brown, I. Johanson, W. Thelen & T. Elias
    Nature Communications volume 11, Article number: 5646 (2020) Cite this article


    • I think you are the first.

      That’s a really good one! It gives a much more complete analysis of the situation than most other articles of the eruption.

      • Hector,
        It got me thinking.

        The way that they illustrate the Magma supply to the LERZ it is supplied at a depth that is equivalent to the difference in elevation from the summit to the elevation of the eruption (roughly 3000 feet figure 7). If we assume that is correct or figure 6 is more appropriate (deeper supply). Just trying to assume things that will keep the conduit full instead of draining due to altitude differences.

        Eruption rate 131 cubic yards a second, 3510 square feet per second

        If it erupted 1 foot a second from the conduit, that is about a 60’x60′ (feet) area of magma (ignoring pressure, I think it is smaller)

        So how does this still hot pipe (sill, dike, crack) of around 27 miles in length, change what the timing of the next eruption might be?


        • Figures 6 and 7 are not to scale. The depth of the conduit (in this case a horizontal pipe) is about 3 km (1,9 miles). This is known from studies that have located with precision the earthquakes related to the conduit pressure build-up.

          The dike feeding the 2018 eruption intruded from this pipe at a spot just uprift of Highway 130, not Pu’u’o’o like it’s widely assumed. Pu’u’o’o collapsed into a small dike uprift from the cone, then a second larger dike followed from near Highway 130. The dike extended from Highway 130 to the Puna Geothermal Venture, this interferogram shows its location:

          The place or timing when the pipe snaps and grows a new dike is very hard to know. It depends on the deep spreading of the volcano, which moves the flank and creates space for dikes to intrude, and on the summit of the volcano slowly sending magma down the conduit. Different sections of the rift conduit will have different levels of magma pressure and tension from the rift spreading. The area of Highway 130 and JOKA station is worth keeping an eye on, since it sends dikes the way to Leilani. HVO had underestimated the pressure build-up at JOKA that had been going on for ~6 years before the 2018 eruption and which could have been interpreted as a precursor to an eruption in the Leilani area.

          • The JOKA- highway 130 area was also the first place to reinflate after the eruption, beginning even while there was that lava in fissure 8 in early September.

            That stretch of the rift looks very similar to what the east Napau area looked like before the eruptions of the 1960s, and then Pu’u O’o formed above that, its definitely a scenario that could play out in the near future 🙂

        • Since the eruption ended HVO has revised a lot of the numbers and a lot of them turned out to underestimate sometimes by a huge degree.

          As of a few months ago, eruption rate was averaged at about 300 m3/s and at times was over double that, the depth of the channel was much higher than expected. The volume is 1.12 km3 with about half being below sea level at the end of the lava delta and as deep as 1.5 km under the surface. About 90% of the lava was erupted by fissure 8 in 2 months, it was not quite so voluminous as Holuhraun but it was erupted about 3x faster.

          The other thing that might have an effect on future activity, the lava erupted at fissure 8 in the last few weeks was pristine plume basalt that was never near the surface before and was dredged out of the deep summit, so the entire system was flushed out and replaced with new fresh magma in 2018.

  6. Great conclusion to the series. One of the big questions I’ve always asked myself on here is “what would a supervolcano look like before it has its first supervolcanic eruption”? I think you’ve done a great job in helping paint a good picture of what that may look like.

    First off, It’s interesting that you mention the Serdan Oriental field since I’ve always thought it looked like something potentially bigger than just the Los Humerous caldera to the north, especially with the rhyolitic lavas.

    Second, I’m curious. The logic that a broad volcanic field would likely pop up when there is wide-scale melt going on within the crust prior to potential VEI-8 events makes total sense. But do we have direct evidence of this occurring prior to volcanoes going…. supervolcano?

    The reason I ask this is because if you just browse around to most supervolcanic systems (Toba, TVZ, etc), you don’t see any visible volcanic fields like you do in Mexico. Is this just due to the age and progression of these systems + the likelihood that a lot of evidence of the volcanic field is buried?

    • Most likely it’s the age and evolution, but it’s also not technically true, at least if you go down to the vei7 crowd. heck, even yellowstone, with it’s separate flows and west thumb caldera might count as a volcanic field at this point. it’s just the point at which a lot of these volcanoes become rhyolitic that tends to no longer favor volcanic field activity, i assume. a lot of these fields, i think, lean a stage down in the rhyodacite stage. Might be wrong though

    • Yes, both volcanoes have evolved beyond their initial stages of andesitic volcanism and done a lot of landscaping that has buried the earliest lavas. At TVZ much of the surface has collapsed into one caldera or another and if not is covered in rhyolite lavas and ignimbrites, but it did start as a widespread area of mostly andesite volcanism that may have looked somewhat like México. Northern Tongariro is a small andesite volcanic field and perhaps the last surface remainder of how it may have looked like in the beginnings.

      In some places you can still see the large calderas together with active volcanic fields. At Long Valley for example. Calabozos and Laguna del Maule have volcanic fields immediately to their west, the direction in which these systems are expanding. At the Andean Puna the caldera complexes of Incahuasi, Lazufre and Cerro Galan are in close proximity to several volcanic fields.

      So problem with ignimbrite flare-ups is that they typically bury the earliest volcanic activity, that geologists studying them are usually more interested in the ignimbrites than what came before them and even when the initial volcanism is found it is hard to interpret exactly what it used to look like.

      México is a good place to study how large caldera systems start to develop. Volcanic fields include almost-purely andesitic like Michoacan-Guanajuato, mixed andesite-dacite like Chichinautzin, rhyolitic fields like Serdán Oriental to calderas like L

      • To calderas like Los Humeros…

        Some evolutionary trend can be seen that is something like this:

        Michoacán-Guanajuato > Chichinautzin > Serdán-Oriental > Los Humeros.

  7. So, generally speaking for a VEI8 to form it requires the thicker crust on top than your usual arc volcano, and a gradual shift to dacitic and rhyolite. Serdan Oriental is very interesting.

    Still think Awasa and Garibaldi are strong candidates, Awasa/Corbetti in particular has had bradyseism (lake level change), and recently built edifices in the north of the caldera, on the east rim and also a cone in the west. Fumarolic activity is also high, and the area has undergone a lot of extension and sort of sits at a junction between rifts.

  8. Great article. Thank you, Héctor.

    Got me thinking. To get the volume for a VEI8, a large volume of magma (rhyolite or dacite) must collect in the crust before the eruption, which would mean that there is unlikely to be a lot of eruptive activity at the site before the VEI8. Perhaps the next VEI8 would actually be a currently Unknown.

      • I think that is exactly what these volcanoes were, the ones without calderas are all unknown as the VEI 8 hasnt actually happened yet.

        As for extinct, probably not, because to actually keep all the magma hot you will need a massive heat source which will probably erupt somewhat often. I guess if it underlies a former andesitic complex then those andesite stratovolcanoes will be extinct but rhyolite could still erupt at that location, as at Cerro Azul and Quizapu.

        • Possibly, but to get enough rhyolite or dacite for the VEI8, it would take a while to build up in the crust, long enough to be considered extinct. The lower crust is melted by magma from the mantle.

          • Possibly in some cases, but as is shown in this article the only real requirement is that there is a lot of eruptible rhyolite and that doesnt necessarily mean that no eruptions can happen while that magma accumulates. The magma chamber can likely erupt easily with big domes and then keep going, the way that VEI 8s seem to happen in a lot of cases is probably by the collapse of the magma chamber simply from the chamber being too wide for the roof to support itself which will force magma out of the ring fault and it sort of accelerates from there, its not really a gigantic explosion

          • Theres actually quite a lot, just few of them are repeat offenders and a lot are extinct and invisible but theore presence is known from a deposit. Theres ptobably been at least 100 VEI 8s in the Cenozoic, there has been 2 in the last 100,000 years with a plausible case for a 3rd anyway. One thing too, TVZ and Snake River Plain are clustered, there is something like 20 VEI 8 or high 7 ignimbrites and deposits in the Snake River plain but the volcanoes seem to have their final eruption be their biggest, erasing the entire volcano then filling with basalt so its invisible. Yellowstone is still active so has not done this, at least not yet… 🙂

            In a lot of cases the magma chamber might never collapse, it doesnt have to but it could, like all of the areas in the Andes. As I hypothesised the actual eruption is not really directly explosive that part is just how rhyolite behaves at the surface under rapid decompression if it cant degas fast enough.
            If the chamber gets too big to support itself the edge cracks in a big way and the middle sinks, it forces a huge amount of magma out very fast, and then the magma fragments in the low pressure environment. No major plinian stage but the fragmentation and degassing will accelerate and fluidise the flow on the ground so it goes very far and probably pretty fast after a while. It would probably look like fluidised sand flowing, with a lot of dust and being very hot probably glowing. It probably wouldnt look like a pyroclastic flow though, its sort of in between that and a lava flow.

        • Chad, that makes sense to me. It would be interesting to know if stratocone building is truly finished yet at Cerro Azul and environs. After the large dacite effusion of 1846, the Quizapu vent stated erupting small volumes of basaltic-andesite in a series of strombolian eruptions in the 1920’s up until when the great plinian eruption of 1932 produced another large batch of dacite almost identical to the 1846 one. Apparently the plinian eruption closed out with another small volume of mafic scoria such as that erupted in the years leading up to the 1932 dacitic plinian event. The eruption of mafic magma at Quizapu makes me wonder if this particular system is not yet mature enough to generate a large silicic caldera. Perhaps it might even do some more stratocone building before evolving into a more mature silicic system.

          • Quizapú did behave somewhat like a stratocone, erupted repeatedly for a century from the same conduit. The way it first showed up was also unconventional, a person had just stood a few hours before the 1846 eruption next to where the volcano was going to form, but noticed nothing unusual, witnesses did no feel any earthquakes either, it was as if Quizapú simply melted its way through in 1846, doesn’t seem to have been dike fed. So Quizapú looks like a revival of the stratocone building of Cerro Azul. I don’t know what the implications of that would be though.

          • Quizapu is sort of a reverse volcano, it begins silently and gently and then gets more violent for a century before it goes out with a bang. It erupted at least 10 km3 of magma in its short life and the area is not a plume so probably there will be a long time before the next eruption like it in that area. It does make you think about such an eruption happening at any time though, just lava flowing out of the ground with no warning. I think today something like that would be picked up by instruments but that only matters if those instruments exist.

    • My thoughts precisely. However that doesn’t mean the initial stirrings will go unnoticed these days. More likely there will be decades of increasing unrest. I do wonder about the massive uplift under south western USA which must suggest an awful lot of something nasty is trying to get out.

      • I would expect seismic activity before an eruption and ground deformation – but how much, I would not like to guess. My guess would be that some tectonic event (crustal failure) lets a large volume of magma out in a relatively short period. Edifice failure on its own gives us “just” VEI 6s and VEI 7s. “Just” being a relative term – I would not want to be near one.

  9. Thank you for the conclusion to your epic work! A profoundly interesting series, much enjoyed and appreciated. Thank you, Hector.

  10. Thank you for the enlightening series, Héctor. You have painted a much more complex and fascinating picture than a simple pipe-and-chamber diagram. “Slush” takes on a lot more significant role, amongst other states.

  11. Yes, mjf….that does seem excessive. The following argument is not proof that the Costa et. al. paper is wrong, but it is my take about their numbers……..A key result of that paper is that the estimated mass of the co-ignimbrite ash is greater than the supposed mass of the ignimbrite itself. My impression is that as a fraction of an entire eruption, co-ignimbrite ash tends to make up perhaps 25-50% of the total eruptive mass of most ignimbrite forming eruptions. The eruptions having the largest fraction of co-ignibrite ash seem to include those eruptions having large ratios of outflow sheets to caldera filling material, and eruptions involving crystal poor magma. This rule of thumb does not include eruptions with a large plinian phase. Fine grained vitric rich co-plinian ash can be hard to distinguish from co-ignimbrite ash. If an eruption has more than 50% by mass of fine distal ash, the mass of that ash may be inflated by co-plinian ash. As the younger Toba event does not include a plinian phase, a result where co-ignimbrite ash makes up more than 50% of the total eruptive mass does seem suspect. Hence the calculated mass may be excessive.

  12. Thanks for this great article Hector Sacristan. I did not even know about the Mexican example!! One thing I have always wondered about regarding the lead-up to gigantic silicic eruptions is whether the geophysical precursors look that much more spectacular than those of much smaller eruptions. Is it possible that the precursors to these eruptions may be related to the relatively small volume of magma needed to propagate a dike to the surface, and not to the enormous volume of magma potentially available ? Is it possible the large volume of magma potentially available may have already been in place for a long time and does not need to do anything other than sit there, and feed a dike once one becomes established? Is it even possible to distinguish between the the propagation of a dike that ends up feeding a small volume pre-caldera “leak” to one which triggers a run-away eruption leading to caldera formation ? Even If segregation of a large volume of eruptable magma from a crystal mush can happen very rapidly, would the resulting vertical redistribution of mass in the crust be that obvious at the surface ? Also, what of calderas with “top-down” eruption triggers where a roof block starts to founder ?

    • Those are great questions and it’s also hard to answer them because there is still a lot to be understood. Assembling a large body of melt could be relatively silent, for example if a magma chamber keeps overturning exchanging magma with a hot deeper source then the magma chamber could grow by melting its roof but without any remarkable inflation or earthquakes because there is no pressure increase.

      We don’t know either what makes a volcanic system eventually collapse, at basaltic volcanoes we have seen many collapses that have been well monitored and it usually has to do with the elevation the eruption breaks out from (eruptions at lower elevation needed to make the volcano collapse) or with the portion of the rift that snaps open and takes up magma volume. Collapses at basaltic volcanoes are not usually preceded by any strange long term precursor that distinguishes them from their smaller eruptions, only once the dike intrusion or eruption is already going can someone tell that the situation is extraordinary.

      Okataina seems to have had a large volume, 4-5 km deep, rhyolitic magma chamber for 25,000 years that has remained relatively unchanged through numerous rifting events and eruptions both explosive and effusive. What makes the difference of whether it collapses or not? We can only speculate about that, it could be interaction with water that makes the eruption more violent, it could be that the magma chamber has reached a critical size or melt to crystal ratio in which it becomes more unstable, it could be that a large rifting event takes up enough magma away from the volcano to start a caldera collapse… So I don’t think the precursor would need to be any distinct from the more usual eruptions of a given volcano and that the magma chamber could have been sitting there for a long time before (e.g, Okataina), but there are a lot of details still unknown.

  13. The war has begun, the first strike has been made! My retort awaits and I look forward to overcoming the first phase of this new conflict. Great move Hector but now it’s my turn!

      • If I recall, Tallis is the one who usually writes caldera and supervolcano articles. Now he’s declared “war” on Hector who’s moving in on his territory, meaning most likely Tallis is going to try to write better supervolcano articles and series than this one.

  14. I have been looking at the Okataina volcanoes and trying to map the locations of vents, I noticed something interesting. It looks like there are 5 volcanoes formed after the last caldera, Tarawera and Haroharo are the main ones but there is also Okareka which is southwest of the caldera, and Rotoma and Edgecumbe north of the caldera closer to the ocean. The volcanoes outside the caldera are andesite-dacite comes but Edgecumbe is entirely a holocene volcano, so the area seems to be expanding in that direction…

    I also noticed Tarawera before the Kaharoa eruption had not erupted in the Holocene, then has had 2 eruptions within a millennium and in a row, while Haroharo has had more holocene eruptions but none since about 5000 years ago, 3400 if the Rotokawau basalt craters are a part of it. it looks like the two volcanoes alternate somewhat, so it is probable the next eruption will be from Tarawera again. I remember also reading about an inflating source along the coast to the north, which is probably not part of this system but is still pretty close, and right next to the ocean… Would not be a good place to erupt, basically a VEI 6 maar crater.

    • You forgot drawing the gigantic steam column and tephra above it.. massive Pyrocumulus over large basaltic fissures Tarawera was almost like a Icelandic Dead Zone vent in scale, and certainly much larger than one of Lakis individual vents

      • I actually did, but its very hard to see in the dark, also drawing billowing clouds in Paint is not so easy… 🙂 The red in the sky is meant to be reflected flow, but also the eruption was very sudden and probably the entire fissure on the mountain opened in a few minutes which is possibly most representative of the picture.
        I think most of the actual massive eruption cloud was from the Rotomahana section of the rift though, the vents on Tarawera were lava fountains probably looking like the ones on Izy Oshima, which didnt make extensive lava flows. Tarawera was much hotter though, about 1200 C.

        As a side note, I think there is just a very slight possibility Kilauea is erupting 😀

  15. Remember that there weren’t any humans in New Zealand at the time of Taupo’s 186 CE eruption. The Māori didn’t arrive until the early to mid-14th Century. But they did indeed witness the Kaharoa eruption of 1315 in Tarawera and the Rangitoto in the 14th Century.

Comments are closed.