Chiles Cerro Negro. Source: https://www.igepn.edu.ec/chiles-cerro-negro
Over the past 6 years I’ve watched Chiles-Cerro Negro go through 4 different phases of unrest. Like a toxic lover, no matter how much I try, I can’t shake this volcano. I honestly thought I was going somewhere with Grimsvotn and Iwo-Jima but no. Grimsvotn is stable in the sense that nothing has really changed besides a probable failed eruption in 2023, and there is so little study about Iwo-jima that I can’t say much more than what I’ve already said. Other volcanoes have either failed to deliver on their unrest, or their eruptions just don’t interest me. Chiles-Cerro Negro is different, I don’t think there has been a volcano that I wanted to erupt more than this. 1.3 million earthquakes in 10 years and accelerating deformation are only the tip of the iceberg when it comes to this volcano. It’s funny, in February Chiles-Cerro Negro had an average of 300 earthquakes/day and the only thing I could think of is how relatively quiet things were. A 100 earthquakes/day at this volcano is considered normal and boring. After another year of watching this volcano go through another major shift, it’s safe to say the volcano has met its final and potentially greatest obstacle that precludes an eruption, and I will explain why I believe that this volcano is either going to erupt big or not at all.
Epsilon Swarm
In my previous article for this volcano, I listed 4 major swarms as points of interest.
- Alpha Swarm (2013-2015)
- Beta Swarm (2018-2020)
- Gamma Swarm (2022)
- Delta Swarm (2023)
After the 2 seismic crisis of 2022-2023, Chiles-Cerro Negro has seemed docile and passive in 2024 with the whole year seeing no more than 40,000 quakes. However, the previous year saw an explosion of VLP earthquakes that had not been seen before at this volcano. Over 1,200 VLP earthquakes and 1,100+ pulses of tremor took place, with the trend continuing this year. Energy levels for these earthquakes are beating the previous record by a factor of almost 5,000. As such, regardless of the low numbers and relatively low energy, I consider this to be a new swarm due to its novelty for this system and what it represents.
The location of the strongest VLP quakes has been 3-5 km below the summit of Chiles, where there is almost definitely a plug. It is no longer just my opinion, but the SGC (Servicio Geologico Colombiano) has also pointed out this possibility. No significant changes have been noted with the hydro-thermal system, so this is likely the result of magma and/or magmatic fluid moving within the edifice of the volcano. There have also been significant shifts in deformation, but that topic will be saved for later. If there is magma accumulating beneath the edifice, the volume probably doesn’t exceed 31,400-100,000 m3/year, so there is no need for immediate concern, as it would take at least a decade for relatively minimal disruptions to the edifice, and there is little chance of significant build-up of shallow magma with these rates. This is of course assuming that the cause of these VLPs is accumulating magma which it may not be.
I’d say an eruption is completely impossible if things continue at this rate, but things rarely stay the same at this volcano, and this swarm could be our potential canary in the coal mine with this system. In fact, with context, I might consider this swarm to be the scariest out of all the swarms so far.
In order to understand why this swarm might be the scariest of them all, despite being the weakest, we need to understand the extremely convoluted and complicated setup of Chiles-Cerro Negro and how we got to this point in the first place. Before I go into detail about this, I just want to say that this is all my speculation, and to be frank, I am still somewhat confused by this setup, and my hypothesis is my way of reconciling the oddities in this system. It doesn’t help that there is new major information and/or development at this volcano almost every year. Over the past 6 years, I have read almost every weekly, annual, daily, and monthly report from the IGEPN and SGC. After reading hundreds of reports and with some additional perspective I’ve gained from reading about the recent Fagradalsfjall eruptions, I think I’ve come close to a definitive answer to the overall cause of the unrest at this volcano.. I will go over the entire 12+ year history of the unrest so far and list points of interest.
Alpha Swarm
In 2011, the Cumbal volcano to north of Chiles-Cerro Negro, started a new unrest period relating to it’s hydro-thermal system. Nothing initially spectacular, but the SGC raised the alert level to yellow in 2012 nonetheless. There was no magma intrusion, but this is not required for a restless hydrothermal system. This period of unrest has continued on to this day with variation. The volcano saw 8,000 quakes in 2011, but in the present, it’s not uncommon for there to see over 24,000 earthquakes in a year. We can see this trend even more clearly within in the inclinometer data.
It is slightly odd for this volcano’s hydrothermal system to be experiencing such a long and slow increase in activity over 14 years in conjunction with the magmatic system of Chiles-Cerro Negro. Both volcanoes share a regional fault system, but Cumbal’s unrest started before the first swarm in 2013. I had long since dismissed a possible connection until 2022, when I realized that Cumbal’s hydro-thermal activity was likely regulated by tectonic faults, a lot of which it shares with Chiles-Cerro Negro. In any case, this is our first point of interest.
It is unclear when exactly the swarm started, but seismic activity was first noted in February 2013, and the swarm was recognized in November of that year, as new instruments were added. Unfortunately, as this was the first real sign of life from a long thought dead volcano, there is almost no meaningful data about the volcano before the swarm besides very basic details about its structure. Although the crisis technically started in 2013, it would truly peak in 2014. A strong area of inflation developed to the south of Chiles-Cerro Negro with peak rates of 20 cm observed in May. Interestingly, the swarm would have 3 peaks with significantly low activity in between. Despite the intrusion likely beginning some time in July or earlier in 2013, it wouldn’t be until October-November 2014 that we’d see the peak with 240,000 earthquakes, including a damaging M5.8.
In the month preceding the crisis, the hydrothermal feature showed a significant rise in heat and acidity, but as the crisis started, the system stabilized despite the escalation in volcanic activity. This gives further support to the fact that the hydrothermal system is more influenced by the tectonic system. As the volcano was building stress, the hydrothermal system was disrupted, but as the stress was released, the system recovered. This fact is very important in putting current events in context, so keep a note of it. As we move on. Seismic activity would slowly begin to wane, reaching moderate levels in May 2015. Uplift would stop but begin again with no associated seismic swarm. The new deformation pattern that would follow for years to come would be that of 18-20 km long “crack” or tilted dike (Crike, I’ll call it) that would act as a conduit for magma to flow from Potrerillos to the Chiles cone. This will also be extremely important.
From this swarm, there are 3 important facts that are crucial for understanding this volcano.
- Cumbal volcano started having seismic activity in 2011, 2 years before Chiles-Cerro Negro’s swarms started in July 2013
- The hydrothermal system reached it’s peak in 2014, likely caused by the buildup of strain preceding the seismic crisis
- Some type of crack or dike had developed following the seismic crisis that facilitated the transfer of Magma from Potrerillos to Chiles
Geological Neighbors and the 2016-2017 quiet period
Nothing much of note would happen in the 2 years following the crisis. Seismic activity fluctuated but remained mostly stable, and deformation was stable.
We are left with one lingering question. Why is Potrerillos deforming and what is it’s connection to Chiles-Cerro Negro? Is it a coincidence that Cumbal’s hydrothermal activity spike just 2 years before CCN would go crazy? Let’s see. Chiles-Cerro Negro is surrounded by 3 other significant and 2 minor volcanoes. Porterillos, Chalpatan, Cumbal, Horqueta, and Chiltazon. These volcanoes are within 30 km of CCN and are experiencing interesting shifts. Poterillos and Chalpatan have seen uplift and seismic activity, Cumbal has seen hydro-thermal activity, and these other poorly researched no-names are in the subsidence zone.
The volcanoes are born from the subduction of the Nazca Plate beneath the South American Plate, which is the progenitor for all of the Andean Volcanoes. This complex is built on top of thick basaltic flows from over 2 million years ago despite CCN proper being far younger. The complex has experienced several eruptions, with the most significant ones occurring around 10,000 to 15,000 years ago. It’s eruptions are a healthy balance between effusive and explosive, with the last eruption possibly happening around 2,000 years ago. Shown by further study, the volcano may not have been as dormant as once thought.
The Potrerillos-Voladero system is encircled by andesitic lava hills, a morphology indicative of a ring fracture from a small caldera collapse. This ring is most prominent along the southern, eastern, and northeastern margins, with the latter segment marking the boundary with the Chalpatán Caldera. While no eruptive deposits are visible within the caldera, roadcuts along the trail from El Voladero Guard Station to Yegua Rusia expose volcanic material linked to gravitational flows.
Notably, a white ash layer rich in silica and altered to clay under acidic condition was identified in these exposures. This layer, bearing biotite and hornblende, is interpreted as a product of hydrothermal or phreatomagmatic activity from the caldera. Crucially, it lies below the 10,000 years ago El Soche marker ash implying Voladero’s eruptive phase likely occurred between 12–20,000 years ago.The Chalpatan caldera, a distinct Plio-Pleistocene volcanic structure formed approximately 2.0 million years ago, represents a significant, albeit comparatively less studied, component of Ecuador’s Northern Andean Volcanic Arc. With an estimated diameter of 6-8 kilometers, it still boast a potent geothermal system with magmatic component and has numerous fault across it.
With so many volcanic features in the region being connected directly or indirectly to the unrest at Chiles-Cerro Negro. I don’t think it’s unreasonable to assume that this could just be one big volcanic field. The area’s history is poorly researched but it’s not without a degree of merit when consider the scale of unrest at this system.
During this period the amount of magma required to cause the observed deformation (pre-gamma swarm) was calculated to be at max 10 million m3/year from 2016-2020 with likely higher rates in 2022-2025. Which means our magma intrusion has likely accumulated 90 million m3 on top of whatever amount of magma it started out with.
2018-2020 Beta swarm
Formerly the most lacklustre swarm from this volcano, consisting of “only” 210,000 quakes. This swarm would be the most well-documented and the swarm that would spurn my interest in the volcano. The cause of the swarm wasn’t clear; it wasn’t hydrothermal, the magma intrusion didn’t get any bigger, no significant magma movement was noted up until the very end of the swarm, and yet up to 4,000 quakes/day took place during these years. With 2 distinct unrest areas, the most quakes took place at Chiles-Cerro Negro but the fastest deformation taking was at Potrerillos. Chalpatan didn’t miss out either with the activity overlapping into it’s region.
The activity, barring one period, took place right where the “crack” or dike is. The swarm would conclude with a swarm of 30-48 km deep LPs that happened on the other side of the volcano, at Cerro Negro and further NE up to 10 km away from the Chiles cone. Oddly, there was no uplift where these quakes took place, in fact the area of uplift is surrounded by a ring of subsidence of varying rates but usually 1-2 cm/year. INSAR data is not very good in this area due to poor coherence caused by the rough terrain, but the general trend is seen on all available maps and covers an area of roughly 30km by 30km. This subsidence is not seen on any other GPS stations on Southern Colombia or Northern Ecuador outside of the Chiles-Cerro Negro.
Given the events that have followed this swarm, it either had something to do with the pressurization of either the “crike” or magma chamber or heralded the recovery from the 2014 seismic crisis. Spoiler warning for the future events, but what’s happening right now couldn’t be happening without the volcano doing one of these things. After this swarm, Chiles-Cerro Negro chugged out 2 seismic crisis, a 3rd swarm and is likely about to do another crisis. Just think about the scale of this, a potential pressurization or recovery event caused 210,000 earthquakes, I am not aware of anything remotely on a similar scale happening anywhere else. Just due to it’s strength and with that our “Crike” can over pressurize and/or recover without such significant seismic activity. It would seem likely that this was more related to the magma chamber.
The Great Quiet and 2nd Seismic crisis (Gamma Swarm)
Following the suspicious movement of deep magma, the volcano seemingly shut down. Seismic activity went anaemic with rates as low as 10 quakes/day, and uplift stopped. However, the inclinometers went insane with changes of over 100 microradians/day and the ring of subsidence continued unchanged. At the same time, Cumbal saw an uptick of activity. 2021 would be the strangest year for the volcano, given that the 2nd crisis would happen with just a couple weeks of buildup following this year. It doesn’t make sense for the unrest to have paused, considering that the 1st seismic crisis had over a year’s worth of strong build-up. I am genuinely at a loss on what was happening at this volcano during this time. My idea is that during this period, after the pressurization phase, there was brief stability period as the crust reached it’s limit and briefly stopped stretching and cracking. That ,or magma supply was somehow paused but that doesn’t really make sense to me as the “Crike” was still active following this period. For the later solution to be true, it would mean that the “Crike” would have a pretty significant volume of magma for it not to solidify after over 15 months of no supply.
Starting Late May 2022, the 2nd crisis would begin with over 200,000 quakes in 4 months and another damaging magnitude 5.6 quake. Interestingly, deformation rates quadrupled from it’s previous rates to 12 cm/yr before the swarm started. At the same time, the swarm of rapidly shallowing LPs quakes took place indicative of magma rising from the depths of down to 38 km below the surface to just a few km below sea level. Once again activity was concentrated exactly where the “Crike” is. Uplift would abruptly pause and restart following the swarm, and the inclinometers would start to stabilize after the swarm ceased. The most seismic energy was released at Potrerillos and Chalpatan although CCN proper still saw major seismic activity.
With the fact that significantly more magma movement and uplift was noted with this swarm than with the years before, it’s safe to assume that something big broke and the “crike” expanded allowing more magma to flow through it and once again, all of the activity lined up with Crike perfectly. At this point, it should be pretty clear that the deformation and seismic activity that we’ve been seeing is related to whatever structure developed after the first seismic crisis and not to the magma chamber directly. As we’ve established, the deformation follows the crack or dike perfectly and without error. A measly >10,000,000 m3 isn’t going to do much for a 1,200+ km3 magma chamber but it would keep a small magmatic structure from solidifying. This isn’t just my speculation; this is the official explanation. The following is from Pilicita Masaband, Mothes, 2025:
On average, vertical deformation reached ~40 cm of uplift over ten years, and continued even during non-swarm periods (Figure 12a). Both GPS and InSAR methods showed the same trends (Figure 12b), distributed across several stations or control points in the InSAR time series located between Voladero and Lagunas Verdes. Given this specific case, the deformation sources were modeled using SarScape software, concluding that the Potrerillos deformation source is 18 km long and 5-6 km wide, resembles a “crack” opening, and its depth drops between 6-8 km below the surface (Figure 13a). The other source is Lagunas Negras, which is shallower, and its displacement suggests a tilted dike (71°). In both cases, a magmatic intrusion is considered to be the cause of the deformation observed at the surface.
The mostly seismic activity and most visible deformation that has taken place since 2015 has never been directly related to intrusion into the chamber proper; the accelerating deformation and increasing seismic activity is the direct result of a magma intrusion from the magma chamber. The subsidence that surrounds the volcano is likely the result of magma leaving the magma chamber. This is somewhat similar to the current situation at Fagradalsfjall, where the activity we see is the result of a deep magma chamber sending a constant supply of magma to a shallow level that would later serve as the main source for an eruption. The implications for this happening at Chiles-Cerro Negro is not pretty, this would mean that we don’t need to wait on the chamber to be pressurized or get more and more molten. The magma reservoir is already primed and ready to erupt and what we are seeing how is not the system oiling it’s gears like other volcanoes, we’re seeing it try to break through the crust and relieve itself. This notion is supported by current events and what happened the following year.
3rd Seismic Crisis (Delta Swarm)
Not even 6 months after the end of the second crisis, the third crisis began on March 11, peaking immediately with 11,000 quakes on the next day. This swarm would almost be entirely be located at Chiles, right at the top of our proposed “Crike” and would have more LP quakes than any other swarm before it. Despite peaking early, it lingered throughout the entire year producing over 270,000 earthquakes, more than the previous swarm but not as intense. Uplift abruptly paused and restarted no less than 3 times during this year but a peak speed of 36 cm/year was observed at Lagunas Verdes to the south of Chiles. Towards the end of the year inclinometers would start to show more instability, a trend that has maintained itself until now. Shallow LPs, VLPs, and tremor has been observed within the edifice of the volcano following this swarm mainly taking place 2-5 km below the summit.
In my opinion, the swarm was the result of shallow magma hitting the plug and destablizing the area. Cracks around the volcano following the swarms have been noted, but still no direct effects on the hydro-thermal system. It should be noted how quickly the volcano recovered from its tantrum in 2022 to produce another one just 5 months later. This speaks to how quickly the volcano,”the crike” more specifically, is now accumulating strain. During this period, the trend of subsidence surrounding the volcano might have completely reversed with more areas undergoing uplift that we’ve not seen before. As a result of these swarms, the top of the “Crike” is now just 1-2 km below the surface with magma or related products potentially leaking into the edifice.
The fact that this swarm took place so soon after Gamma Swarm with no build-up and no seismic activity relating to pressurizing gives credence to the notion that if pressurization processes were the principle cause of Beta Swarm it would be from the magma chamber. In the 10 month span encompassing the totality of Delta swarm, only 6 months had high seismic activity and the other months saw only moderate levels of activity. If we were to assume that the “Crike” was accumulating strain in the weaker months and add up the 5 months repose period between the swarm we find that at minimum 1.47e+12 Joules of seismic energy was gathered in 11 months. This would give the “crike” a minimal seismic energy accumulation rate of at 1.37e+11 joules/month. If we were to assume that all of swarms’ energy was accumulated during the repose period, then double the number.
The 4th seismic crisis and the Current Swarm
Now that we have context, it becomes clear why the Epsilon Swarm is the scariest so far. These unending VLPs are indicative of the “Crike” being capable of effectively delivering magma right to surface, only being stopped by a strong plug. A large felsic pressurized magma chamber is trying to break through a strong plug and erupt. We also know that the area is rapidly accumulating strain given that the system completely recovered from a seismic crisis and produced another in just 6 months. Since February, a general uptick in VT quakes have been noted, in pulse like miniswarms lasting about 1-2 weeks. This signals that things are starting to break again. Whether the plug will fail or not is a different story. About 3 of these mini-swarms have taken place, each one stronger than the last, with the recovery time in between dropping quickly from months to weeks.
It would seem that something has happened with our “Crike”, the uplift at Laguna Verdes, where the top of “crike” is, has completely stopped and the uplift at Potrerillos has slowed significantly but inflation surrounding this area has increased. Normally, this would be seen as good but last time something like this happened we got a large swarm a year later. As general deformation and fluid-motion related seismic activity hasn’t stopped, it’s obvious that magma hasn’t stopped flowing through our “Crike”.
After over 10 years, the fumaroles and hot springs are now once again reaching temperatures close to what we saw in 2014 preceding the 1st and biggest seismic swarm at the volcano. The Alpha and Beta swarm were probably the only recent swarm that was directly associated with the chamber and the recent activity was related to expansions and instability in the “Crike”. This could mean that the chamber could be reaching stress levels not seen since 2014. Remember, Fagradalsfjall had to produce multiple failed dikes before ultimately erupting, it took 2 years of this before we got magma. The pressure was always there but after producing failed dikes, the chamber would need to accumulate strain to create more. The process paused but never stopped. The dikes and swarms helped the volcano by weakening the crust. Something similar is likely happening at Chiles-Cerro Negro but instead of producing multiple dikes or intrusions, it is expanding and strengthening one. This fact is also supported by the fact that the areas surrounding the “Crike’ reversed from deflation to inflation. So we not only have our “Crike” directly stressing the plug and accumulating strain on its own, but the magma chamber itself may throw it’s hat in the ring. Just one of these things would cause another seismic crisis, but the potential of these 2 things happening around the same time might cause the biggest seismic swarm in this volcano’s in it’s 12 years of unrest.
We have 3 options for the current swarm, either the current swarm continues to escalate into a full blown seismic crisis akin to delta swarm, the aforementioned simultaneous Crike and Chamber crisis, or it maintains it’s relatively moderate scale before tapering off. In my opinion, the first option is the most likely.
Caldera chances and Conclusion
I don’t know if I am 100% correct in my assessment but the fact is that we’re not waiting on the magma chamber to ready itself, it is already primed. I can’t help but to hammer this fact in, a large felsic pressurized magma chamber has triggered a magma intrusion that has remained active for a decade and is still trying breakthrough a plug. Thanks to the plug, the volcano is expanding it’s intrusion and gaining pressure. This is somewhat of an unprecedented situation for a felsic system in modern volcano monitoring. Even if the plug maintains itself for another 12 more years, as long as the chamber is capable of sending magma to it’s intrusion and stays overpressurized, the danger level won’t change. Only an eruption or a larger, failed dike can stabilize the chamber now.
I’ve warned about this volcano’s chance of producing a caldera-forming eruption for years and I’ve only grown more confident in this possibility but there could be one saving grace for this volcano. The magma chamber is deep, 8 km below sea level and 10-13 km below ground. This isn’t a favorable setup for caldera-forming eruptions, the deeper the chamber the more pressure it takes for it to produce a caldera. A shallow chamber needs less pressure to produce a caldera than a deeper chamber. With a chamber as deep as this, it should take a lot of pressure to produce a caldera forming eruption, pressure that may not have had enough time to build. With recent studies highlighting the possibility of eruptions within the Holocene. The dormancy argument is far weaker now. The fact is that this system has only produced 2 calderas at the most so far, so we know that this system has a high threshold to cross before it can do such a thing.
We don’t know the current pressures of the chamber but we might be able to calculate how much pressure it would take for caldera collapse with this formula:
Pu = 1/2 μρg L/Sc H^2
μ is the static friction coefficient of the fault, ρ is the density of the host rock, g is the gravitational acceleration, Sc is the basal area of the caldera block, L is the perimeter length of the caldera block, H is the depth to the magma chamber. Unfortunately for me, I don’t know how I would calculate the static friction coefficient of a fault but if anyone else wants to throw their hands on this issue I’d be happy. The official numbers for structure of the Magma chamber is that it’s 15 km wide and 8 km deep. I doubt this because the dike is sourced nearly 20 km away from Chiles but the chamber still produced LPs another 10-15 km away from top of the dike which could indicate of a minimal width of 30-35 km and consistent magma movement was noted with depths up to 48 km which could indicate a maximum chamber depth of 40 km. It wouldn’t be very scientific of me to use these numbers like they’re confirmed by study but it’s not crazy to say that the chamber size is underestimated with some of these inconsistencies. The most important information that we need now is how much pressure the chamber has now and what are the dimensions of it’s magma chamber?
Chiles-Cerro Negro has consistently defied expectations and thrown us many curveballs, and now as more information is uncovered, the less conventional the volcano becomes. The undeniable fact is that we are in desperate need for more information concerning this volcano.
Images
The images above are from Patricia Mothes et al., Variable Ground Deformation Rates Since May 2022 at Chiles-Potrerillos Volcanoes, Ecuadorian-Colombia Border, presentation given in 2023. https://fringe2023.esa.int/iframe-agenda/files/presentation-488.pdf
References
Geshi et al. 2023: https://www.nature.com/articles/s41598-023-34411-5
Insar modelling of Chiles Cerro Negro, EO Open Science, YouTube presentation, Yepez, Mothes et al:
Koch et al. 2021: https://www.sciencedirect.com/science/article/abs/pii/S0895981121001784
Thanks for this great update on Chile Negro in Ecuador/Columbia, Tallis!
Can we imagine the structure of the magma chamber? Are there horizontal layers of different far evolved magmas or a different mixture of magmas? This has likely an effect on the development of an eruption.
How would the first month of an eruption look like? Would the volcano go very fast from 0 to 100% or do some more gentle phreatic eruptions first for a while? Pinatubo and St. Helens did small phreatic things first, before the famous great bangs happened. Unlike this Vesuvius often did “sudden death” eruptions that happened so immediately that humans had no chance to escape.
Tambora is an example for a great caldera forming eruption. It did perhaps some small phreatic eruptions since 1812, but in April 1815 the development was very fast and suddenly within five days from first ordinary Plinian eruption to the main Caldera Eruption. A Caldera Eruption is for me the label for the classical caldera forming eruption that leaves a wide caldera behind like Tambora or Crater Lake (Oregon). It is the size between a classical Plinian eruption of a volcanic mountain and a Super Caldera Eruption (VEI8)
How much has Chiles-Cerro Negro with neighbouring volcanoes in common?
I can’t imagine that it would erupt like Tambora, the chamber is far deeper and dynamics would completely different. I’d imagine that due to how deep the chamber is, if it were to produce a caldera forming eruption, it would have a long plinian phase and since the volcano seems to love earthquakes, it’d be unlikely for things to escalate so quickly. I think the initial phreatic and magmatic activity could come up suddenly and quickly, if once magma would break through the plug, it would start to interact with the extensive hydrothermal system, nasty combo. I think it would do a gradual build up like Pinatubo or St helens
This volcano is not really comparable to the others in the region, the closest thing I can think of is the Cerro Bravo-Machin complex but even then that’s almost a completely different situation with only similarity being several volcano being fed by a deeper chamber.
So if the volcano likely can’t collapse into a caldera, it can still blow a lot of ash into the atmosphere and create a big explosion crater. How certain is it that the magma is mainly explosive? Some volcanoes do effusive eruptions with viscious magmas, f.e. rhyolite lava flows or a dome eruption. The magma of Chiles Negro is Basaltic Andesite to Dacite. GVP mentions “andesitic and dacitic lava flows”.
That reference is outdated. In the FRINGE meeting, the IGEPN states that there was relatively recent explosive eruptions and these deposits were rich in silica and water.
A lot of basaltic andesite or andesite at arc stratovolcanoes is felsic lava with mafic crystals that average an intermediate composition. Actual crystal poor basaltic andesite or andesite melt is as fluid as basalt and flows easily, Stromboli erupts basaltic andesite to trachyandesite and is fluid. Villarrica and Yasur have lava lakes of basaltic andesite.
CCN would probably be better compared to a crystal rich felsic magma, basically granite in physical appearence although not in average composition. It is still most likely to be effusive but the vent clearing stage could still be a big explosion, not what Tallis gives as worst case, but probably up to a sizable VEI 5 depending on how open the system is and how much magma is shallow enough to decompress explosively. Something like Santa Maria, explosive opening that opened up a path, and 20 years later magma was able to recover and find its way up again.
That is my guess, first eruption is a blow out, how big is yet to be seen. Then some point later the rest of the magma finds its way up more peacefully. Not likely a caldera.
Great update, Tallis! Your interest/excitement is contagious.
If you were to hazard a guess, how do you think the plug will fail?
Would it fail due to pressure, which I think would produce several smaller cracks/vents as the plug starts to loose continuity at an ever increasing rate, or would the plug fail suddenly due to increased heat from below that softens the roof over a wide area thus allowing a single catastrophic/caldera failure?
And lastly, if the plug remains intact despite increasing pressure/heat, might a new conduit form as magma pushes around the plug till it finds a weakness and opens up a new path to the surface?
I think the plug will fail after either enough strain has accumulated or the intrusion expands again. Another odd thing is that it doesn’t seem like this system is particularly hot as such It’s more likely to fail from pressure. It’s possible that the magma goes around the plug but we’d have to see magma accumulate in the form of a sill first. This plug isn’t just a big block but a layer of older magma so I don’t know if it do such a thing.
Thanks Tallis for the additional thoughts.
My own instincts are that the swarms/dikes could be in part from magma at depth slowly/episodically building another escape route…ie. your “Crike”.
With so many volcanoes/vents in the vicinity, and each one has their own conduit/plug, the lithosphere must be highly fractured, so other weaknesses in the crust are probably present that could evolve before making a final push to the surface thus forming a new vent/conduit. In theory, that would suggest the current plug will continue to hold as there are easier (developing) escape routes for magma to migrate.
Tallis, I am safely rooting for the end of your torture and madness 30 some odd degrees North of CCN!
Part of me is too!
Always looking forward to your articles on the biggest bangs!
I am curious. How big of an eruption do you think this volcano will likely cause based on all currently available data? (i.e. low-end VEI 6, high-end VEI 6, low-end VEI 7, etc. [or using examples like Pinatubo, Quilotoa, Hyaynaputina, Samalas, Hatepe, etc. if a historical analogy would be an easier comparison])
And also, how sulfur rich do you think the magma at Chiles is?
The size of the eruption will depend on how quickly the intrusion can transfer magma since there is no way in hell that it’ll accumulate the magma needed for a big eruption in our lifetimes with these rates unless we can get an even bigger rupture and more magma comes in. All 3 scenarios are possible here and I am personally leaning towards something akin to Haetepe for an analog but this isn’t really based on anything concrete as we don’t know how effective the intrusion will be when it comes time to erupt, However this intrusion might have that potential consider that magma from 38 km depth may have gotten sucked up into the intrusion preceding the Gamma swarm
Haetepe? wow. That’d be spectacular if it wasn’t so equally terrifying.
Thank you for the articles and reply!
Wonderful article Tallis, I think your thinking about CCN is correct, I think it’s just deciding which one breaks first, the plug or the wider area. I think another volcano could be a similar eruption would be the millennium eruption of 946-947 CE, could be a better fit.
I am not privy to the details, is this setup similar to the Millennium eruption? if it is how so?
I think iirc that the millennium eruption had two stages of eruption, I believe that CCN will have two stages of eruption as well with a small break in-between.
An earthquake has happened nearby: https://www.eltelegrafo.com.ec/noticias/nacionales/44/sismo-magnitud-4-8-registro-cerca-volcan-cotopaxi?fbclid=IwY2xjawMN0A1leHRuA2FlbQIxMABicmlkETBaTlA0bTdVa3lZTjBkeFlQAR4DprQisGOCpmcI9yE3F0N9IG2jKOspH_2K7POmmY5Lw2tqS1rjIvEqedj1yA_aem_ixXmBZbkl_YK_S-8Vnsaug
… close to Cotopaxi. It is both the second highest volcano and mountain of Ecuador after both the highest volcano and mountain of Ecuador Chimbarazo. Cotopaxi’s eruption behaviour remind a bit to Mt. Rainiers typical eruptions: often phreatic, sometimes with risk of lahars or pyroclastic flows. 1877 a lahar traveled both into the Pacific Ocean and the Amazonas basin. A completely different volcano to Chiles-Cerro Negro
Local tsunami in Alaska caused by big landslide (Geohub made a video about this): https://earthquake.alaska.edu/index.php/major-landslide-southeast-alaska-fjord
At Sawyer Island in the fjord the tsunami grew to a local mega tsunami height of 100 feet = 30m
Interesting analysis of the seismicity and deformation, thanks! Though I didn’t see any geochemical or gas data, which is also important for understanding unrest.
I was following your argument up until this formula: Pu = 1/2 μρg L/Sc H^2
You were talking about “caldera collapse” here. But in order for a caldera collapse to occur, we must first breach, and empty, the magma reservoir – so surely the relevant stress calculations would involve the hoop and radial stress fields?
I indeed read this equation as the shear pressure required to cause failure of the caldera ring fault. For vertical shear, this requires over- or underpressure of the magma chamber, some of which can come from the lower density of the magma compared to the surrounding rock.
An unusual star at Langjokull. That is in Iceland, in case you wondered, but not an area rich in stars
Rock fracturing from the looks of the drumplots?
Kilauea has broken its normal trend of recovery. SDH has completely recovered and is going well beyond even the peak point of E30.
But UWD has shown an exponential decline in supply as it has approached the trigger point, and is now almost flat. This is at the same point that the E30 runup had 1 or 2 DI events, so this point now might actually be the limit of the system working as expected. It shows now pressure to erupt is too high for only the Halemaumau magma chamber to be able to do it, and now pressure needs to build in the whole summit system too.
E31 is probably going to be different somehow than before. Like E30 its pretty likely flank vents will be created or reactivated. But the uplift at SDH compared to UWD also brings the whole broader summit into play too, and possibly the SWRZ. It also makes it much more likely that the intervals get longer and volume per episode, possibly also intensity and fountain height, get significantly larger too as the year goes on.
The GPS data show that the difference between minimum and maximum points of deformation has grown during the recent episodes. The swinging increases. HVO’s website doesn’t load now, maybe there is an internet problem.
I expect a breakout in the eastern rift zone could become likely.
Only if there is inflation observed along that rift, like there was September last year, the ERZ is a magma storage area too, each pit crater is basically a satellite volcano, with similar behavior to the paret just smaller scale. As of yet neither rift connector is seismically active, so despite seemingly SDH being pressurized more, its still less than last year, because there is an open vent to relieve pressure.
At some point that will change but its still far from clear yet. HVO is still analysing the samples but the INSAR suggests the south fissure of E30 was a shallow satellite vent of the existing vents, not a new intrusion from the magma chamber. E31 is about to begin and might change this or not.
Major Hurricane Erin makes a cruise through the Caribic Sea: https://zoom.earth/storms/erin-2025/
Now it’s a category 4 hurricane. It won’t cross any land, but comes very close to the Bahamas and will send high waves the the US east coast. After transition into an extra-tropical storm it may still cause significant weather afterwards in western Europe.
Erin is quite a storm. It went from category 1 to 5 in 24 hours which may be a record for intensification out in the Atlantic ocean. (Gulf storms can do this faster because the water there is so warm.) It is also so intense that it is upwelling colder water even underneath its core – I had not seen that before, but it acts to self-limit the storm’s intensity. Luckily it seems to miss every bit of land along the way.
In September 2005 I experienced the impact of ex-Hurricane Maria in Bergen (N). Although the city is famous for most rain in Europe, this rain (150mm) was much more than usual with landslides. https://www.terradaily.com/news/hurricane-05zzv.html
Ex-Hurricanes still keep a lot of heat and energy with them. They can build a strong warm front that brings a very late summer heatwave in Europe, can develop a strong September storm and rain.
https://www.youtube.com/watch?v=sTXKIz2NGvg
Episode 31 will likley start up at any moment
And thus, it was jinxed.
or called a few days early
Same as last time the episode has stalled right at the cusp of beginning, its probably the same cause in both cases. The exact cause of DI events is still unknown but the best ideas are magmatic crystals flowing back down the magma conduit and temporarily blocking the flow up. Or the magma chamber walls collapsing in and doing similar. Its a sign of pressure increasing all over. High pressure will also kill the supply rate, which will go down significantly until he episode starts, showing as a slowing or stopped upward tilt.
E31 will probably be a few days off, at least 1 more. But it will probably be the most intense and voluminous episode yet, E30 was 8 million m3, E31 might be over 10 million. The low fountains of E29 and E30 are an illusion as the fountain is clearly diverted, if the overhang fails or vent location moves, high fountains will resume. Its also very likely flank vents will show up again if pressure is this high.
The pause duration between episodes has increased markedly, from a few days in January to 2 weeks now. Some of this is because the vent is much higher now, so more pressure has to build up before it can erupt. This pressure comes from the gas, and the gas supply is probably fairly constant or slowly declining over time. The high fountaining since March is probably related to this higher gas pressure. I don’t have numbers for the volumes of each episodes but my guess is (speculation alert) that if you divide by the duration of the intervals (i.e. use volume per day), you’ll find the same reducing rate over time that is seen in most effusive eruptions around the world. They run out of gas. Should go electric.
The volumes are visible on the HVO website, in eruption intormation. There is variability, but the episodes can kind of be grouped into 3 phases. E1-E4 is phase 1, with deflation of the summit. E5-E24 where most episodes were between 1 and 5 million, E16 an outlier at 6 million. Since E24, every episode has been over 5 million, and bigger than its predecessor. Length has not increased though, only intensity. Up to E30, the combined volume is 145.1 million m3, and 349.1 million since December 2020. So 44% of the way full from 2018, but 20% is from the last 8 months…
I dont have the means to plot it out, but I know someone else here has done it, their name I cant remember right now. Their plot shows that the eruption is if anything getting stronger though, not weaker, even with growing interval length. Kilauea isnt a normal volcano and this isnt a normal eruption…
The name is J.O. (credit to person who made graph) and, according to the graph, the eruption is indeed intensifying (albeit with progressively longer pauses), with episode 30 being the most intense of them all so far…
https://www.volcanocafe.org/the-kamchatka-earthquake/#comment-167730
That is not the most relevant number.
The important number is not the eruption volume per episode. It is the volume per day, including the time the eruption pauses. That is the supply rate of the eruption. Over the past 8 months, it has hovered around 0.7 million cubic meters per day. The peak values (excluding the December start) were in March/April at 0.8, but May was low so the April spurt probably depleted the resources a bit. At the moment it hovers around 0.65 million cubic meters per day – declining a little but largely constant.
The eruption rate during the episodes indeed increased after April. It is now around 0.6 million cubic meter per hour. Interestingly, during an episode, during an hour it erupts about the amount of magma that accumulated over one day. That is why the duration of each episode in hours is similar to the preceding pause in days.
So the supply rate is fairly constant but the pressure needed to start off an episode is increasing and is now 2-3 times higher than it was in February-May. Based on this, the next episode could start as late as 21 or 22 Aug. (Famous last words)
Yes, thanks:)
Thats exactly what I said before Albert… Supply rate is pretty constant, but pressure increases as the vent gets more elevated. The pressure also temporarily slows the supply rate at the cost of pressure increasing further, and the tiltmeter isnt necessarily going to see this unless it has a secondary affect like a DI event, which has maybe just occurred. But this doesnt necessarily affect supply rate average, remember E30 also had an intrusion to tge flabk fissure for about 50 minutes before fountaining, and that might be upwards of 1 million m3 extra magma
High pressure isnt necessarily just gas pressure too, its also hydraulic pressure. Thats probably most of it actually, SO2 and H2O wont exsolve at the magma chamber depth, so pressure on that is probably pretty low until fountaining starts and runs away.
At some point pressure will be enough to erupt somewhere else. But its still not that close. Nowhere else outside the caldera rim shows uplift of intruding magma. I will make the bold claim that a lower ERZ eruption is more likely than this stopping by itself though. As before, Kilauea isnt a normal volcano, it isnt restricted by magma supply when erupting this way.
I think given past history, that the next eruption outside the caldera will go out the back side down low on the SWRZ. The map linked below references Kilauea lava flows that indicated an eruption before 1823 that everybody thinks of following the 1790 caldera collapse. It notes that the the age is unconstrained but was sometime between 1790 and 1815. This suggest that there were at least 2 eruptions before the big blowout eruption over on the ERZ in 1840. Other sites suggest the 1st eruption on the SWRZ was relatively long lived, but a much lower eruption rate than 1823. The 1919 Mauna Iki eruption adds more evidence to my argument following the long-lived summit eruption prior to the 1924 mini-collapse of the caldera.
There was 50 years between the 1790 collapse and 1840 eruption. If the early 19th century is a guide, we have at least another 15 years of golden showers (of tephra) before the caldera springs a leak and Kilauea soils her backside all over the lower SWRZ. Taking this prediction a further step, the next significant ERZ eruption will occur in 43 years.
Mark it down now! I predict there will be a Lower ERZ eruption in 2068!!!
Theres a lot of differences between now and back then though, which werent so obvious before. Maybe the most obvious is that BOTH rift zones erupted last year, they have already woken up, and unlike in 1790 the rift connectors seem to be largely undamaged by the 2018 collapse.
The eruptions southwest of the caldera after 1790 are complex. There is the 1823 flow, a rapid drainout associated with south flank slip and probably shift to the ERZ. But there was also the Kealaalea flows, that were lava in the caldera flowing down a crack and re-erupting far down the rift, a lot like in 1919, or the flank vents of Pu’u O’o. Slightly younger are the Kamakaia hills, which was a true SWRZ eruption fed from the connector, but erupting evolved magma and not basalt at least initially. It seems to have been a large intrusion as a small cone sits far down the rift near the 1823 flow and is inline with the other Kamakaia vents.
All of these probably happened in the later part of the 1790-1823 interval, as the oldest vents required a mostly filled caldera, and even at current rates of 0.2+ km3 a year, the 1790 caldera was probably about 5-8 km3 in volume, so would have taken 25-40 years to fill that much. Its a lot of rough numbers so not too specific, but its pretty likely all the SWRZ eruptions were after 1820, and represented failure of the summit system to hold all the magma, which eventually went east in 1823.
The current situation is comparable, but the floor might need to be higher to have the same force with the smaller volume. And while there is an open vent the lava level needs to be higher anyway. Question is if it needs to be so high the vent ends up higher than the rim and overfliws, its certainly not unlikely. Im not convinced the current eruption style is about to end, it is getting stronger if anything, wider gaps with no glow at all between now, and larger volume episodes with truely enormous heat emissions.
Thanks Tallis, just got around to reading this, very good write up and a lot to think about.
Here are some thoughts I have.
1. The “Crike” clearly seems to be a conduit with a significant volume of magma moving up from depth. But I don’t think it can be confirmed quite yet that there is no overlying or shallow magma chamber. Many, or perhaps even most volcanoes magma chambers aren’t as well defined as we would like to think. More a series of sills and dikes with crystal mush in between. And for what would be more easily defined as a proper chamber, there is very often a larger deep chamber that feeds a smaller shallow chamber. This is very well documented in volcanoes like Sakurajima, which has it’s own smaller shallow chamber that’s fed by the very large deep Aira Caldera magma chamber.
In this situation, whether there is a smaller shallow chamber remains to be seen, and even if there is one, it could be mostly crystallized for all we know. I wouldn’t be surprised if the various surface features in the region (chalpatan, potrerillos, chiles-cerro negro, etc) all were fed at one point or another from a deep chamber that has taken various different paths to the surface.
2. I don’t personally think we can get a truly accurate estimate of magma intrusion size based on surface deformation alone. This goes against popular thinking, even from a lot of volcanologists, but I think my thinking makes sense, and I’ll explain why. The popular view is that a magmatic intrusion can be measured by surface deformation, because the magma is clearly being added into the crust below, which means any increase in surface level is a direct relationship to the amount of magma added. But in my view, this is a lot more distorted because any “intrusion” that occurs is already occupying space within the crust. The popular view only works if the magma being added were exogenous. But that’s not the case – the magma rising to the surface is already part of the system below. Just as we would see with rising wax in a lava lamp, a magmatic intrusion wouldn’t simply push up surface ground, but could also push *down* any crust that it is intruding into, which is possible from any space this magma vacated.
This is not to say that intrusions would not cause surface deformation, but I don’t necessarily think the volume is a 1 for 1 relationship with the amount of surface deformation. Since a lot of the deformation could potentially just be a “swap” displacement (which explains the subsidence around the ring), I think that the intrusion volume is a little more difficult to pinpoint and may even be larger than what surface deformation is implying.
3. Estimating any size of an eruption in my opinion could be really tough here unless we could somehow measure the amount of existing magma already available in the system. And even then, it’s tough to really gauge because we would need to know how much is eruptible, and then there is the realization that non-eruptible magma can shift to an eruptible state fairly quickly if depressurization occurs (IE, a large enough eruption that can either reduce the overlying pressure enough, or cause a caldera collapse, destroying the overlying roof).
4. Do we know for sure that this is felsic magma?
The most recent products have been dacite and I don’t recall any recent intermediate products.
I suppose, that the only way to properly to ascertain the size of the intrusion, is through study which we don’t have. Same thing for eruptible magma. The uplift we is in spite of the surrounding subsidence.
Deep stuff is probably basalt or andesite, as a melt being fluid and hot. But there are no obviously basaltic or andesitic young vents anywhere in the area too so it suggests there is either very dense crust in the area (pretty plausible, its a mountain range) or a widespread area of inactive magma, which is also pretty plausible in such a setting. Both would result in felsic volcanoes.
I think maybe the volume of the swarm isnt so important exactly, by virtue of the thick crust any volcano would have a huge magma system in this setting, most of it is plutonic. The area isnt particularly prominent for massive calderas generally, but there are a lot of huge effusive felsic volcanoes, Cumbal is mostly lava flows. Not to say big explosive eruptions are absent but are rare, VEI 5 is probably the biggest a near future eruption is likely to be, followed by voluminous mostly effusive activity. A caldera collapse comparable to Hatepe is a pretty extreme suggestion really, that event was crazy even among caldera collapses, it was a supereruption in all but volume.
Volcanoes in neighbourhood are Galeras and Reventador. Galeras has Andesite and Basaltic Andesite, had two caldera collapses and the strato volcano grew in the caldera. Magma here is more intermediate than Chiles-Negro. Reventador probably did an eruption 1936 that was falsely noted as an eruption of Chiles-Negro. Reventador also does predominantly intermediate magmas, but can also do basaltic or dacite ones. This volcano also sits inside a caldera. It seems typical for volcanoes in this region to sit in calderas. This means that there must have been caldera forming eruptions … maybe like the one that CCN can do?
If CCN’s deep magma chamber doesn’t allow an actual caldera formation, it may still blow out the summit that blocks a possible eruption now. Maybe first there happens a felsic Plinian eruption, and afterwards a more gentle intermediate longterm eruption follows that builds a strato volcano again in the explosion crater.
They arent that close, way too far off for a common source above the mantle. Also again the andesite is really dacite with mafic crystals, andesite melt is much more liquid. Reventador is more of a true andesite lava but there are still crystals. None of the major stratovolcanoes in the equatorial Andes have any significant mafic lava component as I understand. They are like gigantic versions of Santiaguito or Bagana, mostly effusive but just really huge. Explosive vent clearing is likely but theres no good way to kniw how big that will be until it happens. Although, because it hasnt broken through yet at only a couple km deep I personally dont think it is about to blow up, eruption before 2030 is likely.
More the point was that a caldwra eruption would be very unlikely, and a large caldera eruption even more so. The area has dangerous volcanoes but they are dangerous more for the population density close by, than because any of them pose a risk of particularly big eruptions.
The size of the eruption, depends on 3 variables, how much magma is available, how much pressure is accumulated, and how effective the conduit is.
First, with the given chamber metrics and traditional volcanology rules(40% threshold). We find that the system has 560 km3 of melt at the minimum. If you want use the number I threw in using context given by the unrest th
en we will get a much bigger number. In any case magma is not an issue.
Pressure and the conduit is much harder to figure and needs proper study but given context by the unrest, it’s likely to be significant. Not only has the system, generated and sustained a single intrusion, something Taal and other volcanoes in recent memory couldn’t do. Involving a crack several km long and 2-3 km wide, yet this didn’t stop this volcano like it would others. This situation has no known analog and no known peer besides Ioto. I don’t think it’s right to assume that a minimal VEI 5 is the most likely scenario.
Correction:18 km long
Question for Tallis after his response to the above comment by Chad. Is Chiles really that different from the others in the Northern Andes? It seems like the Northern Andes aren’t known for explosive eruptions. Galeras and Quilotoa seem to be the only ones, and the biggest of the 2 was from Quilotoa and that was only 21 km3. Unquestionably a big eruption, but still far and away from something like a Haetepe-like event.
The only system that is remotely comparable is the Cerro Bravo-Cerro Machin complex but even then there are huge differences. The key takeaway is that the same chamber that feeds Chiles-Cerro Negro feeds Potrerillos and likely fed Chalpatan as well. It’s also possible 3-5 other major volcanic features are also related as well. This is WAY different then Galeras or Quiltoa, on a fundamental level. It’s not wise to look CCN in isolation. especially when we see the magma is coming 20 km away from the clone
I mean more that the actual evidence is that the eruptions here can be big but not usually explosive, 2 calderas in 2 million years, only one of which has direct evidence of connection, not high odds really. Theres also a lot of evidence that supervolcanoes are probably actually very hard to erupt explosively, which is how so much magma can accumulate, Yellowstone has hundreds if not over 1000 km3 of rhyolitic lava erupted since the last VEI 8. The calderas in New Zealand also have similar activity, and same for the Andean volcanism further south, probably VEI 8 progenitors but on the surface mostly effusive. Crystal rich magma also tends not to be typical of primed calderas or their progenitors as far as I understand. This seems to be a reawakening volcano not a primed monster. Maybe in 50-100 years of heating it will be very different.
And again, thick crust by default will lead to huge magma systems just to breach it, every volcano in the area probably has a complex pathway, just not one visible now for either being open or currently inactive. CCN is certainly a place to watch but nothing it is doing screams VEI 6+. Its a possibility but the system doesnt have a habit of doing it like was the case at Pinatubo.
Wait 50-100 years? For what? The system is already pressurized and molten. This intrusion is a testament to that. The exact size of the other systems is unknown but the structure, it’s safe to say though that the current situation isn’t and hasn’t been replicated recently.
I mean for the crystals to settle out, that would take decades. Its not a hard rule but calderas tend to be more crystal poor, stratovolcanoes more crystal rich, its a factor of magma rising immediately or collecting. CCN at least in existing state is a stratovolcano complex, crystal rich lava, seemingly mostly effusive by volume.
I dont think it will take more than 5 years for an eruption to start at this rate, and not unlikely next year, but it would probably take decades to heat up enough of the dormant magma to let it settle and potentially fuel a bigger caldera. I have doubts about that at all.
I would hazard to guess the existing intrusion is probably not felsic, it is probably hot magma (basalt and/or andesite) that may or may not be interacting with existing felsic magma in the shallow crust. That would explain the structure and locations of quakes, deep magma rising up. None of this really gives any indication of the eruption size potential, just shows there is ongoing supply of hot magma, making an eruption likely.
The only way to test this is if an eruption begins.
My base guess if an eruption were to occur would be something similar to Chaiten or Cordon Caulle, except Dacitic instead of rhyolitic. As for size, I think that’s a total crapshoot, and depends on a lot of variables that we simply don’t have any way to understand. That includes whether there were other periods of uplift and magma injections in the past before any monitoring occurred, how much magma already exists in any shallow magmatic systems, and then how much could be mobilized in an eruptive event.
I think once a main plug starts to break or a surface conduit forms, it could erupt very quickly with an explosive initial eruption followed by a long period of blocky lava flows or dome formation.
Also, if the magma chamber is indeed over 8km deep, that would make it very difficult for a proper caldera to form.
To address your caldera comment, based on what I read in the above comments, that caldera formation possibility was mostly, if not entirely settled as being very unlikely because of the deep magma chamber.
Also, after reading the rest of the comments here, it seems like Chiles isn’t likely to produce as big of an eruption as what’s been previously believed (i.e. Pinatubo, Haetepe, etc.). Based on what I’ve read (and by all means correct me if I’m wrong), but it seems that if Chiles produces an explosive eruption, it’d barely reach over VEI 5.
Nothing is really settled here to be honest. That’s probably true for almost any volcano – too many unknown variables for anyone to say they know something with complete certainty.
That being said, we do not know tif there is or isn’t a shallow magma chamber. In fact, I would personally believe that it DOES have a shallow magma chamber, which is where the “Crike” is pushing the intrusion into. The question is more likely that of how large a potential shallow chamber is, how much of the magma is just non-eruptible semi-molten rock, and how much could potentially be reactivated into an eruptible state.
As for other comments talking about volcanoes in the region. I wouldn’t say there is little history of caldera eruptions or explosive volcanism. Of the 4 nearest volcanoes to the north (Galeras, Cumbal, Azufral, Dona Juana) 3 of them do have existing calderas. These are not necessarily huge calderas, but they do exist, representing likely large vei 6/7 eruptions at some point in their history. Volcanoes in this region also tend to be dacitic, which there is some commonality between them. If you go to the south, the volcanoes also tend to have this similarity, and there are other small calderas at Chachimbiro, Cuicocha, and Imbabura. Then of course you have to factor in the two existing older calderas at the Chiles-Cerro negro complex.
From looking at the region, I would say that somewhat explosive dacitic volcanism is not uncommon. These volcanoes don’t tend to build huge strato-edifices since the Dacite is too brittle and viscous. But they do push out big domes, lava flows, and can produce some large collapses or small calderas. For dacitic volcanoes, I would assume almost all eruptions would start out explosively, but I would also guess most would then evolve into dome-building or blocky lava flow phases one the explosive buildup has cleared. You see this type of behavior in the Cascade range as well. Mt. St. Helens and 3 Sisters are good analogs.
Now, large caldera volcanism does not seem to be a thing in this region. Most of the calderas are small to mid-sized, or are large collapse scars. It would seem that these volcanoes do not tend to build large enough shallow magma chambers that would be needed to get a larger caldera size (over 8 km wide). Why? Could be that most of the main storage in these systems is much deeper, which would limit the size of any potential caldera formation. Or alternatively, it just may mean that the shallow storage tends to fail before enough magma accumulates to build any bigger magma chamber.
All that said, I think when CCN is ready to erupt, the final phases will make things very obvious what is about to happen, and it will probably occur fairly quickly. The closed-nature of this system and viscousness are probably why this has been such a noisy volcano so far, but I would expect it to get a lot noisier before any eruption. Clearly this would be easy to spot by looking at earthquake depth rising quickly in a narrow column above the region we are now calling the plug.
I largely agree with cbus here. The Northern Andes eruption history is relatively well studied, and I once collected information from various articles into an excel where I have the age of a number of young eruptions in the area (particularly explosive, which are more studied).
The volcanic arc is one of the most homogeneous in the world in terms of composition. It erupts mostly crystal-rich andesites and low-silica dacites. Caldera-forming ignimbrites are extremely rare, and I can’t recall anything younger than the eruption of Chalupas 200,000 years ago that formed a 20 km wide caldera. That said, erosion is very high due to the rainy environment and glacial activity so some calderas/ignimbrites may have been eroded away. For one, Puracé has a suspicious structure as big as 30 km long and 20 km wide.
While proper ignimbrites are rare, normal plinian eruptions that sometimes reach into low-end VEI-6 are common. Atacazo has two VEI 5 eruptions in the last 5000 years, Azufral has 3+ large plinian eruptions closely spaced around 4000 years ago, Cerro Bravo and Cerro Machin have a few Holocene plinians idk how big, Cuichoca formed a 3 km caldera 3500 years ago which may have well been a low-end VEI-6, Guagua Pichincha did VEI 4-5 eruptions 1000 and 2000 years ago, Pululahua went plinian four times and then erupted a 0.6 km3 lava dome between 2700 and 2200 years ago, and finally Quilotoa had a 20 km3 plinian eruption about 800 years ago. I’m probably leaving some events out.
Effusive eruptions are also common, with highly viscous lavas that form massive lava domes or coulees, and are sometimes associated to lateral blasts or massive vulcanian explosions. For example, Santa Isabel has three Holocene andesite/dacite lava flows that are aligned, maybe formed in the same fissure eruption, which have individual volumes of about 0.3, 1.1 and 2.1 km3 (fully effusive as far as I can tell). Chachimbiro is known to have erupted a large rhyodacite dome that produced a lateral blast 5000 years ago, and dome-building of Doña Juana in 1899 was associated to explosions with deadly pyroclastic density currents.
So a VEI 5-6 or a massive lava coulee, either of the two seem like possible scenarios to me (based on the background the area has).
Now that Kilauea has shown a significant deflation, one is forced to ask “did the magma go into inflating some new chambers under presssure?’ This is the 2nd time (at least) that we have seen a deflation near the peak of the eruption episode.
I am curious as to where that magma is going.
Kilauea is in full DI mode. Looks like 31 is going to be similar to 30…
I meant to add the SDH…
?fileTS=1755652962
The real answer is that no one really knows yet. But the most likely answer is that the magma doesnt go anywhere, just that either a piece of the roof falls in and sinks, or crystals settle out and flow down the conduit below. The result is pressure upwards is reduced so tilt falls.
The fact DI events have happened at the cusp of the next episode twice now suggests it might be the second option in these cases, pressure in the magma chamber cracking the roof and letting it fall in. Its temporary but this should probably be expected in future events.
Interesting thought, Chad, magma breaking the roof and releasing pressure. Would there be any siesmic signals too?
Is the deflation caused by a minor intrusion? Maybe it’ll precede new side vents as we had one during E30.
The tilt pattern gives a suggestion of what is going on
First, note it is a small deviation, so we are looking at a minor event compared to the rising pressure since the last ‘event’.
Second, it is quite sudden. This is not related to cooling or crystals which act much slower. It is an escape hatch opening somewhere allowing some venting.
Third, it is repeatable. It is near-identical to what happened at the same phase in the process during the previous event: at the same tilt and showing a similar shape. The timing suggests it is not far below the surface, perhaps 100 meter (the number is purely speculative)
Fourth, during the previous event there was a full recovery after the deflation ended to a level that would have been reached without the deflation. So ‘venting’ is the wrong word: the pressure temporarily reduces but it suddenly comes back.
And finally, it started with a minor earthquake mid-July, followed by a larger (but still minor) one late July. Something broke.
My feeling is that it is an escape into the lava lake in the crater, below the solid surface. Once the pressure in the conduit rises enough, the underground path into the lava lake is forced open allowing some magma to take this route – but not reach the visible surface. But after a while, the magma level in the conduit is higher and now it is beginning to break through the lid on the main path. Magma and gas always take the easier route, so it flows past the escape route, the flow reduces pressure in that direction (Bernoulli) and the lava lake pushes back. The real surface eruption follows.
This is speculative, of course!
The DI events being crystals flowing down the base of the magma chamber into the conduit and temporarily obstructing the supply is suggested by HVO themselves if I remember. The roof collapse model I think was from Hector.
The only bit that needs to be fast is the obstruction of the supply into the magma chamber, not how fast that obstruction itself moves. The obstruction means no new pressure is added, and magma might drain down following the obstruction, so tilt falls. When the obstruction is gone or magma supply pushes through it rushes in and rapidly recovers. DI events tend to basically equal to a net 0 deviation on the existing trend. There havent been many recently but they were very common up to a couple years ago. The very steep inflation between episodes now mutes the D part of the DI, so it is a gradual decrease.
My prediction, in the next 3 days at some point the tilt will rapidly shoot up and recover the deflation, maybe over 5 microradians in a day, and E31 will start while it does that.
That crystals can plug up the conduit is not under dispute. But they don’t fit this event, as explained. Same for roof collapse: neither would predict the repeatability. As for your recovery prediction, that is indeed likely what will happen, as it did last time. But it doesn’t distinguish between the models.
I did provide a logic behind why this might be exactly caused by growing pressure though, in another comment. Crystal settling isnt going to do this with rapid response but high supply would logically bring up more crystals that eventually find a new higher equilibrium and that might be what we see.
But given it has happened now at exactly the same time and elevation point on the tiltmeter, it suggests the pressure is the cause. Using the other option that some DI events are from flaking off of the walls and roof, that would be a volume change, but also the cracking itself would be kind of a mini intrusion and remove pressure a bit too. SDH records mostly the massive deep primary magma chamber, its not a shallow process. There are quakes at the summit on occasion but notably just before this DI was one that actually showed on the main map, might be important.
SDH and UWEV may be more sensitive to shallow changes than you think. The locations are some 2 km from the vent. 20 micro radians of change in tilt corresponds to a change of 4 cm over that distance. There is’t much GPS data available at the moment, but OUTL is working. It shows a northward/southward movement during the tilt excursions of a couple of cm, so there is an effect on the local ground of that rough amount even at that distance.
We don’t know the actual movements inside the caldera. But if we assume upward movement there of 1 meter, then the tilt change would suggest a depth of the magma/gas of a few hundred meters and if it is 10 meters, that becomes ~ 50 meters. Very rough numbers!
OUTL shows that the region has been deflating since the start of the events, which shows that the erupted magma is not being fully replaced at depth.
The magma chamber has deflated because there is an open vent, so pressure was relieved. Not because the eruption is faster than supply. The GPS at UWEV shows shows slight uplift despite that being uphill and on the side of Mauna Loa. The interferograms also show deformation is only within the caldera fault, no sources outside are active at present.
Same as the ERZ, its not magma backflow, just lack of inflow.
The data I see from UWEV has not been updated for a year. BYL, on the other hand side of the caldera, shows no upward change this year.
https://geodesy.unr.edu/NGLStationPages/gpsnetmap/GPSNetMapMovable.html?latz=19.393&lonz=-155.110&zoomlev=11
This shos it, though not as nice.
Thanks, I did not have that link. As you say, more complete but harder to read. I think the red trend line should be ignored for the ‘up’ component: blue gives the data. Here, UWEV shows no obvious upward change over the past 6 months, at last one strong enough to recognised. I haven’t looked at the actual numbers though
Red is real but not fully checked yet, i think. Back last year I saw it switch and data stayed the same. I dont know why the HVO site doesnt show the data of most stations past then, it predates the budget cuts and hasnt really been clarified, it seems to be only a site connection rather than a fault with the system.
UWEV looked very slightly upward trending to me, not really inflating, but its different to OUTL. Both are just outside the caldera. I think the default vertical movement of all the stations in that part of the island is down, if magma isnt being fed in then the south flank slip will slowly distribute the load. When magma actually does move either in or out the change is very much faster, as seen on the ERZ a year ago with the sudden spike up from probably only 0.1 km3 of magma or less, basically reversing over a year of the last t years trend of slow sinking. Slight inflation probably is real though, as in magma supply caused, but deflation is background unless obviously caused by something.
Thats ny interpretation anyway. It looks like tge DI has bottomed out, some point soon the supply will catch up and surge, probably starting the episode. If the average eruption rate does decrease from this there is probably magma building up in the primary chamber, which could be consequential later. Slow magma seeping past Pu’u O’o after 2011 was what eventually caused the 2018 eruption,or at least allowed it to be in the lower ERZ instead of high up like prior conduit failures of Pu’u O’o.
New insar, no real deformation outside the caldera since the 14th. Mostly noise from the tephra blowing around and lava lake, so not very useful…
http://www.mounts-project.com/static/data_mounts/kilauea10/2025/kilauea10_20250814T161632_20250820T161535_VV_ifg.png
Volcano Watch mentions sinkholes in the tephra deposits on the south rim of the caldera that “opened as a result of the new fissure in the south wall of the caldera”
We see a sharp inflation rate now at 1:36 pm Hawaii time, Thu Aug 21st 2025 https://www.usgs.gov/volcanoes/kilauea/science/monitoring-data-kilauea
HVO volcano watch about the south fissure of E30, right on time
https://www.usgs.gov/observatories/hvo/news/volcano-watch-fringes-and-fractures-episode-30
Its unclear if it was from the existing vents, or a new dike up, but it wasnt big. However the dike was longer than the fissure, cracks opened above the cliff and formed sinkholes in the tephra. Seems it was at least bigger than the insar made it look and followed existing cracks. At some point lava might have enough pressure to erupt on the rim this way.
Looks like episode 31 is kicking off.
Not quite there yet but beginning to build up. Some spatter and strong hydrogen flames now
Fascinating! You can see the spike on the SDH tilt, just before the onset of deflation. Nothing like that in episode 31 so far – which admittedly hasn’t started yet
Using Fibonacci analysis on the 1 year EWEV GPS, I believe if a strong eruptive episode drains the summit much below -0.025 meters on the graph, this eruption will officially stall for 3 or 4 months as the pressure recovers. As the pressure continues to build with stronger episodes and longer pauses, I believe the SWRZ will finally crack the closer the EWEV gets to +0.125 meters on the graph. At this point, I can see the possibility of an SWRZ eruption starting in December and rising the longer this pattern continues.
?fileTS=1755628418
Just curious: Are you the first person to apply Fibonacci analysis to volcanology?
The usual little article about Hekla:
https://www.ruv.is/english/2025-08-20-hekla-could-erupt-at-any-time-451363
Current inflow seems to be around 0.5 m³/s. Also funny how auto translate did a better job for me than the english homepage version…
0.5 m3/s for 25 years is a bit under 0.4 km3, and 0.2 km3 at the low end. But its broken the cycle of the 20th century so theres no real way to say its close or not. In 2100 though it could be over 1.5 km3 if it keeps up, so its reasonable to expect the next eruption to be on the large side.
RÚV has another article today:
Tension building in Brennisteinsfjöll; major earthquake expected eventually (19 Aug)
South of Reykjavík and a bit to the east of the Svartsengi fires. It would make sense that the strain from the latter could set things up for a big tremor. There’s also the water injection wells for the geothermal power station nearby, which would add to the mix. It’d be interesting if a quake occurred then set Hekla off.
You also have the SISZ quakes in 2000 and 2008 that showed a westward progression and only released about half the pent up strain since the last sequence of large quakes. They might have transferred a bit of strain towards Brennisteinsfjöll.
Now, SISZ and Reykjanes are separated by the Hengill triple point, so it’s not obvious in what way strain is propagated between the regions. There have also been quite a bit of activity lately in the eastern part of the SISZ, around longitude 20W, right next to Hekla. If we get a quake in that region, Hekla might be affected, while a quake at Brennisteinsfjöll probably will not have much influence on Hekla.
As far as I’ve understood it correctly, Hekla is predominantly independent from SISZ, although it sits in it. SISZ’s earthquakes can occur around Hekla, but aren’t Hekla’s “private earthquakes” that would happen initially at greater depths and move quickly towards the surface in case of an eruption.
A bit mystery are fissure swarm eruptions of Hekla outside the central volcano. The last happened 1200 years ago, short before the Vikings came. Can the SISZ cause some kind of rifting in the EVZ part that drives Hekla? Hekla, Tindafjallajökull and the Myrdalsjökull volcanoes are in the region where SISZ and EVZ meet each other.
Langjökull is preparing a Jökulhlaup, the strongest since 2020, maybe it’s gonna be bigger than 2020: https://www.vedur.is/um-vi/frettir/jokulhlaup-hafid-ur-hafrafellsloni-austan-langjokuls
Probably predominantly caused by climate warming…
Not to be a slowdown here, but appearently the USGS site for Kīlauea is going really slow. I’m using a phone but in spite of very strong connection, it takes minutes to load. It’s a little faster on the computer, but still takes minutes. Anyone else experiencing this? Is there anything going on?
Nvmd, just a few things I had to do to make it quick.
Same here. This morning it was working fine.
View of the pond on Monday at Kīlauea.
Is it possible that under/near CCN there is shallow magma chamber except it’s all mush..and this mush is undergoing rapid rejuvenation at depth and causing a change in pressure enough to support a Crike? So, some pressure/expansion in the shallow chamber zone during each seismic crisis could be thermally driven from mush/crystal-melting and not necessarily only from a change in magma inflow from below.
It’s possible that the Crike is the crustal zone directly above where mush is melting the fastest thus the pressure the greatest?
https://link.springer.com/chapter/10.1007/11157_2017_21
https://insu.hal.science/insu-00576480/document
Gotta read this a couple of times and let it soak…
Tokara Islands earthquake swarm continues.
Tremors are currently located to the northeast and southwest of Akuseki-jima, but not immediately near the island itself. Interesting distribution.
This might help – https://www.jma.go.jp/bosai/map.html#7/29.636/131.204/&elem=hypo&contents=earthquake_map&lang=en
A July 2nd article https://www.japantimes.co.jp/news/2025/07/02/japan/society/japan-earthquakes-tokara/
Etna does what Kilauea was supposed to do: https://www.youtube.com/watch?v=Px4atfg6GF8
I do believe I see a smidge of ropy pahoehoe here and there…
Clearly ALOT more crystal rich than the Kilaueas smooth shiney lava that looks like liquid aluminium close to the vents. Here at Etna its almost becomes Aa lava instantly as soon as it leaves the vent despite being fluid the lavas flows crust starts to gets torn into clumps “rough pahoehoe” or “spiney pahoehoe” very quickly almost as soon as it leaves the vent it haves a “Aa” like surface its crystal rich. This because its so gentle and spattering can be classifyed as a “hawaiian style eruption” for Etna but the lava is clearly not as fluid as Kilauea or Nyiragongo of course even if its still fluid enough to form lava tubes and rough pahoehoe flows
The video also shows, how gasrich the magma/lava of Etna is. The lava doesn’t flow out quietly, but does many gas-puffs. The temperature of Etna’s lava is a bit lower than Hawaii or other plume volcanoes. It’s usually in the range of 1000-1200°C. This contributes to the higher viscosity of the lava.
Another strong earthquake in the Drake Passage. There was at least one other earlier this year.
https://www.volcanodiscovery.com/earthquakes/22087250/2025-08-22/02h16/magnitude7-Drake-Passage.html
I like how there’s a felt report from Antarctica.
I don’t know the tectonic environment well, is it related the subduction zone off Chile?
Just wanted to add a note for for future article authors (and for anyone else who had to think a while to work it out) can you include the full text the first time you use an acronym – in this case VLP I’m pretty sure stands for Very Long Period – I knew what it meant (magma on the move) without working out the acronym but I couldn’t find the words for a little too long.
Otherwise great article
Ongoing Spattering in Kilauea’s northern crater. Is it already escalating?
https://www.youtube.com/watch?v=sTXKIz2NGvg
It is heading for the escalation but slowly.
Looks to be trying its best, but eventually it fails to a mere gas-piston cycle. That would mean it’s gas rich, based on how intense the strombolian eruptions go. It’ll eventually break the cycle for sure, but it’ll take some time.
On that same note, saw that the V1 camera livestream might be view-botted…
(About 34-45k more than the typical 100-400 views during less exciting events to 1-3 k during episodes…)
yea viewbotting also happened on afar last time
The spattering vent has filled in the crater of the north vent, it seems to be draining into another hole instead of properly overflowing. But deflation hasnt started yet.
With this development though if it doesnt make new fissures there could be a return to high fountains again now that a narrow vent has reformed.
Yeah I came here to say the vent geometry looks like it has returned to a configuration that might produce high fountains…
Vent has begun overflowing at just before noon HST, I’d say we’re about to get underway here!
I’d say episode 31 is well underway now. Looks like alot of the degassed magma has been pushed out.
Yes about 20 minutes ago.
Mac
She’s spoutin’ all over the place now.
It’d be interesting if it fountained 1000-1250 ft. *horizontally*!
Its looking very close to doing that now, actually…
I wpuldnt be surprised if the overhang blows out or erodes away this time, or in E32. While still open its obviously unstable to have a diversion like this long term.
These fountains being never vertically tall also mean basically the entire output stays hot landing and the whole crater floor is flooded completely. The MIROVA heat emission reading is going to be enormous.
Use V3 cam or V2 cam, V1 has too many people on it, having issues.
Mac
Looks like it’s trying to go sideways again…
They have moved the V2 & 3 as if they are expecting the S fissure to reactivate??
HVO has said if the fountain was vertical it would be well over 150 meters tall, so not quite the height of the episodes in July but its very premature to say high fountaining is over. At some point the fallout will accumulate enough in the north side of the crater that a cone will grow there too, though that might take a long time if the intervals are multiple weeks as standard. At the same time tbe south vent is still active if not participating, the overhang diverting the fountain could also fall in.
Alternatively the fallout area becomes a lava lake and evolves into a shield with no change in eruption style
There was a significant shift from tall lava & tephra fountains towards moderate high pure lava fountains. Although the height of the lava fountains has decreased, the volume of lava and the amount of deflation has grown (from -15 to nearly -25 microrads) on UWD station.
Looks like two or maybe even three active south vents now 8pm HST
I think it might actually be a new fissure in between the north and south vents, although “independent” from either…
Its in the same spot as the south vent but does seem to be a new crack, and goes up to the glowing patch that has been there over a month now but not erupted.
I wouldnt be surprised if the cone splits and merges the vents with how powerful the north vent is yet being diverted sideways. Even as I have typed this the south vent has got a bit stronger… At some point whatever is diverting the north fountain has to fail.
If you watch the early degassed fountaining from the V1 cam you can see the conduit is located under the southern rim of the cone’s crater so it’s being deflected into Halemaumau by the rock above it.
You’d think with that amount of heat and pressure it would have failed by now.
I believe the higher secondary fountain is not actually the South vent at all – but the crack that has been glowing at night above and behind the north vent.
The true south vent is far left as viewed on the V3.
If thats the case the rim diverting probably hasnt got long left…
Hypnotic
MIROVA thermal emission for Kilauea, comparing to Nyiragongo and Nyamuragira. 14+ GW and its probably been around this now for about 11 hours, and likely a couple more left. So potentially around almost 200 GWh of thermal energy, which is enough to heat every house in the state 10x over.
https://i.imgur.com/UJqZ0MV.jpeg
That was a very abrupt shutdown of eruption on Kilauea.
Yes! Next episode: 7 Sept
Pele’s inland delta in Kilauea’s Caldera, like Okavango with lava:
Kilauea’s episodes head towards -15 microrads as minimum value during the effusive phase and towards +10 microrads as maximum value before the effusion begins:
New post is up! To be or not to be: the google volcano battle
https://www.volcanocafe.org/the-censured-volcano/