Rome’s world’s weirdest caldera

Italy is a fascinating country, and when it comes to volcanology, Italy has been arguably the most influential location in the planet. The first ever detailed description of a volcanic eruption came from Pliny the Younger writing about the eruption of Vesuvius in 79 AD. Vesuvius was also the first volcano to be watched over by a volcanological observatory, in 1841. Three types of eruptions plinian, strombolian and vulcanian were described in and named after Italian volcanoes. The very name volcano comes from the island of Vulcano. While In Sicily, according to legend, the philosopher Empedocles jumped into Mount Etna believing he was immortal. And was this before or after Zeus threw Mount Etna on top of the monstrous Typhon?

In such an exuberantly volcanic region it is easy to miss some of the other volcanoes that haven’t done so much historically, but are still there, and some of them are genuine wonders. Here I’m going to talk about Colli Albani. A large caldera right next to Rome. The arguably weirdest caldera in the world in terms of chemical composition.


 The Roman Volcanic Province

Colli Albani is part of a remarkable cluster of large caldera volcanoes. The Roman Magmatic Province. The cluster includes four caldera systems generally classified as separate. These are from north to south: Vulsini, Vico, Monti Sabatini, and Colli Albani. The calderas started their major caldera forming eruptions at about the same time 600,000 years ago. They have become inactive in the past 100,000 years, except for Colli Albani which still shows signs of intrusive activity.

The exact number of caldera forming events produced by these volcanoes seems unclear. The Vulsini Complex had at least 5 ignimbrite caldera-forming events. Vico had four caldera-forming ignimbrite eruptions. The largest eruption of the Roman Magmatic Province may have been the 300,000 BP event that formed the Bolsena Caldera of Vulsini, which had a bulk volume of 460 km3, the caldera is now filled with Lake Bolsena.

Vico caldera seen in Google Earth.


The weirdness of Colli Albani

So what makes Colli Albani so special? A unique chemistry. Colli Albani has done foidite ignimbrites, something as far as I know no other calderas in the world have. Of course, throughout the history of the planet there must have been other calderas similar to Colli Albani, but none of them remain active. Remember foidites? My last article was about foiditic magmas, and then Jesper’s Nyiragongo series also touched this magma type. Foidites are some of the most potassic magmas in the world, very rich in certain elements such as potassium, phosphorus, barium, among many others, and depleted in silicon and aluminium. Foidites include magmas such as melilitites, nephelinites, leucitites, and kimberlites, among many other less common designations. Carbonatites are also related to foidites, given that carbonatites come from nephelinites through a partition between silicate and carbonate melts. Colli Albani’s magmas would be classified as leucitites.

Last time, I was talking about Kaiserstuhl, an extinct nephelinite-carbonatite stratovolcano in Germany, and nearby Urach, a field of melilitite maar volcanoes. However, this is not the only province of foiditic magmas in Europe. One of the most important active foidite volcanoes of the world is in Italy, and it is the only one that has made it into a caldera.

We could distinguish four series of magmas given the way that they are classified in the TAS diagram:

-Low-K series (low potassium): Subalkaline/tholeiite basalts, basaltic-andesites, andesites, dacites, and some rhyolites.

-Medium-K series: Alkalic basalts, trachybasalts, basaltic-trachyandesites, trachyandesites, trachytes, and some rhyolites.

-High-K series: Basanite, tephrite, phonotephrite, tephriphonolite, phonolite, and in some highly evolved cases trachyte.

-Extreme-K series: The rocks clasified as foidites in the TAS diagram.

This is called a TAS diagram. I have coloured the classification according to the Google Earth maps that I show later.

Approximate fields spanned in a TAS diagram by several foidite volcanoes that are discussed in this post, or that were discussed in my previous article. Note Colli Albani is not as potassic as Nyiragongo or Ol Doinyo Lengai but is pretty close to it. Oricola corresponds to monogenetic pyroclastic cones close to Colli Albani.

Foidite volcanoes are usually isolated in terms of their chemical composition. The nephelinitic Ol Doinyo Lengai is the black sheep among the trachytic volcanoes of the Ngorongoro complex. Nyiragongo nephelinites contrast with the basanites and phonolites of the other Virunga volcanoes. Colli Albani is a similar case.

Roman Magmatic Province calderas have compositions that are less potassic than Colli Albani. Vico is a medium-K volcano, that has erupted trachytes and rhyolites. Monti Sabatini and Vulsini are more potassic than Vico, and straddle the boundary between the medium-K and high-K series. To the south of Colli Albani lies the Campanian Volcanic Province where volcanoes have mainly medium-K compositions, with the notable exception of Vesuvius which is a high-k phonolitic-phonotephritic volcano.

Map in Google Earth. Each circle is a chemical sample of lava from the extensive EarthChem catalogue. I coloured them according to the series they belong to, in the same way as shown in the TAS above. Red is low-K. Nearly absent. The only active low-K volcanism in Italy is in the Aeolian Islands. Green is medium-K, and blue is high-K which are the dominant series in this area. Purple is extreme-K lavas, the volcanoes that belong mostly to this series are Colli Albani and some of the monogenetic volcanoes like Oricola, or San Venanzo.

Same map as above but showing the Campanian volcanoes.

There are other foiditic volcanoes in Italy appart from Colli Albani, however they are monogenetic. These monogenetic volcanoes make up the so called Umbria-Latium ultra-alkaline district and are located to the east of the Roman Magmatic Province. They are San Venanzo, Acquasparta, Polino, Cupaello, Oricola, and others. Umbria-Latium volcanoes erupted melilitites and carbonatites. Eruptions were mainly explosive forming maars/diatremes and pyroclastic cones. Some lava flows are also present but usually of very small volume. These eruptions have been usually attributed to hydromagmatism. However, I think they were probably driven, at least partly, by the very high gas contents of melilitites, because wherever there is melilitite you find maars and diatremes.

The Umbria-Latium ultra-alkaline district might still be active, however eruptions here are extremely rare and minuscule in volume. Melilitites are some of the most geochemically interesting magmas but also some of the rarest. It is not unusual to find a lonely melilite maar or tuff cone volcano with no other volcano nearby, which probably formed in one day, was a VEI-2, and effused a 200 metre long lava flow. The way of melilitites is modesty.


Caldera-forming Colli Albani.

Colli Albani is 10-km wide caldera that has collapsed repeatedly generating massive pyroclastic flows. The first dated activity of the volcano was at 608 ka (ka means thousands of years before present). This was also at about the same time explosive activity started in nearby Sabatini volcano, which happened at 591 ka. It has been noted that the eruption histories of Colli Albani and Monti Sabatini appear to be connected, and that they have often entered eruption phases at the same time.

Colli Albani. Image from Luca in Wikipedia. It can be found by clicking here, where you can also learn about cloud types.

The first caldera-forming ignimbrites of Colli Albani are less well known. There are 4-5 ignimbrites, depending on the publication, that took place during 2 eruption periods, at 561 ka and 530 ka. The volume is not well constrained, but they seem to have been at least VEI-5 eruptions. One of the 561 ka explosive eruptions was the Tor de Cenci ignimbrite, that is shown in the map below.

A series of larger caldera forming eruptions followed: the >34 km3 bulk Pozzolane Rosse eruption at 456 ka, the 9 km3 DRE Pozzolane Nere at 407 ka, and the >50 km3 bulk Villa de Senni ignimbrite at 365 ka, which ended the caldera-forming period. As such ignimbrite eruptions occurred at intervals of about 50,000 years, although sometimes there may have been two ignimbrites that occurred in more rapid succession. It is not too easy to know how much time elapsed between certain explosive eruptions, and it is difficult to know which ash deposits formed in  separate eruptions or phases of the same event, making things unclear.

The eruptions, as typical of caldera-forming ignimbrites, were very powerful. A caldera-forming ignimbrite is a particular type of pyroclastic flow of enormous extent which is thought to issue from the ring fault of a caldera through a ring dike. Each Colli Albani ignimbrite started with a plinian/subplianian eruption which rained scoria around the volcano, this phase was relatively small in volume. Each eruption then intensified and a massive pyroclastic flow issued from the caldera, spreading over a radius of 20-30 km all around the volcano, obliterating everything in its way. It would have blown trees to the ground and suffocated animals to death and finally buried them under meters or tens of meters of ash and scoria. Caldera-forming ignimbrites reign supreme among eruptions in terms of intensity. If one of these pyroclastic flows were to occur today, then it would devastate most of Rome. Luckily for Rome, Colli Albani doesn’t seem to be in the caldera business any longer.

The extent of two of the caldera-forming ignimbrites of Colli Albani. The area is as reported in some of the publications cited at the end.

In caldera-forming ignimbrites, the magma type doesn’t seem to matter much. Doesn’t matter if it’s leucitite like Colli Albani, water-rich basalt like Okmok, or rhyolites from the Altiplano-Puna, the eruption intensity is relatively similar. The leucitites of Colli Albani are silica undersaturated. None of the ignimbrites has had more than 50 wt% SiO2. The Pozzolane Rosse eruption, for example, had only 40-44 wt% SiO2. A magma like this is very fluid and gas rich. Normal eruptions will usually be explosive but still have low eruption rates. However, a caldera-collapse is different. Leucitite has enough gas to fragmentate the magma into tephra. The caldera collapse brings in enormous eruption rates that generate the violent pyroclastic flow-style of activity. The only magmas that seem incapable of producing caldera-forming ignimbrites are the gas-poor basalts from intraplate or mid-ocean ridge environments, probably because of the very little volatile content that they carry.

Following the final ignimbrite, the Villa Senni eruption, Colli Albani entered a very active period from ~365 ka to ~350 ka in which many vents erupted from a ring-like area around the caldera, making scoria cones. These eruptions were effusive and explosive. They produced lava flows, but also generated subplinian eruptions that deposited scoria for tens of kilometres to the east of the volcano. Nearby Sabatini volcano also has a similar pattern of concentric fissures which envelop the Lago di Bracciano caldera.

Scoria cones of Colli Albani marked in red circles. The approximate caldera rim is in white.

The three calderas of Monti Sabatini are marked in white. From the left these are Bracciano, Baccano and Sacrofano. Scoria cones are in red, maars are blue. Note the ring-like distribution around Bracciano and Sacrofano calderas.

The distribution of fissures around the caldera suggests to me that they were fed from cone sheet intrusions. Cone sheets are intrusions which face inwards towards the centre of a volcano and, together, multiple cone sheet intrusions make the shape of a cup or an inverted cone, and where they reach the surface, concentric fissures open up that usually run parallel to the rim of the caldera. Such intrusions are typical of caldera systems worldwide. There is a link with caldera resurgence. In order for magma to intrude, it must be able to push away the rock. Cone sheets happen by pushing the rock upwards. Because they face the caldera, they must push the caldera floor up, and if the caldera is already rising up because of an expanding magma chamber below, then the cone sheets will intrude effortlessly towards the surface, taking up the stresses generated by this resurgence.

Drawing of a resurgent caldera which is feeding cone sheet intrusions (red).

Thick successions of scoria fall deposits and lava flows separate the major ignimbrites of Colli Albani. Many scientific publications focus only on the ignimbrites, so that it may seem that the 561-365 ka period of Colli Albani activity simply involved a couple of ignimbrites and there was little intervening activity. In reality, I suspect that very roughly half of the volume erupted from 561 ka to 365 ka, was possibly erupted from circum-caldera fissures fed by cone sheets, similar in style to the activity that followed the Villa Senni eruption. This activity probably would have accompanied caldera resurgence that eventually culminated in the ignimbrite eruptions.

The “reverse” of a cone sheet is a ring dike. While cone sheets thrive in emerging calderas, ring dikes intrude in collapsing calderas. Ring dikes work by pushing the ground downwards. They need a trigger to depressurize the caldera. Most caldera-forming ignimbrite eruptions of Colli Albani started with a scoria fall deposit, which corresponds to a subplinian/plinian eruption. These eruptions would have depressed the floor of the caldera by extracting a large amount of magma from it. The subsidence allows a ring dike to intrude, in turn the dike lubricates the caldera ring fault with magma, and the roof collapses faster, magma shoots out of the ring dike in a caldera-forming ignimbrite eruption, the more the roof collapses the more the dike opens up and the faster it erupts.

Representation of a caldera-forming ignimbrite eruption fed from a ring dike.


Post-caldera Colli Albani

Following the ignimbrite eruptions, Colli Albani became less active, even so it has kept some activity to recent times. After the ignimbrites, activity became centred inside the caldera were it built a small stratocone, the Faete stratovolcano. Faete is known to have been active in 290-260 ka, although may have started earlier. The volcano was mainly constructed from lava flows and subplinian eruptions, but also involved maar style activity that blew away the top of the cone into a 2 kilometre wide crater.

Since 200 ka, Colli Albani has been dominated by maar eruptions. Several maars have been constructed, including three large polygenetic maars southwest of Faete stratovolcano. The largest maar is Albano, which erupted 7 times starting 45 ka. Albano makes up a massive crater of 3 x 4 km, and 200-400 meters deep in relation to the rim of the crater. The explosions of Albano ejected large metre sized blocks of old lava and carbonate rocks and deadly pyroclastic surges swept across the landscape to distances of 15 kilometres to the northwest and southeast of the vent.

The last dated event of Lake Albano in 4.8 ka may have been non-volcanic, and consisting of a lahar. This seemingly repeated in 406 BC, when, as reported by Plutarch, the lake surged over the surrounding hills, unleashing a flood of water that destroyed fields and vineyards, and then poured into the sea. This event may have been related to hydrothermal activity.

The young maar volcanoes of Colli Albani shown in blue. Green corresponds to the ejecta from the maars, mainly pyroclastic surges.

Lake Albano in the Albano maar. The bottom of the lake is 170 meters underwater. Image from Ra Boe is found here.


Will Colli Albani erupt again?

It seems to me that Colli Albani is a waning volcano well past its prime. However, it will probably erupt again.

The Global Volcanism Program reports that Colli Albani experienced inflation between 1993 and 2000, which might still be continuing. The inflation rate in 1993-2000 was about 2-4 mm per year. This is, of course, very little compared to other volcanoes of the world. The inflation is, however, in an ominous location. Inflation is centred in the Albano and Remi maars. Likely there is a magma storage under the large polygenetic maars which is refilling with magma. There is also active seismic activity. 1100 earthquakes took place under the Albano maar in 1989-1990 at depths of 4-5 km.

Should a large maar eruption take place, consequences would be devastating. There are well over 100,000 people living in the area that was affected by earlier pyroclastic surges of the maar eruptions. Two of the larger towns in this area amount to 80,000 people. Many other towns amount together to a probably greater number. So even though an eruption is unlikely to happen anytime soon, Colli Albani should be carefully watched.

One publication has also suggested that Lake Albano could produce a limnic eruption given the CO2 that is being dissolved into the lake. This is, however, considered by them to be an unlikely possibility, given that the concentration of CO2 is presently far from saturation.



Colli Albani has it all. It has bizarre leucitite magmas, the power of a large caldera, and a bit of a mystery regarding Lake Albano, with its earthquakes, inflations, and floods. It is also a volcano that can teach us about calderas.

Calderas are calderas, regardless of their magma composition. It is true that rhyolite calderas usually erupt larger volumes. However, leucitite, trachyandesite, or basalt calderas, all of them are frightening systems capable of VEI-6+ eruptions. The ignimbrites of Colli Albani would each have destroyed areas of 1000-2000 km2 within a very short time-span.

In reality, volcanic eruptions are very complex. A lot of factors. There is viscosity, yes, but there is also volatile content, magma chamber size, conduit width, or groundwater interactions. Simple rules like: the more a volcano pressurizes, the worst the eruption will be, or the more viscous the magma, the more explosive, fall short to explain the enormous diversity of volcanic phenomena out there. And in a world full of volcanoes you can still find a volcano that is a one of a kind gem, like Colli Albani.





Late Pleistocene ultra-alkaline magmatic activity in the Umbria-Latium region (Italy): An overview


Integrated approach for the reconstruction of stratigraphy and geology of Quaternary volcanic terrains: An application to the Vulsini Volcanoes (central Italy)


Origins and energetics of maar volcanoes: Examples from the ultrapotassic Sabatini Volcanic District (Roman Province, Central Italy)

Colli Albani:

The Colli Albani mafic caldera (Roma, Italy): Stratigraphy, structure and petrology.

A large K-foiditic hydromagmatic eruption from the early activity of the Alban Hills Volcanic District, Italy

CO2-driven large mafic explosive eruptions: the Pozzolane Rosse case study from the Colli Albani Volcanic District (Italy)

The 1989–1990 seismic swarm in the Alban Hills volcanic area, central Italy

Discussion on ignimbrites:

Classification of ignimbrites and their eruptions

301 thoughts on “Rome’s world’s weirdest caldera

  1. Looking at the cone now, it does seem to have gone down the typical path of small to moderate basaltic eruptions, unlike the last eruption which behaved more like a big eruption but on a small scale. First a fissure with a moderate output opens (10-100 m3/s) that forms into a cone, and then that cone basically walls itself in and becomes a bubbling lava lake with some strong fountaining and strombolian activity, with lava flows on the surface mostly at some distance away from the cone itself fed by tubes under the cone.
    I dont know what exactly is a good name for this sort of activity, but it is different from what many larger eruptions are. Large eruptions with a low stable effusion rate make shields or perhaps small stratovolcanoes and often have episodic activity with brief high effusion that trends into continuous slow effusion (Pu’u O’o is the best example), I think last years eruption almost did this but was not able to form a stable open conduit within the upper crust. Other large eruptions simply begin as lava floods and then become a powerful single vent, that then pretty abruptly stops when the pressure is gone, and most are related to caldera collapse or subsidence. This sort also seems to be 2 subtypes, one where a rift opens above a magma chamber, and another where a rift opens next t oa magma chamber. The latter will see a dike form very fast and probably do eruptiosn of moderate volume but very high intensity, like Krafla. The other will see a huge volume of magma drain, and if it erupts will see a long lived eruption at high effusion rate, like Holuhraun or Leilani.

    What we have here really isnt any of these styles, it is a higher base effusion rate than what a shield would have long term, but also nowhere near what a proper rifting fissure would do if there was a magma chamber involved, and it is well known there definitely isnt a crustal chamber involved unless one has formed only in the last year. It really is exactly like what I proposed in my speculative article last year about the Hell Machine, except that such a machine is evidently not a runaway process like I proposed, which is probably a very good thing 🙂
    Not all of these types of eruptions are small either, both Tolbachik in 2012 and La Palma last year would fit into this, and they are the 3rd and 4th biggest basaltic eruptions of the 21st century. Lanzarote would fit this style too, and no-one can argue that was small in any situation. The eruption of Hallmundahraun up at Langjokull I always thought was a shield, but I actually found something of a very detailed and well researched video about a first hand account that shows it was a more powerful eruption that had strong fountaining and some faster lava flows, but still nowhere near the scale or intensity of Eldgja or Laki, so maybe also fitting into this category of mild intensity eruptions.

    I guess maybe this style could be called Fournaisian, given Reunion seems to nearly exclusively erupt in this way outside of caldera collapse. It might also be called ‘effusive strombolian’ but that really doesn’t give the most accurate image. Most books would call it hawaiian but that I think is better descriptive of proper rifting lava floods, especially as the word was made to describe the intense eruptions from Mauna Loa.

    All this really shows, is that all of our categorizations are way too simple to describe the real world 🙂

    • Yes very similar to Piton De La Fournasies small cones, and the viscosity seems similar too many of Pitons more strombolian Looking eruptions

      The current meradalir lava is much more viscous than the 2021 s Geldingadalir lava. Not as viscous as pure ”strombolian” But more viscous than Hawaiian.. perhaps ”viscous hawaiian style” Althrough it is a bit more fluid than Fimmvörðuháls that was pure Etnean in viscosity

      Hawaiian style is any eruption with very very fluid lava 🙂 small or large
      Thats what the criteria in geology books are. Althrough Mauna Loa is real ”mini flood basalts” very much

      • There actually was shiny pahoehoe erupting the other day, when the cone breached. The lava also looked just like it does now back in the early days of ladt years eruption, it was really only when a proper conduit began to form and particularly when a lava lake existed that the lava was erupted as pahoehoe. It has only been 2 weeks, give it time 🙂

        I think in terms ofthe potential devastation it could cause, a major flank eruption from Etna is probably a lot scarier than an eruption at Mauna Loa. But eruptions like 1669 are very rare, the last one that was similar was in 1030, and that was rather smaller in volume. Mauna Loa probably cant be beaten in terms of the combined size and intensity of its eruptions though, 1950 opened along the entire 20 km within 2 hours of the eruption starting, before any of the flows had even reached the sea. Actually, the effusion rate in that time makes even the infamous numbers for Nyiragongo look quite tame, that first 3 hours might have erupted most of the entire volume, which in total was 0.36 km3, that is going on a direct 1/1 with a VEI 5… 33,000 m3/s. That might well be as close as we can get to visualising a flood basalt, even if it doesnt last long. Most eruptions there are far less intense, but even at that, many probably go over 1000 m3/s to begin with.

      • Yes 1950 was an insane event .. how tall was the fountains ?

        • Hard to say, but reportedly over 100 meters the entire length at least where numbers are given at all… But actual observations were made only later when the eruption had gone down to ‘normal’ scale so there is no real data on that early stage except for distant observation from HVO.

          • And the $64K question, still unanswered by the whole of the volcanological discipline, is why exactly did ML erupt that much? It was only eight years since the last RZ eruption, and 24 since the last SWRZ one. Where was all that magma hiding? The EQs didn’t really reveal a huge slug of magma arriving at ML suddenly, I don’t think any observer thought it’d be any different than 1926, 1919 or 1907.

          • I think it was a failed caldera collapse, one where I guess the dike broke out too early to reach a low elevation. But the fact that even the 1949 cone that still had lava in it didnt collapse is a bit strange. I guess maybe the entire magma system must have been ruptured t ofeed the dike, not just the upper levels, so that deflation was never concentrated. This might also be why it has been so inactive since, only really being able to inflate continuously since 2002, and 1974-1985.

          • I have a different idea regarding 1950. If one looks at the historical SWRZ eruptions, which were 1868, 1887, 1907, 1916, 1919, 1926 and 1950, a regular pattern shows up. Each eruption generally happens uprift from the previous one. In fact 1868 and 1887 were very exceptional eruptions that were the lowest elevation SWRZ eruptions in over 2000 years. Given their huge length the dike intrusions feeding the 1887 and 1868 eruptions must have very roughly had volumes of 300 million cubic meters or so, almost as large as the 1950 eruption itself.

            The decrease in dike lengths I believe shows that the stress in the rift was changing. There would have been more tensional stress in 1868 allowing dikes to intrude more effortlessly. Space would have been gradually filled in by magma. Each event would have taken higher pressure to happen The pressure in 1950 would have been reaching a very high point, as shown also by the large long-lived summit eruptions that happened in 1949 and 1940. So the 1950 dike was the first to have high enough pressure to overflow along its entire length. 1950 would have been more forceful than earlier SWRZ eruptions which were more passive rift-filling events.

            That said there is also some overhype regarding the 1950 eruption. The largest eruption by volume was the 1859 eruption, which also had a very intense start with a curtain of fire nearly 6 kilometres long.

            The 1868 eruption stands out for its very high sustained effusion rates, lava was running into the ocean at such high rates that a sizable tuff cone grew were lava was exploding against the water, and a stationary thunderstorm developed above. If taken into account dike volume + eruption volume, then I suspect that the 1868 event may have been the largest in the historical period of volume. The eruption volume is not well known because most went into the ocean.

            Most other historical and prehistorical fissure eruptions have had intense fire curtains with kilometric lengths and initial eruption rates possibly in excess 1000 m3/s, including summit, radial, NERZ, and SWRZ eruptions.

    • Even the avarge start of a Mauna Loa eruptions is stuff of insanity .. 1950 Did 5 Fagradalshraun volumes in just over – night ! .. Infact Mauna Loa is probaly the most scary basaltic volcano at current

      • It’s nothing to sneeze at, that’s for sure. A lot depends on which RZ is the winner this time.
        NE tends toward shorter fissures, more gentle slopes and a longer distance to developed areas.
        SW unzipped from the summit to 6000 ft within hours and immediately headed toward 30 deg. slopes. Plus, there are residential developments within rock-throwing distance of the RZ.
        The radial vents are the wild card, and that’s no great deduction on my part.

        If the SW or a radial erupt, my azz will be on the next plane to KOA. If it’s the NE, I’ll stay home and watch livecams.

        • You live in south Kona? I was not able to see any of the Kona side when I visited, but it looks really nice and very different to Puna where I was.

          • Oh, I see so you are going TO Hawaii to see it, not trying to escape 🙂

            I think there is a good chance of a SWRZ eruption, the whole idea of the rifts alternating is not really very reliable but with the last two eruptions being NERZ events it would seem the SWRZ is maybe due for something.

            Radial, well there is no real data how they start. 1859 was a deep eccentric eruption, with a very high lava temperature (1220 C) and powerful fountaining. 1877 was by contrast a shallow lava lake draining, like what happened in 1823 on Kilauea, or what Nyiragongo is infamous for. The radial vent in 1852 might have also been an eccentric eruption but happened only right outside of the caldera, for some All of the other radial vents were associated with a larger eruption on the NERZ so maybe are better considered something different.

            I think that the eruptions from Kilauea Iki are also eccentric eruptions, with a different path than Halemaumau, although not a separate volcano. Both of them largely bypass the main magma system so erupt much hotter lava.

        • RE: Plus, there are residential developments within rock-throwing distance of the RZ.

          They certainly do like to roll the dice in Kona. I recall my stay a the Four Season, $1million dollar homes were for sale, on the old lava field, right under the shadow of Hualalai

  2. Great article about the newly discovered ~8km Nadir impact crater off the coast of West Africa near the Nadir Seamount.

    The most interesting part of this is that it is possibly concurrent with Chicxulub and may have also occurred at or very near the K-Pg boundary, though it was a much smaller ~400m impactor. Still would do quite a bit of damage today.

  3. Apologies for jumping around topically here, but I have another conceptual question floating around my brain this morning.

    In the Chiles Cerro Negro article I brought up Tambora and Hector / Chad and others shared some good insights.

    Tying some things together, I’ve learned how large caldera collapse eruptions can occur in stages, sometimes separated by several human lifetimes or more but geologically very close together (as if being analyzed by a far future observer).

    Tambora’s massive VEI 7 occurred without razing the entire edifice; only the top several km’s collapsed / evacuated. Is it possible this is sort of an awkward “mid point” in Tambora’s caldera phase, and sometime in the future another very large event may occur that “completes” the process and forms a more typical large, “flat” caldera? More like the stereotypical Crater Lake / Mt Mazama, or analogous to other large Indonesian calderas.

    Despite how enormous Tambora 1815 was, it kind of appears to have only have made it half way to the caldera phase. Of course it could be special unique characteristics of this volcano, or some other explanation.

    And it seems this might apply to Rinjani as well? Is this plausible?

    • I’m thinking that’s extremely unlikely, Large Stratovolcanoes such as Tambora usually go through cycles of destroying and rebuilding themselves. Samalas and Tambora has produced large VEI6+ eruptions before their VEI 7s. These aren’t the classic caldera volcanoes like Toba and Yellowstone.

    • I think maybe it is more of the shape of the magma chamber, in those volcanoes maybe the chamber is smaller than the edifice. Rinjani is actively building a new stratovolcano so I doubt it has any dangerous large storage now, the actual mountain of Rinjani was I guess just not exactly centered on the magma chamber, it might have been a somma structure of a repeat offender but repeats in these cases are not close together. Better to look at the not-collapsed volcanoes nearby, like Agung or Sangeang Api. I used to think neither was close, as they only erupt basaltic andesite or andesite, not rhyolite, but then Tambora was a variant of andesite so I guess that idea doesnt hold… Rinjani was a dacite at first but became andesite and basalt later and that was most of the eruption, as well as what it erupts today. So these volcanoes really do just get too big and force collapse themselves instead of evolving.

      • Good stuff, thank you Chad! Rinjani as a somma system is definitely interesting.

        And I think when I first began learning about this stuff, I got the impression that large scale explosive volcanism naturally should require dacite or rhyolite. Then you see learn about some large volcanoes erupting andesite or basaltic andesite and you start to realize there’s an enormous spectrum of forces playing out at each individual volcanic system.

        • Calderas are a whole different class of volcanism, which we have only really gotten a small peak at. Kilauea in 2018 was a small caldera, and stayed quiet, but large calderas even at Kilauea seem to be very different. I have not seen a lot of data, but Kilauea might have produced ignimbrites in the Pleistocene, when the Pahala ash was being made. It might be getting close again, given the extent that pyroclastic surges have gone, although these are not ignimbrites. Kilauea also looks like it did some proper flood lavas from its summit at that time too, it must have had a much bigger magma chamber than it does today. I dont know if Hector has seen some of this stuff but might be able to give more information 🙂

          The current article really shows magma characteristics has not got much to do with its ability to do an ignimbrite. Colli Albani has magma that is probably just as fluid as Kilauea. In fact so does Vesuvius, where the very term ‘plinian’ originated. The actual type case of a plinian eruption was from a mostly mafic eruption and fluid magma…

          • Yup, well I have to be wrong for a while before I can start to be right, you know?

            I’ve been getting more involved in asking you guys this stuff so I can better sort this information in my mind and reject the things that aren’t well supported.

            From a lot of articles this year I really did learn how little magma chemistry ultimately matters in the explosivity of a volcanic system. It’s not that it doesn’t matter it all, more just how the characteristics of the volcano matter more.

            And that’s wild re: Kīlauea. I’ve heard about it doing ignimbrite s in the Pleistocene, but I’ve never looked into it much. Is it theoretically possible for it to loop around to that kind of activity again, or was that characteristic of the morphology of a past version of Kīlauea?

          • Vesuvius 79AD and 1631 was mostly phonolites, yes?

            I thought it switched to more fluid and less evolved magma after its phonolitic reservoirs were drained, but that was just my working understanding.

            Pretty sure I remember reading phonolites can be fairly fluid as well at higher temps, as it’s not incredibly silicic.

  4. At volcanodiscovery there is a new map of the lava flows until August 15th. Apparently the lava has moved under the old ceiling. I have never seen lava on the eastern edge on the mbl ( panorama fromKeilir) cam.

    • It didn’t move under the old ceiling. The lava got there shortly after the eruption started (within the first week) but there were several days with no visibility due to bad weather during that time. It never got that far again during the last ~10 days.

    • The eastern edge moved toward the east during several days when nothing was visible due to low clouds. There were comments from observers that it was close to overflowing the eastern gap, but I didn’t see any pictures of the lava.

  5. Well, seems like I came back from my vacation (sailing) right for the moment when the current vent is shutting down.

    On the GPS systems it looks like the pressure is ramping up fast now, but I am reticent to say to much since there are to few datapoints to be sure.
    But, if I am right new vents should pop up in the not to distant future. I expect them to form a little bit further towards Keilir. Ie, same vent progression as the last eruption.

    • And even faster eruptions?
      Any curtains of fire next time?
      Well the startup of this eruption had a small curtain of fire, While Ögmundarhraun had fissures km long and perhaps only lasted two weeks. Kapelluhraun the same, If Krysuvik blows the two examples .. then Hafnarfjördur coud be in serious trouble .. While Brennsteinsfjöll haves the capacity of threating Kopavogur. But to reach settled areras requires eruptions in the northen parts of these volcanic systems, either eruption rates haves to be gigantic ( fast moving Aa ) or you needs tube feed pahoehoe .. slow and steady and insulated eruption

      But it does look like that the eruptions are moving north, Krysuvik haves a tremedous ammounts of pahoehoe so not all eruptions are fast, some are slow and may last years, even a decade

    • I say, go 500m straight north from the northernmost point of the first fissure, then draw a circle with a 500m radius centered around that point. My guess is that at least part of the next fissure will be inside that circle.

      • Sounds likely yes

        Still quite small the eruptions been in speed

    • Oh am I glad you are here Carl, love the insight to what is (possibly) going to happen in-light of the decrease in lava output.

      • I’m really looking forward to what he will uncover after this boat trip! remember last time Carl went sailing he exposed Etna being a fake tourist attraction!

        What’s it gonna be? That Krakatoa was just stolen Chinese firecrackers?

        • Crateri Silvestri didn’t appear fake to me last June. As for the summit, volcano tourists weren’t going there anytime soon.

    • Thanks for the heads up Carl. I can remember last year when I thought the eruption was ending, as vent #1 and #2 languished, but all of a sudden new vents appeared in the north (around noon time, local time, if my memory is correct) so I ended up with egg all over my face. Thanks, again.

  6. A commenter on one of the YT streams pointed this out to me, and I have to agree. If you pull up the visir stream and then scroll back to about 5 hours ago (from the time of this posting) and then quickly switch back to the live view, you can see a clear rise in the overall height of the cone relative to the distant hills in the background that doesn’t appear to have come from spattering, but rather the whole thing moving upward en masse. It’s a subtle effect and initially I was skeptical, but I couldn’t find any evidence that the camera angle has changed during this time, and it does not appear to be an optical illusion due to changing lighting conditions or anything. I then started stepping back and forth in ~30 min increments, and you can see that the change is indeed incremental and not due to a sudden shift (although it appears to have done the most swelling over the past 2 or 3 hours). I’d post some screenshots, but I don’t want to fall afoul of copyright issues.

    • So you can! Its visible on the Langhóll camera too a bit, around where that baby volcano used to be.

    • Endogenous growth, this is common in shields and especially in places where an eruption is trapped by topography, HVO has a great recent timelapse of Kilauea doing the same thing since last year. It really didnt happen at Fagradalsfjall last year though because that was sitting on top of a hill.

  7. But on the visir cam it is only the main cone. The rest of the cone on the right does not lift. At least I can’t see it.

    • Look closely at the part on the right. Over time it lifts and sinks and very slowly moves to the right, away from the rest of the cone. It’s one big and very slow lava boat!

  8. Submarine flows flow freely underwater:
    Lava is a great insulator, and that allows it to flow underwater, without no problem.
    Lava haves so very low heat conductivity I guess the cooling rate for underwater lava flows does not much differ from land, the water can only chill the surface of active submarine lava flows, and under the crust it remains hot.
    Submarine have glassy surfaces, but crystaline interiors showing They cool slowly underwater as well.

    With low heat conductivity and high heat capacity, basaltic lava can move quite freely underwater without cooling, as these amazing Photos shows. The drained pillow lava Shell with liquid interior, and the faster sheet flows .. that formed lava vortexes in the flexible crust underwater, are good signs of low conductivity and low viscosity. The lava stalagtites and drips on the underside of these crusts, also show how good at insulating the crust is from the cold seawater.

    Lava can flow a long time underwater, as long as the supply is feeding it.

    Knowing How low heat conductivity and high heat capacity lava haves, you cannot chill a lava flow with water very much, the interior of the lava flow will remain hot and push foreward.
    Heimeay is often called ”’humans stopped lava flow with water” But most of the Heimeay lava flow flowed Not towards the town either, so they only chilled a wall. Had the lava chosed to flow right through the town, then the water woud have not worked at all I guess, as they cannot chill the flows interior

    • Good Photos of lava vortexes underwater, been flowing turbulence as well underwater, I guess the seawater did not cool it very much when it was flowing

      Nope you cannot stop a lava flow with spraying water on it: the heat conductivity is too low and high capacity is too high

  9. is a live drone video now (4:01:08 PM PDT) of the Meradalir eruption on Thursday August 18, 2022. The eruption is languishing now, all we have is a lava pool with occasionaly gas bursts stirring up the lake, but no overflow looks apparent.

    • To me, what is most interesting about this video is the obvious surface lava flow from the SW lobe of the lava lake to the NE lobe. Is this just the surface expression of an internal loop flow, or is the lava in the NE lobe draining out below to feed the lava tubes that must be present below the surface.

  10. Have there been any infrared topographical photos released publicly of the current eruption? I would think that someone has a good flir camera and some drone has been busy mapping the thermally hot areas? This data seemed hard to get during the previous eruption.

  11. Not much activity left, now.
    The next few hours will be interesting, to see how long it takes before something will give.

  12. There are now some more small quakes happening again, not located yet but thingsmight be breaking. I dont think we will see lots of quakes before a new fissure though, it has only been 2 weeks, so most of the dike should be still molten except maybe the top few tens of meters. But it might rather make a new fissure at another location than the existing fissure reopening again.

    Does anyone have the links to the deformation data? Carl says it is now inflating and that would make sense but how much could be telling to if we get a new fissure in the next day or if it might take some weeks or even months again.

    • Well it looks spent today,but I take Carl’s and Chad’s information as a give.Its not a carbon copy of the 2021 eruption.

    • Of course it’s possible but not necessarily likely that the eruption is petering out because it’s ending, yes?

      All of the things stated about the deep, continual supply of magma under this system suggest a relatively large supply, so the events of the past few days could be taken as an “access problem” in a sense?

      In short, we’ll probably ultimately get another vent and a renewed, vigorous eruption is where I was going with that. The activity before Fagradalsfjall II seemed quite intense for this to be “it.”

      • It could end, technically it has to eventually… but it is unlikely that it will stay quiet too long really. USGS defines an eruption as over if 3 months go by without it starting again. If it erupts again even in the same place then that is a separate eruption, likewise if it stops and another fissure opens in Natthagi, or near Keilir, way away from the existing vent, and it erupts in 1 month then that is still part of the eruption that began on August 3rd. But it is all fed by the same magma, so it is all sort of the same eruption.

        Fagradalsfjall has got a magma supply rate of about 15 m3/s. That isnt a lot compared to a typical eruption, but for an actual deep supply rate that is huge, it is equivalent to 0.45 km3 a year. Kilauea, which has the highest long term supply rate of any single volcano, is only about 3-5 m3/s. The whole of Iceland long term in a typical is probably only slightly higher than that. So Fagradalsfjall right now is basically like all of the rest of Iceland funneled into a single spot… Given that these rifting episodes on Reykjanes often last decades per volcano I doubt we will see more than a year go by without an eruption in some form happening at Fagradalsfjall until at least 2040. By then everyone who lands in Iceland will get to see a new skyline driving to Reykjavik 🙂

        Something about these rifting events too, the first stage is often not as volcanic because all of the magma goes into the rift to fill it up. It is only when the rift is full that eruptions happen a lot more, so eruptions will only get more frequent and probably also more powerful as this goes on.

      • There is a large supply right now into the area, as shown by the series of 2 sill intrusions, 3 dike intrusions, and 2 eruptions that have affected this area since 2020. However note that the December 2021 dike died out quite rapidly. Which means the supply to the December 2021 dike was interrupted, if not then the intrusion would have kept slowly growing until erupting. Perhaps the pressure was insufficient to sustain flow into the December 2021 dike? This is why I’m not sure if the supply will hold up this time or not.

        • It may be similar to what happened with the December 2020 eruption of Kilauea that died out much sooner than I expected because pressure did not keep with the rising elevation of the caldera floor, and magma, instead of erupting, went into the inflating the summit magma chamber.

          • Kilauea only really stopped for a few months, barely longer than the 3 month limit. And when it did start it was very strong and also in part from the same place as the earlier eruption. It basically was a resumption of the same eruption just that the pause was longer than our arbitrary definition…

            Would expect that the supply rate has not really changed at Fagradalsfjall, and that you are probably onto something with all the Reykjanes volcanoes being connected below 7 km. In that case since February 2021 there have been 4 intrusions at about 3-4 month intervals after last years eruption ended.

          • I tend to think of the Reykjanes volcanoes, at least going from Eldey to Hengill (maybe a bit further north, to Hromundartindur and Grimsnes) are like a single volcano with a weird oblique structure. South of Eldey it becomes a proper ocean ridge, with spreading at a right angle to the volcanism, and north of Hengill it becomes a graben that has similar characteristics to a MIR again but on land, before going to large volcanism at Langjokull again.

            It is a bit hard to say, maybe calling it a single volcano isnt the right idea, but they do seem to interact. I guess that the magma last year shot through and erupted fresh, but might end up more compositionally similar to the magma from the middle ages with time. It would be nice to find the exact location of the first eruption in the last cycle, and if it has got a more primitive composition than the later eruptions. Or of eruptions at Brennisteinsfjoll, which seem to be the same style as those at Fagradalsfjall now, if those are more primitive than the eruptions at Krysuvik which should involve some storage in the crust.

      • Strange, the dike in December was smaller than this one but still shows as deflation at the FAFC station, where this intrusion is hard inflation that is still ongoing. It is like the eruption was always too small… I am much more ocnvinced we wil lsee somethign break out soon now, either the existign vent has a paroxysm or we see another fissure open somewhere, or just the whole original fissure line opens again. I expect we will see lava escape Meradalir too, it is already high enough to spill over just has lost its feed, a new flow will escape immediately.

        • Something interesting is that Thorbjorn clearly deflated with the dike intrusion as seen in the SKSH GPS seen here:

          The dike would be expected to have pushed SKSH up, the fact that it moved down likely means there is a local signal from the sill complex below being drained. The deformation is fast and significant which I think means there is a good connection between the sill complex under Thorbjorn and Fagradalsfall. It could explain why the dike grew so rapidly at first and why eruption rates were high when the eruption started.

          • Not sure about the dike is influencing the Thorbjörn segment in a comon rift situation. In Reykjanes it is an Oblique Rift we are dealing with. When rift is the cause of the intrusion (more or less sucking up magma, underpressure by moving plates), some parts of the crust may go down in this area.

            In casus where no magma is involved, tension, stress is releaved by larger magnitude eartquakes. Result is then too that some segment move up, others may go down. I think it has to do with rift at first, magma intrusion following.
            Note that Nylenda gps did not stop moving. Rift is the driving proces.

  13. One thing I will note is that the Lava thickness is 25-30 meters surrounding the crater and 35-40 meters for the top of crater rims. Perhaps this is impacting the lava output too?? I know the 2021 crater was over 100 meters tall, but perhaps circumstances throughout the eruption allowed lava to flow up it more easy. This one we are still early on and it’s nearly half the height above the surface compared to the other.

  14. If this decrease in activity continues, I think this fissure has a week left at most. Even in the past day, the spatter has decreased in height. Lava may be reaching 10-20 meters. I can’t see much from here except the end or new fissures opening up.

  15. If this decrease in activity continues, I think this fissure has a week left at most. Even in the past day, the spatter has decreased in height. Lava may be reaching 10-20 meters. I can’t see much from here except the end or new fissures opening up. This eruption is becoming quiet for now.

  16. Show’s not over yet! Must be going into a pulsing phase.

    • Yes these are odd breakouts. The problem is the tremor isn’t going up. I cannot explain what has been going on recently but most evidentiary would be continual spattering, lots of smoke and rise in tremor for the more short term indicators of a new vent. Something to keep a eye on, perhaps this really is a indication that something is going on.

    • After watching awhile, a 3rd spot seems to be wanting to set up, to the right of the two vents now. 1 am local time. Thick smoke obscures it most of the time.

      • Yeah. The next few days will be extremely interesting. Perhaps this is just a tease. It is good to note that the same output decrease led up to new vents in 2021. Today is now day 17 of this eruption and new vents opened up on day 17 of the 2021 eruption. As time goes by here, it is making me more anxious with the uncertainty of a possible end or if new vents will start up. I hope this eruption doesn’t end early, that would kind of be unfortunate, or perhaps we were spoiled too much with the extraordinary 2021 eruption.

  17. The tremor indicates this fissure will end soon. Around the 5700-5800 level for the blue plot. That is around the on-threshold for the 2021 on and off phase.

    • Breakouts and flows I believe, more or less along the flow paths from a week ago, and a quite overexposed webcam making it look rather dramatic.

      Dawn is coming now, so it should be possible to see more normally exposed views of it, soon.

      • Thanks Erik,yes I’ve come to that view now-breakouts….the smoke enlarged the outbreak to a point is it a new vent.Bring on day light I say.

    • Probably related to the 3.2M quake that occurred nearby, we’ll need more data points to know more though.

    • The spike could be the M3.2 quake at Reykjanestá. The low frequency curve seems to level out the last hours, so let’s see what happens.

      • Or maybe the wind is masking the decline in the curve… Just a thought.

        • I found a graphic on someone’s Twitter feed that gives a nice comparison to last year:

          • It is here good to notice that the tremor level was higher during this eruption compared to the previous one.

  18. Well no more spattering that I can Visibly see. A pause for now in the very least.

  19. The GPS-systems is indicating a slight rise in systemic pressure.
    So, even though the current fissure has mostly played out, there is the potential for a new fissure to open up in the direction of Keilir.

    But, for now and for this vent the show is over.

    • Ok thanks! A short and sweet eruption….not the last ,waiting for Fad III.

    • Carl, are you aware of any maps which show the typical behavior of fissures starting up away from the rift zone, but growing closer to the source, as new fissures erupt, during the last cycle of the Reykjanes fissure eruptions? I am curious to see if past flows in this region of Iceland show the same behavior, something we might be able to use in our current situation.

  20. On the topic of this article, the volcanoes around Rome, I found this piece of information on GVP about Monti Sabatini.

    “During the third phase, hydromagmatic eruptions formed the Sacrofano Caldera and minor lava flows occurred in the post-caldera stage. In the western sector, a stratovolcano and the Bracciano Caldera developed, along with fissure eruptions that produced flood lavas greater than 100 km3, air-fall tuffs, and surge deposits.”

    Apparently this thing decided to challenge both Iceland and Hawaii to a biggest lava flow competition and won by a long shot… This really needs to be investigated, even if Sabatini is inactive it might say something about the capabilities of Vesuvius and Etna that we have never considered.

    • They might be speaking of 100 km3 as the cumulative volume of all lava flows and not an individual event.

      • That would be most likely, but the way the describe it really needs better wording then. It is also still notable in any case, even Etna might not have that volume of lava flows erupted on a large scale in its history, and it is very different from the other volcanoes in the viscinity of Sabatini that seem to produce lava flows only in smaller activity, building their cones, not in large events like this. Maybe the caldera of Sabatini is not an explosive caldera, or at least not only formed that way.

        • I think the Roman Magmatic Province calderas, in general, had abundant lava flows. Colli Albani has many lava flows. Some of them are 15 km long, which is not bad. Lava flows are typical of circumferential fissure eruptions in caldera systems, like in present-day Pantelleria, Tofua, Los Humeros, or Laguna del Maule systems. But caldera formation is usually related to ignimbrite eruptions from what I’ve seen, and Sabatini has a number of ignimbrites if I recall right.

        • Los Humeros, for example, has voluminous trachyandesite lava flows which issure from fissures along the southern margin of the caldera:

          And of course Yellowstone has the largest lava flows in the planet at up to 80 km3. Okataina has done lava flows with over 10 km3. Although Okataina and Yellowstone rhyolitic lava flows are related to a rift type volcanism, and erupt from long linear dikes, which is a bit different from other calderas.

          But yes, caldera systems do impressive effusions.

        • I dont know if you saw my other comment, but looking at the south coast cliffs of Kilauea, the hilina volcanics look like layered ignimbrites and lava flows of large scale, not like today where the lava flows are large but slow long lived shields. Kilauea in the Pleistocene might have been a flood basalt/ignimbrite rifting caldera. The layering is quite visible even on google earth on the cliff below Hilina lookout, which is on a hill that is the only part of Kilauea with a pre-Holocene surface.

          • I do wish the cliffs of Hilina Pali and others were better studied, it has been decades since the last time someone took a look at them and published something. As far as I know there are no major ignimbrites though, there are, however, pyroclastic surge and airfall scoria deposits of much greater thickness and extent than those of the better documented explosive events of the past 1000 years.

          • The way things are going now, with a large summit caldera and apparently rapidly expanding magma storage in the ERZ, it might not be too far in the future before suc hactivity becomes the norma again. Both of those sill complexes on the ERZ are probably under 1500 years old, the one under Makapohui and Napau probably began forming when Kanenuiohamo did, probably about 1000 years ago or whenever the dates are. But probably at least half of that complex is probably formed this century, some in the 1960s and 70s, but most of it probably from 1975 to 2020 with the growth of Pu’u O’o. The complex down at Heiheiahulu probably formed mostly in the first half of the 18th century, maybe in the 1740s when the first flows down to Kaimu happened, and then with the growth of Heiheiahulu and then the fissure eruptions further east in the 1760s and at the time of Cooks visit. The summit complex maybe is more ancient, but it has probably expanded in that time too, although possibly also moved south so that now the northern part of the caldera is inactive.

            Maybe in 1000 years we will see a much wider summit comples that might include all of the summit out to the Koae faults and large circumferential fissure eruptions from the Koae and Kaoiki areas flooding the land south of the summit again. It is frustrating sometimes how HVO has got such a detailed view of the summit in the past 2000 years but has hardly looked at flank eruptions even to the point of thinking that the last 70 years has been ‘unprecedented’ and is a freak event. And nothing at all from the Pahala ash and Hilina formation apart from the idea that it was created the same way as more recent eruptions, which might be very erroneous assumptions. Really, 2018 was an eruption that was way bigger than they ever thought Kilauea could do, bigger than ANYONE thought it could do… But the caldera it made, was tiny compared to the size of the existing total caldera, not even 1/10 the size. So evidently eruptions a lot bigger than 2018 are possible and have happened even pretty recently, given the caldera is only 500 years old. And the Power’s caldera was even bigger than todays caldera, which implies that the caldera we see today still has more collapsing to do before it is fully formed…

            It is hard to find a good comparison to what Pleistocene Kilauea would be like, Sierra Negra might have the caldera size and scale of eruptions, but it doesnt have a long rift zone or any explosive activity. The Icelandic volcanoes do have rifts and large calderas but individually none of them are nearly so active, and neither feature dual summit/rift activity, they have one or the other but not both… Add to this that Mauna Loa was a glaciated volcano that probably erupted explosively most of the time and Hawaii was a completely different place back in the ice age.

    • 25km north of the city, which has as many as £4m people present during the height of tourist season.
      That Alberto scenario regarding the Diego Armando Maradona and a classic saturday night at the football would be amplified in the case of Rome. Though probably a better organised evacuation.

  21. I am getting real tired of the IGEPN. It looks like Quakes have been sparse at Potrerillos and Chalpatan but have been ramping at CCN proper over the past couple of days. A 4.2 was recorded just at 1.1 km depth at the system, and I can see the damn tremor on the seismographs. Where’s the hydro-thermal data? Where’s the deformation data? Where are the LP earthquakes taking place? Where’s the magma now? Is the hydrothermal system being pressurized? Why has deformation exploded? What’s the cause of this swarm? I can only speculate and I am starting to lose that ability to do that because the IGEPN won’t publish any good data.
    For the love of God, can someone else contact them? They won’t answer any of my messages.

    • What’s the cause of this swarm?

      It is possible that IGEPN doesn’t have a clear answer to that question, so prefer not to speak up.

      There is always the old way, when the volcano explodes everyone panic and run for your lives, just joking. I hope they make a timely prediction before the eruption. But who knows what will happen really.

    • Welcome to the joys of Spanish volcano watching.
      They data hog like crazy there.
      It is an infuriating part of life, but nothing that we can do anything about, so please do not blow a blood vessel about it. 🙂

      • I guess you’re right but it’s almost torture to be robbed of this sweet data at a time like this

        • I can’t speak to the data or what the volcano is likely to do, but the fact that the population adjacent to the volcano has been undergoing bi-national evacuation drills at least suggests a decent level of concern.

          In other words, if an eruption were considered unlikely, I’m not sure they’d be wasting time and money in large scale evacuation drills.

          I think the situation is tenuous at best. I really hope this won’t be a surprise eruption that catches people down there unaware, but you have to believe there are at least semi-competent scientists in the area tuned into the complex like hawks.

          What will CCN do? Well short of getting better information that will allow you guys to get a better read, this one may just come out of nowhere. Or not.

  22. California Geothermal Consultancy request!

    Hello, if you have knowledge about geothermal energy and Californian resource permitting in general, we at Mantlepower would like to have a chat about a potential consultancy gig.
    Please contact me directly at if you are interested, or if you know someone who might be interested.

    P.S. We normally do not accept work ads like this, but since half of the Admins work for Mantlepower we sort of can give ourselves a bit of leeway. If you have any similar things you want, please contact a friendly admin first for permission. Regardless of this, I have punished myself with a cup of sub-par coffee. D.S.

  23. Carl what sort of timescale do you think before the eruption resumes at Fagradalsfjall? It looks like it has well and truely stopped now, not even spattering. Would assume it is on the order of days if pressure is still high.

    • I need a couple of more datapoints from the GPS before I can say anything really, but I would say within days. The GPS-systems will be our friends now.

      • 🙂

        At least it isnt likely to erupt in a dangerous location, there are few people likely to be north of the eruption area. Although I suppose it ispossible the dike might extend, if it erupts south of the 2021 cone or in Natthagi that might be a different story…

        Probably just a case of days, weeks or months, Fagradalsfjall is probably going to be Icelands most active volcano this decade, it will erupt many times more.

        • Fissure eruptions almost always have follow up eruptions upstream of the fissure/dyke. And the conduit to the mantle is very near Keilir.

          • Carl:
            Do any current geology maps of Iceland show this? I am curious to see a map of this behavior.

          • I have a map somewhere, but it last happened at Fagra last time. And most famously at Laki.

          • Sometimes, but not always. Holuhraun opened a small fissure closer to Barðarbunga than the main vent, but it was short-lived and the main vent remained the main vent for the remainder of the eruption. And Hawaii 2018 was all over the place, with fissures opening in an essentially random order, before eventually a vent near the middle became dominant.

            Even last year’s activity was not so tidy: though the first fissure to open was the farthest from Keilir, and the next two opened progressively farther north, the remaining fissures then opened in between existing ones, and the eventually dominant vent was in the penultimate fissure to open and the second-farthest south of the lot.

            Terrain is likely playing a role here: in all these cases, fissures were also at varying elevations, and it takes more pressure at the source to force magma to farther fissures, but also to the surface at higher-elevation ones. The small second Holuhraun fissure was at higher altitude. The eventual dominant vents in 2018 and 2021 may have been compromises between altitude and distance from the source, and in particular lower than the vents that were closer to the source. (At first. “Nar”‘s cone eventually became higher than any other vent at the site, but by then the others had likely solidified internally. And toward the end the “Nar” eruption was jetting sideways out of the base of the cone, and surfacing at other points nearby, rather than having to climb up out of the cone, as a workaround.)

            Another factor might be the local stress field. All of these eruptions occurred in extensional settings, and a vent in a spot with more tension might open more readily than in a spot with less, other things like altitude and distance-from-source being equal. The question is whether there can be much in the way of short-range variation in the tension. In the top few hundred meters, I’d expect the answer to be “yes”, with nearby elevated terrain to the sides of the dyke counteracting tension locally with their compressional force as they’d “like” to sink and spread out. So a fissure between two hills might be preferable to one on a hill, but not to one on flat ground a little farther from the source or a little higher. This might have contributed to the notable absence of fissures opening within Nátthagi, where both hill-squeeze and distance from the source would have both been factors against, and only elevation in favor.

          • I wouldnt say Kilaueas 2018 eruption was random, it did show uprift progression. The initial dike started from a magma storage area that is underneath the 18th century shield of Heiheiahulu, a bit west of the main highway 130 that crosses the ERZ. That dike went as far as Pohoiki road before eruptiosn began and those generally speaking migrated uprift almost opening vents back to the start of the dike west of the highway. But when the south flank slipped and the big quake happened it changed everything, the pressure increased and the dike began advancing again, so eruption resumed further east and new vents opened. Those vents had a moreprimitive composition with more of the new stuff. Fissure 17 was made from the new magma heating up and mixing with a pocket of dacite magma that was already known to exist under the PGV area since 2005, so was sort of its own thing. On May 18th fissure 17 fountained to a great height (well over the 100 meters HVO reported) and fluid lava began to gush out of the eastern fissures, before those also began to fountain. The center of major lava effusion went from F20-22 on May 19th to 23, F13 and 6 on May 23-25, F1,3, 21 and 7 on May 26-27 and then F8 and 2 on May 28, which also saw F24 open. That night was when the fissure eruptions peaked, the only time it produced a proper curtain of fire about 90 meters high and 2 km long, with an output of over 1000 m3/s. F8 then formed its cone from this. The eruption peaked in July of 2018, that is when the caldera beyond Halemaumau began to subside, it reached peak effusion rates of over 2000 m3/s bulk rate, which is probably the highest eruption value recorded in any lava flow over a prolonged period in recent time.

            There are some places this progression probably doesnt work though, when the volcano in question is not flat, or is very flat with no peak above the dike. Mauna Loa doesnt progress back to the summit on the timescale of a single eruption, usually it is the opposite. Mauna Loa is not a steep volcano by any means, but it is also definitely not level. Kilauea ERZ has an inclination of under 2 degrees, while Mauna Loa is about twice that, 2.6-3, and rather steeper at high elevation. The Icelandic volcanoes are even flatter than Kilauea, Veidivotn 1477 vents happened along 70 km of rift and there is only 200 meters elevation difference from the highest elevation to the lowest, and that is mostly because of older terrain giving an uneven surface. Laki craters are very similar, although over a shorter distance. Grimsvotn and Bardarbunga are both rather tall though, summits more than 1 km above where those eruptions happened, so presumably would push magma down more than Kilauea does, but not nearly so much as Mauna Loa which is twice as tall again.

          • Kilauea 1955 bounced all over the place, too. It even stopped for two weeks. 1840 started high on the RZ and went ‘downhill’, 1960 was limited to a small fissure.
            I’ve given up on trying to predict the future based on the past. More enjoyable that way!

          • Kilauea has a lot of ways to erupt at that elevation. 1955 was really two eruptions, because it was two dikes, one began under Pu’u Honuaula where PGV is, and went east to erupt quickly and die off quickly too, with high erruption rate. The other started in the same place as where the 2018 dike did and it erupted more slowly and built up a large cone.

            1840 was completely different, a dike breaking directly out of the upper ERZ conduit and moving at a shallow depth all the way to the LERZ, no fountains and a very high olivine Mg content like magma erupted at the summit, it was probably a lava flood like what Nyiragongo does or what happened down on the SWRZ in 1823. It was a very different mechanism to most ERZ eruptions that begin from magma that is already in the rift.

            1960 was a dike that mostly mirrored the first 1955 dike, starting under where PGV is now, only the greater pressure in the volcano after 1959 and rapidly increasing basal supply rate meant the eruption was much more powerful. 1960 was as strong as 2018, but the magma accumulation at the summit was more limited so it stopped quicker. The volume has been increased from earlier estimates to a conservative 0.25 km3 and might be rather more, it was almost a caldera formation event and probably the biggest 20th century eruption from Kilauea before Pu’u O’o started.

  24. There still seems to be magma supply present. At this point there is still noise. Even with the 2021 eruption’s on and off phase the blue plot would go between 1000-2000. Right now it is at about 5000. That’s another sign that brings hope that new fissures will pop up or this vent will reactivate. It is still decreasing, but not fast which may only mean a slow retreat of magma currently.

  25. Wow, after watching the livestream on YouTube very intently, I could still see some spatter come up every once in a while. I guess the pause isn’t official yet. The day time illusion along with the very low activity has made it impossible to see on every other stream.

  26. Current drone livestream from Isak shows some lava is still flowing from a tube, though nothing but a small glow remains in the crater itself. Seems this fissure is still (barely) alive -for now.

    /Non working link fixed by an Admin without a clue of how he did it.
    It is now proven that Admins work in mysterious ways.

    • You can also see hole within the crater where the lava is laying beneath. I wonder what the cause is. Maybe there is enough lava for the fissure to remain active, but first more supply from the mantle has to reach it? After all the average output for the first few days was notable higher than the supply rate. In this case activity should probably pick up within the next few days.

  27. According to Volcano Discovery a 3.9 Magnitude Earthquake with a depth of 0.6 km and located 1.5 Km SouthWest of Kelir occurred a few hours ago. Unsure if this is adjusted or preliminary estimates but this stands out for sure to me.

    • Only 50.5% reliability on, this needs official manual checking before anything can be said about this.

    • That’s a false alarm. Quality is 50.5 and there’s no sign of it on the drumplots. The automatic system got a hickup and it will never survive manual review.

      • Thank you for letting me know. Well that false quake is sure a tease then.

  28. It appears that both “Ofu-Olosega” and “Ta’u” have had their aviation codes raised to yellow as swarms seem to be occuring under both volcanoes.

  29. I initially swayed to the prospect that new fissures will pop up, but it is becoming clear that is less likely. There are no new cracks, the tremor is dropping fast and also quakes are at a low level. Even the last few deformation measurements aren’t signalling rising movements compared to the week and a half before. Unless inflation starts within the next week I really don’t see this resuming. Perhaps we can expect a new eruption nearby or somewhere else in the Reykjanes in the next year between spring and fall. There must be plenty of supply deep down, the conditions just have to be almost perfect for big and lengthy eruption

  30. And that was it at the 2022 eruption of Fagrafjall. It is though possible that a new fissure will open on the same dyke to the NNW.
    Rest in Peace little eruption, you were a cute one!

    • What the hell went wrong? This was supposed to last a few months, not three measly weeks.

      • I am glad it could be over so that way Taal, Cerro Negro, and Ioto can finally take over the discussion.instead our tourist volcano.

    • Cool little eruption. Great while it lasted. Especially the show on the first day.

    • Thanks Daniele Bianchino. Colli Albani is a volcanic wonder.

      Your page is amazing, lots of information, maps, and images, for someone like me who loves visualizing the information. Those models are great too. Have you ever thought of painting them according to different volcanic units? Geologic maps are sometimes true works of art, combining them with those models could have interesting results.

    • Just another thought. On landing at Fiumicino this past June, my view from the window was to the South. I assume that the edifice which piqued my curiosity then was indeed Albano.

  31. Thank you all. In fact, I had thought of coloring the model of the Alban Hills also on the basis of volcanic units. Sooner or later Im doin’g I try and show you. Buyers generally ask for gray or neutral concrete models 🙂
    Here other models of volcanoes, if you are curious,

Comments are closed.