Puelche and Calabozos. Calderas of the Maule

Last article we saw a newborn felsic system, Laguna del Maule, whose so far brief history has begun rather impressively, with possibly two VEI-6 eruptions, a number of smaller plinian eruptions, and some sizable lava flows. In the article before that one, we saw Domuyo, a monster volcanic system that is very lazy. It could throw a cataclysmic eruption, but never does anything. Now we will meet two monster felsic systems with a more impressive history behind. Puelche and Calabozos. We are in the Chilean Andes, 200 km south of Santiago de Chile. And this is a land of hidden volcanic beasts.


Puelche volcano

Puelche is a volcano of which very, very little is known. Not much can be found in scientific literature. The article with the most information on this system is a 1999 article led by Wes Hildreth, who has also been one of the main researchers into Calabozos and Laguna del Maule systems. This 1999 article describes Puelche as the largest cluster of Quaternary rhyolite lava flows known in the Andean South Volcanic Zone.

The area of Puelche is covered in snow much of the year, and is barren, uninhabited, and of very difficult access. Some of the flows are showered in white pumice from the 1932 eruption of Quizapú. Erosion is extensive. Deep gorges cut into the volcanic field. The rhyolite lava flows now form the flat tops of mountains, perched up to a kilometer above fast-flowing rivers, over precipitous slopes.

The composition of the field is bimodal, with most flows being rhyolites, and some flows being basaltic andesites. The rhyolite flows have 72.1-75.5 wt% SiO2, and the basaltic-andesites 54.5-54.9 wt% SiO2. About 21 km3 of rhyolite lava have survived erosion, a comparable volume of pyroclastic material is likely to have been erupted. The flows form a ring of about 15 by 11 km, in a similar arrangement to Laguna del Maule. Two of the flows have Ar/Ar ages of 213 and 354 ka (thousands of years ago).

This earlier information is what is known. However, after inspecting this area in detail, I’m convinced that this volcano has an earlier, much grander eruptive history. I have a map in Google Earth with dike intrusions from this section of the Andes. As part of this map, I studied a remarkable cluster of dikes in the north and south walls of the Puelche River. From the first moment, the dikes stood out for their length and thickness, which, after being used to the few short dikes of stratovolcanoes, struck me as being very prominent. One dike has about 15 meters thick, and at places up to 60 meters thick or more, and can be followed continuously for 6 kilometres, while discontinuously I think extends over 12 kilometres. The dikes are not simple-shaped; they jumped between parallel segments, and sometimes thinner secondary dikes seem to branch off obliquely from the main intrusion. They have variable thickness. Some are a meter wide or less; I can’t see them if they are thinner than this. Other dikes are up to tens of meters thick. Even the same dikes vary a lot along their length. All are very dark-coloured, probably basaltic-andesites in composition. And the swarm is very dense in places.  Some brighter coloured portions of the valley walls, where the dikes have strongest contrast and can be pinpointed most easily, show dense intrusive complexes, where you wouldn’t be able to walk any more than 30 meters without running into a dike.

It didn’t take long to realize the dikes seemed to extend from the Puelche Volcanic Field, located immediately to the ENE of the swarm. Presumably the dike swarm would have been the sub-volcanic complex to a field of basaltic andesite cones and lava flows above, in the shape of a plateau, that has been all but removed by erosion. Plateau volcanism has also happened in the area of Laguna del Maule and Calabozos, and judging from the dike swarm in Puelche too, so that it seems that felsic systems in this area are actually bimodal, consisting of plateaus of basaltic andesite lava which develop into calderas or ring-like fields of rhyolite flows, generally located in slightly back-arc positions.

Another clue to the history of Puelche is to be found across the border in Argentina. There, a thick, extensive stack of white-coloured ignimbrites, pyroclastic flow deposits, dominates the landscape. I had long been aware of those ignimbrites, but never bothered to look much into them. I had previously assumed they were probably ignimbrites of Calabozos. But then, upon mapping the ignimbrites of Calabozos, I realized something; they were not from this system. This would have been obvious rapidly. The white ignimbrites are very different in aspect from those of the above-mentioned caldera. Their surface is a lot more worn down and more brightly coloured, but it is their location away from Calabozos that makes it impossible for that to be their origin. The article by Wes Hildreth mentions these ignimbrites as a 650-m thick stack of several welded ignimbrites east of Invernada River. The youngest of them is dated at 1.06 Ma (millions of years ago), too old to be from Calabozos. But there is no more information. The ignimbrites are thickest and most extensive immediately east of Puelche, where they form a broad platform, hundreds of meters thick, that extends 35 km downslope. The distribution, I think, implies an origin in the exact location of the Puelche Volcanic Field. These layers would have been completely removed on the Chilean side by deep glacial erosion that has also taken away the basaltic andesite lavas there, but in the Argentinian side they were better preserved.

My first motivation after learning about the ignimbrites was to find if there were any signs of a caldera structure in the topography, and I think there are. Valleys in the Puelche area follow a circular structure that runs around the ring of Late Pleistocene rhyolite flows. I think this elliptical structure might be the former caldera. If this is correct, it would have measured 17 by 14 km across. Perhaps the floor inflated upward, which added to later lava flows, largely concealing the caldera.

Map showing the circular structure that might be a caldera. The ignimbrites make sloping platforms to the east. Topography map from https://maps-for-free.com/.

Puelche would have likely sourced multiple caldera-forming eruptions, possibly VEI-7 events, where pyroclastic flows travelled to distances of over 35 km and covered the landscape in a cumulative thickness of up to 650 meters of pyroclastic rocks. Do we know the age and composition of these massive eruptions? Maybe we do. There are two ignimbrites near Laguna del Maule, the Cajón de Bobadilla and Laguna Sin Puerto ignimbrites, which are dated at 1.5 Ma and 0.99 Ma respectively. I think these ignimbrites could have come from Puelche, since they are found within 20 km of the circular Puelche structure. Because Bobadilla is very thick, it was thought to be a caldera-filling ignimbrite, but I think it could simply be a valley-filling ignimbrite which might be correlative with the one dated at 1.06 Ma east of Invernada River. Both the 0.99 Ma and 1.5 Ma pyroclastic deposits are of rhyodacite to dacite compositions. The earlier one carries fragments of rhyolite lava flows which must have been blown away by the explosive eruptions, while the latter has pieces of andesite.


The Calabozos Complex

The Calabozos Complex lies 40 kilometres north of Puelche, and includes several central volcanoes, a plateau of basaltic-andesite lavas, a well-hidden caldera, a few massive dacitic lava flows, and 1000 square kilometres of a snow-white pumice desert, easily identifiable in Google Earth as a massive white patch from a thousand kilometres eye altitude or more, legacy of the plinian eruption in 1932 that fell just short of a VEI-6. Like other volcanoes of the area, it’s remote, although not as much as Puelche, and it’s little studied and heavily eroded, but also barren and arid which highlights young volcanic features in incredible beauty and detail.

Andesite and basaltic andesite lavas are dated at 2.02, 1.09, 1.08, 0.55 and 0.35 Ma, which makes a lava plateau on top of which have been constructed the later stratovolcanoes. Hildreth mapped three caldera-forming ignimbrite sheets, dated at 0.8, 0.3, and 0.15 Ma. The older ignimbrite only outcrops in a few places and its volume is not well known. The younger ignimbrites, of 0.3 and 0.15 Ma, are estimated at 250-300 km3 and 175-250 km3, respectively. Both ignimbrites are densely welded for the most part, which happens when the searing heat within a pyroclastic flow welds the pyroclasts together. They lack basal plinian deposits. All three ignimbrites have rhyodacite to dacite composition.

I think the reason we know there is a caldera is that there are ignimbrites, because the caldera itself is nowhere to be seen. I’ve looked many times at topographic maps, geologic maps, and Google Earth, and I still can’t picture the exact caldera rim. The ignimbrites point to the area that must have been in the caldera, but topographically the caldera has practically no footprint, with the floor being as high as the surrounding landscape. Two volcanic edifices, Descabezado Chico and Cerro del Medio, are built within the former caldera area. They have probably contributed to hiding the caldera structure. But the main reason the depression is gone is probably a combination of resurgence and glacial erosion.

The northern half of the presumed caldera area doesn’t have any lavas from Descabezado Chico and Cerro del Medio, and instead is floored by the ignimbrites, so the reason that this portion of the floor is as high as the encircling area must be that it has been pushed up from below. Resurgence is common in calderas, after all. A dacite lava flow, possibly from Cerro del Medio, follows the eastern moat around of the resurgent dome, is undeformed, and has been dated at 0.096 Ma. The authors of in the Hildreth article consider that this flow must postdate the main resurgence of the caldera. Main resurgence would have happened within 50,000 years of the last caldera collapse. If the lava flow marks roughly the elevation of the former caldera floor, then the resurgent dome is about 650 meters tall, and given that the dome covers very roughly 100 km2, 16 by 10 kilometres in size, then its volume is 65 km3. My guess is that a volume of 65 km3 flowed into the shallow magma chamber of Calabozos since the last caldera-forming event, but mostly in less than 50,000 years following this eruption. This is larger than the volume of magma emitted during the Tambora eruption in 1815. But it is also less than half the volume of uplift at Domuyo.

A system of ground cracks extends to the NNE of Descabezado Chico. They seem to be the surface scarps of normal faults, with a small downthrow. Most scarps face eastwards. If you have two spots in a balloon and you inflate the balloon, the distance in between the spots will increase, and if you are dealing with an inflating caldera, the same will happen, with the rocky surface getting stretched apart and eventually breaking up, making a graben system along the top of the resurgent dome, typically elongated since most resurgent domes are elongated. This is an axial graben system. So I think the fault scarps were formed during inflation of the Calabozos caldera floor. But we already knew about that inflation, right? The cool thing is that these are scarps are very low but very sharp and sometimes on soft ground, for example there are two low scarps that cut across a debris covered talus but are very well preserved. It is impossible for such scarps to have survived 100,000 years, given that even a small amount of erosion would erase them, and thus I think they are almost certainly postglacial, from the last 15,000 years. That means there have probably been major episodes of caldera inflation during postglacial times and that the magma body is very much alive and kicking.

There are other two other signs of the sleeping monster here. First, there are numerous hot springs which issue along the bottom of the valley, on the east and north sides of the resurgent dome, making an arc along the edges of the presumed caldera structure. Water in the hottest hot springs happens from the northern end of the caldera, which is where the valley cuts down the deepest and probably closest to the magma chamber. These hot springs are known as the Baños de Llolli, and water issues at boiling temperatures in them. There are also reportedly some sinter deposits in this same location, although it is nothing too spectacular as far as I can tell. The second piece of information comes from a mega-thrust M 8.8 earthquake that occurred in 2010.

Following a 2010 M 8.8 mega-thrust earthquake, several volcanoes in the area affected by the earthquake slip experienced deflation. The volcanoes that deflated were Nevados de Chillán, Descabezado Grande, Calabozos, Planchón Peteroa, Tinguiririca, and Cerro Overo (Caldera del Atuel). The volcano that deflated by far the most was Calabozos. Calabozos deflated up to 15 cm over an area 30 by 15 kilometres across, centered on the system of normal faults, and spanning Descabezado Chico and Cerro del Medio. Only Cerro Overo which is possibly a caldera volcano, although probably of smaller caldera size than stated in GVP, deflated by a similar amount of 13 cm, but over a much smaller area of 20 by 12 km. Other volcanoes deflated by smaller amounts and over smaller regions than Calabozos. Watching volcanoes deflate after mega-thrust earthquakes might be a useful way of finding about the location and size of magma chambers. Unfortunately, the study that reported this did not extend its interferograms into the area of Laguna del Maule and Domuyo.

One wonders why the volcanoes deflated. The authors proposed that the release of hydrothermal fluids during the earthquake was the most likely cause for the subsidence. But I would like to propose another option. At Calabozos the deflation area matched very well the resurgent dome, and not the hot spring areas, so I think it could have been magma-driven deflation. The earthquake will have relieved compression built up over decades. It may have thus created extension and opened space for magma underground, causing magma to drain from upper magma chambers into deeper levels of the lithosphere, the volcanoes with shallow storage may have seen some of their magma be removed to fill up space underground, the larger the amount drained, the bigger their magma storage.

Appart from the Calabozos caldera, there are five central volcanoes and numerous scoria cones spread out over the whole area. The central volcanoes are Cerro del Medio, Descabezado Chico, Descabezado Grande, Cerro Azul, and Manantial Pelado.

A map of the Calabzos complex. I place the somewhat shield-like volcanoes, made of lava flows and with multiple scoria cones, in red. I put the stratovolcanoes, or likely eroded stratovolcano remnants, in orange. I marked the historical craters of the Quizapú eruption cycle in yellow. While the postglacial monogenetic vents of bimodal dacite and basaltic-andesite compositions are in green, and seem to form a north-south belt along the western end of the complex.

Cerro del Medio and Descabezado Chico are shield-like edifices made up of lava flows, and with multiple vents, with each volcano having at least 4-5 different scoria cones and or craters, and both are located within the Calabozos Caldera. Cerro del Medio is very eroded and looks relatively old.

Descabezado Chico is younger than Cerro del Medio. It last erupted during the Holocene, producing the Escorias lava flow that we will see later. Apart from the Escoria’s lava flow, there are two scoria cones that are well preserved and are probably from the latest Pleistocene. One of them forms the summit of the shield structure. From this small scoria cone at the summit extend thin lava flows, I’m guessing of dacite composition (because they look a lot like the Escorias flow), and have preserved channels, wrinkles, and other features. Older, thicker flows underlie the younger, thinner flows. Descabezado Chico is located along the axial zone of the resurgent dome. It seems as if the resurgent dome itself is a massive tongue that extends north from the shield. I would guess that Descabezado Chico is the central vent of the very Calabozos system, and is possibly above the feeder of the shallow magma chamber. Chico means small in Spanish, and descabezado means headless. It’s the small headless volcano, according to its name. But I’d say its not so chico, but rather a monster system, you just need to know where to look.

We have the two shields, and then we have the two stratovolcanoes: Descabezado Grande and Cerro Azul, which are located ~18 km southwest of Descabezado Chico. The stratovolcanoes form steep cones that tower 1-1.5 km above the surrounding lava plateau, made of layers of red and black scoriaceous material interbedded with lava flows. Given the dark, mafic-looking scoria that makes up these two volcanoes, I think they are probably mostly basaltic-andesite in composition, but this is a guess of mine since there is not much data available. Cerro Azul is truncated by a crater 800 meters wide at its top. Descabezado Grande is truncated by an even larger crater 1-1.5 km across. The name Descabezado Grande alludes to its shape. Descabezado means headless, and grande is big. It’s the big headless volcano. The size of these craters indicates powerful vulcanian explosions, and plinian or subplinian activity in their past. Both cones are weathered, and apart from their historical activity and one small dacite flow (Alto de las Mulas), I don’t think they have had any earlier postglacial eruptions.

The fifth central structure, Manantial Pelado, forms a cluster of highly weathered scoria cones, and possibly deeply eroded stratovolcano remnants. The area between Manantial Pelado and Descabezado Grande includes some additional eroded remnants of possible stratovolcanoes.

The postglacial eruptions

Appart from the 5 central edifices, there are a number of scattered monogenetic vents. Most of these vents are old and eroded, but several vents are postglacial, from the last ~15,000 years. Five of the monogenetic vents carry basaltic andesite composition. Among these, are 3 postglacial explosion craters, known as La Resolana, ranging from 500 meters to 1 kilometres in diameter and located to the W of Cerro Azul. Two of the explosion craters are very deep and expose more than 100 meters thickness of bedrock in their walls, which was excavated by the explosions, while the third crater filled up and overflowed with lava. One of the two deep craters is partly filled with water, and the other one is right next to a lake, so the phreatic layer seems shallow and there may have been magma-water interaction. To the SW of Cerro Azul are a pair of postglacial scoria cones, crowned with craters of 150 and 350 meters diameter, and with thick aprons of lava, made of many overlapping 10-15 meters-thick tongues of lava with narrow channels, which were formed during long-lasting lava flow eruptions. The area around the explosion craters and scoria cones, for a length of 15 kilometres or more, is partly covered in a dark mantle of scoria. This shows that some of the postglacial basaltic andesite eruptions were very explosive, probably subplinian.

The Calabozos complex seems somewhat bimodal, with postglacial eruptions of basaltic-andesite and dacite, but with no andesite-only flows. Massive dacite lava flows have issued from vents all over the plateau, making four major lava flows: Escorias, Alto de las Mulas, Lengua de Vulcano (Mondaca), and Quizapú. Only the age of the Quizapú flow is known, which is historical. All four seem young enough to be postglacial, however.

Alto de las Mulas is very modest. Compared to its huge brothers, it is almost insignificant, with 0.1 km3 volume. It erupted from a short, 1-kilometre-long fissure, from the flank of Descabezado Grande. The flow is a simple sheet in shape, some 20 meters thick along the front, and possibly around 10 meters thick along the edges close to the fissure, so it’s among the thinner flows. I don’t see any signs of accompanying explosive activity, although the entire area is thickly covered in pumice from the 1932 eruption, so it’s hard to tell.

Another flow, Escorias, erupted from the heart of Calabozos, from Descabezado Chico volcano. A 1.8-km-long fissure opened cross the shield volcano of Descabezado Chico and erupted from both sides of the mountain, each side sourcing lava flows. The flows are thin. From the west side of the mountain erupted a flow that only reaches 10 meters thickness at the very tip of the flow, 5-km downstream from the vent, and is thinner than 10 meters elsewhere along the edges. A larger flow erupted from the eastern side of the descabezado, which is 20-30 meters thick along the edges, as far as 16 km downstream. GVP estimates the Escoria lava flow at a volume of 5 km3, which I think is too much. The flow is not thick enough, and the volume is more likely to be around 2 km3. This shows, I think, exceptionally high effusion rates that produced thin flows with scarce crust thickness. The lava flows are dacites with 64 wt% SiO2, but they have a very dark brownish colour, which is almost black. This shows the lavas are high in molten material with few crystals, when that happens felsic melts freeze into a glassy black-coloured material, like obsidian. Explosivity during the Escorias flow eruption was minimal. The west vent formed a minuscule crater, 100 meters wide. The east vent formed a shallow crater a little under 300 meters wide, with a subtle ring of darkish pyroclastic material. Eruption was basically effusive, with pyroclastic volume being minimal in relation to the lava. Ejecta from both vents is dark-coloured, which I suspect could be andesitic or basaltic-andesitic in composition.

Next to the Manantial Pelado volcano, lies Lengua the Vulcano, which, as the name implies, is a gargantuan lengua (tongue) of lava. It consists of two slightly overlapping sheets of lava that fill a deep glacial valley, with the longest tongue reaching over 100 meters thick at the lava flow front. My guess is that the overall flow averages 150 meters thick or more. At the vent, lava piles up to a height of over 100 meters despite being on the very steep slopes of the U-shaped valley. The total volume, I believe, is probably a bit below 2 km3. The large thickness could suggest this flow may have erupted more slowly, however, the channels are very wide, which could suggest high eruption rates. Maybe it was actually its composition that was different and favoured a strong viscosity and increased thickness. Its composition is rhyodacite with 70wt% SiO2, which is more strongly silicic than the Quizapú and Escorias eruptions. The appearance of the flow seems different too, of a cream colour, much brighter than other flows, although there are some parts and bands which are dark like the Escorias flows. I’m not sure whether the colours could have been given by overlying pyroclastic material, but the cream tone of the flow is different from the pure white pumice of the 1932 eruption, and the Lengua de Vulcano eruption doesn’t seem to have produced substantial explosions, so I think the flow is bright because it is very crystal-rich, because it has felsic minerals that give the flow a brighter tone than the glassy lavas of Quizapú and Escorias, but only close inspection could confirm this. The eruption had very little explosivity, with no cone or crater visible around the vent at all. There is only some possible pumice upslope. One article mentions the Lengua de Vulcano eruption as happening around 1760 and cites personal communication as the evidence. However, without more information, I’m skeptical whether this is accurate.

Lengua de Vulcano lava flow.


The Quizapú eruption sequence.

On 26 November 1846, a new eruption broke out from the flank of Cerro Azul stratovolcano. The new vent opened 1.5 km north from the centre of the large summit crater of Cerro Azul. Quizapú was born, or reborn, as this may have been the return from the dead of Cerro Azul. The name Quizapú means “who knows?”, which is what a local said after being asked about the name of the volcano. There was no fumarole or volcano in the place the eruption happened, so of course no one had a name for it. Within 15 days of the outbreak, lava had flowed 7 kilometres to the east.

Not many contemporary details are known of this eruption. No ashfall was reported, so the eruption must have been predominantly effusive. The date of the stop is not entirely clear. Some lavas may have flowed westward until 1853-1854, with the eruption lasting several years, erupting a total volume of about 5 km3, according to existing estimates. The flow field is very complicated. Many tongues of lava were active at different times, most of them with thicknesses somewhere 25-50 meters along the edges, and with up to three overlapping sheets of lava in some areas. The final lavas of the eruption seem to have formed somewhat thicker flows, with 50-150 meters, and in one location up to 150 meters thick along the edge, which only advanced 2-3 km from the vent. The morphology shows this eruption was longer lasting than any other postglacial flows of felsic composition in Calabozos or Laguna del Maule volcanic fields.

The chemistry of this effusive phase is very interesting, the lavas include dacites, andesites, and basaltic andesites. Silica ranges from 53 wt% to 68 wt%, but you can seemingly find every intermediate value. Phenocryst content is about 15 vol%. The more silicic lavas are slightly more crystal poor than the more mafic types. Presumably the lavas erupted are a mixture between the basaltic-andesite and dacite end-members of Calabozos. The flows are very dark coloured.

White pumice from 1932 thickly mantles the ground, a rubbly dark wall of lavas from 1846-53 blocks the way, but is cleaved on the left side by a water stream, and the majestic Descabezado Grande towers behind. Image by Jfbustos, Wikimedia.

Starting around 1907, Quizapú started making frequent ash clouds and steam plumes. In 1914, it sent a plume of ash to 7 km in height. Between 1916 and 1926 explosions became more frequent, sometimes happening daily, and incandescence lighted up the volcano at night-time. Explosions weakened a little but continued until 1932.

Suddenly, on 10 April 1932, Quizapú reinvented itself. The postglacial dacite eruptions of the Calabozos Complex have been entirely almost entirely effusive, but the 1932 eruption would be the entire opposite, a fully explosive eruption. A column of ash rose to 30 km high and spread out into an enormous umbrella. Booming sounds rattled doors and windows in Santiago de Chile during 18 hours of continuous plinian eruption, but fortunately, no one was anywhere near the event when it happened. There wasn’t much property damage either. A catastrophic release of water from a lake impounded by 1846-53 lavas, which took place a few months after the plinian eruption, was actually the most destructive event related to the eruption, and caused damage along the Maule River. Total volume erupted is estimated at 9.5 km3. The eruption opened and ended with minor volumes of basaltic-andesite and andesite, but the bulk of the volume is dacitic with 66 wt% SiO2 and 15 vol% phenocrysts.

Quizapú painted the landscape in a white dacite pumice, which turns pink close to the vent. I presume the pink colour is due to high-temperature iron oxidation. Two pyroclastic flows are visible as pink, fluid-looking deposits that run following the steepest lines of descent from Quizapú to a distance of 6 km westward, and some other possible flows reach 7 km down the opposite side, to the east. Interestingly, there are very few pyroclastic flows. Almost all the ejecta went airborne and then came down as showers of pumiceous material that still today mantle the ground for tens of kilometres downwind. A crater 600 meters wide and 200 meters deep formed in this eruption. Given the size of the crater, I expect the intensity of the eruption was lower than that of plinian rhyolite eruptions of Laguna del Maule, even though, perhaps due to its great duration, the volume was very substantial.

Two weeks after the plinian eruption, activity shifted to Descabezado Grande, with plumes of steam rising from the flanks of the headless volcano for the first time in recorded history. At some point during June-July 1932, Descabezado Grande’s north flank started producing explosions. These explosions gouged out a massive crater 900 meters wide, and 200-300 meters deep. Explosions continued into 1933, producing plumes of ash, at times up to several km high. The new crater, now known as Respiradero (literally meaning vent), ejected only lithic material, old lavas of the headless volcano blown to ash-sized bits.

I think the activity of Descabezado Grande is very interesting and poorly understood. Quizapú is 1.5 km north of Cerro Azul summit crater. Respiradero is 1.5 km north of Descabezado Grande summit crater. Each of the new vents represents the reactivation of its respective stratovolcano, which had been dormant since pre-glacial times. I believe this is what happened: the plinian eruption of Quizapú, of the Cerro Azul conduit, decompressed the entire magmatic system, and Descabezado Grande is part of this system, which in turn, when affected by this plinian-induced decompression, had a buildup of gas bubbles, pressurized magmatic gasses trapped in the volcanic conduit eventually broke upwards, blasting out a gaping hole in the side of Descabezado Grande in a series of repeated gas explosions.

From the lower left corner to the upper right corner, we have four volcanic vents. First, Respiradero, rimmed in gray lithic ashes. Second Descabezado Grande, with its large glacier-filled crater atop a large scoriaceous cone. Then we see Quizapú, surrounded by lava flows of 1846-53 and pinkish proximal pumice of 1932, and rimmed in black andesitic scoria from the final explosions of 1932. And finally, we see sharp-peaked Cerro Azul, with a subtle ice-filled crater in front of the peak, which towers over the Invernada River valley behind.


The origin of the dacite magmas

A 2012 article uses thermobarometry to estimate that the magmas of Quizapú were stored at about 5-6 km depth, while a 2021 article estimates an even shallower depth of 4 km. In either case, the origin of the dacites seems to be from a very shallow magma chamber. The depth of this magma chamber seems to be slightly more shallow than that of Laguna del Maule, as far as I can tell. But where is the magma chamber located exactly? The only place with good evidence for a shallow magma chamber is the Calabozos resurgent dome where there is topographical evidence of a ~65 km3 resurgence that continued into postglacial times, large-scale deflation during the 2010 mega thrust earthquake, and numerous hot-springs that are found around it. Quizapú erupted almost 10 km3 DRE of magma in less than a century. It must have drawn its magma from a very substantial magma chamber, and the only clear option is Calabozos. Descabezado Grande must be part of the system too, given its behaviour in 1932-33, and, by extension, the Lengua de Vulcano lava and Alto de las Mulas lava flows may have fed from the resurgent dome too. This shows an unusual, largely unrecognized, behaviour. Though it could be similar to the Katmai-Novarupta eruption of 1912. A shallow, felsic magma chamber which regularly supplies magma to three neighbouring stratovolcano systems: Cerro Azul, Descabezado Grande, and Manatial Pelado. In fact, Calabozos itself rarely erupts, unless it does so through the shield volcano at the southern end of the resurgent dome, Descabezado Chico. Unlike Laguna del Maule, which has a ring of cone-sheet-fed vents around its chamber, Calabozos doesn’t erupt around the resurgent dome and instead sends magma to nearby well-established conduits. Calabozos is also a volcano proficient at building up magma through resurgence, and eventually producing VEI-7 eruptions, while having a less silicic dacite-rhyodacite composition than the rhyolite of Laguna del Maule. I’d say Calabozos has a greater subduction influence in its volcanism, and in fact is a stratovolcano complex, while Laguna del Maule is more intra-plate-like.


The silicic trio of the Maule-Neuquen

Domuyo, Laguna del Maule, and Calabozos, the three silicic volcanoes of the Maule-Neuquen, are next to each other and have silicic compositions, but are very different to each other. Combined, they make up a vigorous ongoing silicic flare-up. Domuyo, in the Neuquén Province of Argentina, has the most magma but the least eruptions. Laguna del Maule, has the least magma and the most eruptions. While Calabozos has very different eruptive dynamics to the other two, and is also predominantly dacitic, while the other two are predominantly rhyolitic.

This is my take on the three systems potential:

Domuyo I think has the potential for the largest eruption, some 150 km3 of magma, but this is probably a volcano with a really, really long history behind, and that resurgent dome may have been there for many hundreds of thousands of years. In the scale of human lifetimes, the possibility of Domuyo going caldera seems insignificant.

I don’t think Laguna del Maule has enough magma for a VEI-7. Although many would disagree with this. Only a small VEI-6, and maybe not even that at present. I‘m not too sure what to expect from this system in the future. The youngest two eruptions have probably been rhyodacitic. Could it be that the system has “weakened” from its former rhyolitic eruptions to rhyodacitic ones? I don’t know.

Calabozos has rapid caldera cycles. The previous two ignimbrites happened 150,000 years apart, and 150,000 years have passed since the last one. The system has been overflowing dacite from multiple satellite systems, and the historical series of eruptions is incredibly rare. Something is going on, although I’m not sure what. I think Calabozos will be the next one to do a major caldera-forming eruption, a low end VEI-7, possibly. Thankfully, this volcano is in a remote uninhabited area where direct effects of the eruption will not be too damaging, and in the southern hemisphere so that the impact of a volcanic winter impact will be lessened. The volcano is also probably spent after the whole Quizapú sequence, and it might very well take a few centuries or more for it to erupt again.

And thus this series about the Payenia Volcanic Province, and the Maule-Neuquén silicic volcanoes comes to an end. Although there are many questions that remain about this region. Like, for example, flat slat or no flat slab? I’m not as sure as I was during the past article about this question. I ended up talking a lot more about this region than I meant to, but nowhere near as much as I wished I was capable of.


First, E. C., Hammer, J. E., Ruprecht, P., & Rutherford, M. (2021). Experimental Constraints on Dacite Magma Storage beneath Volcán Quizapu, Chile. Journal of Petrology62(5). https://doi.org/10.1093/petrology/egab027

Hildreth, W. y R.E. Drake, 1992. Volcán Quizapu, Chilean Andes.
Bulletin of Volcanology 54 (2):93-125.

Hildreth, W., Fierstein, J., Godoy, E., Drake, R. E., & Singer, B. S. (1999). The Puelche Volcanic Field: extensive Pleistocene rhyolite lava flows in the Andes of central Chile. Revista Geologica De Chile26(2). https://doi.org/10.4067/s0716-02081999000200008

Hildreth, W., Grunder, A. L., & Drake, R. E. (1984). The Loma Seca Tuff and the Calabozos caldera: A major ash-flow and caldera complex in the southern Andes of central Chile. Geological Society of America Bulletin95(1), 45. https://doi.org/10.1130/0016-7606(1984)95

Pritchard, M. E., Jay, J. A., Aron, F., Henderson, S. W., & Lara, L. M. (2013). Subsidence at southern Andes volcanoes induced by the 2010 Maule, Chile earthquake. Nature Geoscience6(8), 632–636. https://doi.org/10.1038/ngeo1855

Ruprecht, P., Bergantz, G. W., Cooper, K. M., & Hildreth, W. (2012). The Crustal Magma Storage System of Volcán Quizapu, Chile, and the Effects of Magma Mixing on Magma Diversity. Journal of Petrology53(4), 801–840. https://doi.org/10.1093/petrology/egs002

78 thoughts on “Puelche and Calabozos. Calderas of the Maule

  1. Absoutley fascinating Thank you
    Perhaps South America is about to do a sillic flareup

  2. Pueleche is like a fossil supervolcano, only one that is still mostly intact and could revive in the future if conditions are suitable. Well, maybe not a VEI 8 but I think a mid to large VEI 7 probably makes itself worthy of the name.

    And any volcano that erupts fluid felsic lava is interesting 🙂 and flowing 30 km that is crazy, must have looked like a basalt flow, yet it is dacite.

    • Yes, the Escorias lava flow must be one of the longest, if not the longest, known dacite lava flows. It’s quite impressive.

  3. Fascinating story. Every volcano is different. In an area like this, the topography should not be ignored. I once flew in a small plane along the Andes in this region. Every now and then the ground would get very close to the plane. The elevation differences even over small distances are enormous. That will affect the direction underground magma will take, so that also can change over small distances

    • I give people who are able to fly in those tiny planes a lot of credit, especially over terrain as rugged and dangerous as the Andes. What an experience that had to have been Albert!

      I’d be scared witless, and indeed I’ve never been in such a vehicle even just above sea level in the tropics as my parents have done many times, let alone above the Andean plateau!

  4. Beautiful.
    They could all erupt or not, not now at least.
    This system in the Andes is the one that depicts the time scale of plate tectonics best. This is the one system that had subduction most of the time of Earth’s history. The system that devoured the Phoenix Plate and now eats parts of the Nazca Plate. The system that is probably the closest relative and heir of Panthalassa, a lost ocean, that is incorporated in the Andes.
    I tend to believe that the Inka somehow had a feeling for those mystics further north. They build sacred burial sites on Misti and other volcanoes.

    Thank you for that detailed, precise, very scientific and interesting piece of work.

  5. Would the ~100 m below sea level zone to East, near Argentina’s Atlantic coast suggest ‘back-arc’ spreading ??

    Or simply an arm of sea isolated during last glacial low-stand, else would now be a ria… ??

    Disclosure: I’d an aunt and cousin in Argentina, but they vanished during chaos of ¾ century’s right-wing regime. So, either they ‘went dark’, changed addresses, burned letters etc etc, or the teen boy also inherited my ‘Foot in Mouth’ social disease but, in his case, died of it…

  6. Calabozos doesn’t erupt around the resurgent dome and instead sends magma to nearby well-established conduits.

    There are a few other volcanoes that seem to have satellite storage areas connected by conduits more persistent than dikes. The article also mentioned Novarupta, but we can add Kilauea, with (former?) storage under Pu’u O’o and a stable hydraulic link between that storage and the main summit chamber, and several of Iceland’s volcanoes, with notably Grimsvotn and … Thordarhyrna? … possibly being hydraulically connected, and perhaps long lines of connected volcanoes in the east rift zone, with extensions into the “dead zone”.

    These satellite storages and associated satellite vents all have in common an extensional setting, as well. Calabozos and Novarupta are in back-arc settings, Pu’u O’o is on a rift zone of a very large shield volcano, and the Icelandic ones are on a divergent plate boundary.

    The back-arc setting allows satellite storage to become established two-dimensionally around the central chamber, unlike the others, where they tend to be one-dimensionally distributed along a rift axis. The two-dimensional case leads to the obvious possibility that the expanding network of storage and conduits can fuse together into a single, massive chamber, and in the process undermine the overlying crust by removing “support pillars”. It makes me wonder if this is a stage in the development (and, post-collapse, perhaps redevelopment) of large caldera systems.

    The mentioned lack of plinian deposits with the Calabozos ignimbrites also points to a caldera collapse source. Though there may or may not be plinian eruptions at the perimeter when it starts, once the roof begins falling in in earnest the vent becomes so wide it would take ludicrous flow rates of material out to make a tall eruption column: an inverse of the “nozzle effect”. So what you get is a sort of ultra-pelean eruption instead, from that point on. Magma just foams up out of the hole and spreads across the landscape like an overflowing top-loading washing machine, kicking up a co-ignimbrite cloud but no plinian column and not stopping until there’s nothing left in the upper chamber that can decompress and degas. (And then the exposed conduit from the lower chamber will ooze some basaltic andesite into the hole this made.) The last ignimbrite-forming Calabozos eruption shows signs of having been like that, which suggests a very wide vent, aka a caldera collapse.

    It also means the system may be capable of this style of eruption again in the future. Best not to be nearby when that happens, as the scorching suds will overtop ridges and peaks and destroy anything within many tens of kilometers. Terrain that would protect against mere lava flows, lahars, and such is no shield at all against these huge events. Only distance works.

  7. I’ve only been to Argentina once, for work, and flew back north over the northwestern part, over Lima thence to Mexico City and eventually back to my home here in eastern Australia.

    I was struck by the landscape of NW Argentina – bleak. Volcanic! A fascinating place glimpsed out of a 747 window.

    My comment though is about the volcanic complexes to the west, since one of the areas of my work has been in copper extraction. The interesting thing is the copper precincts in Chile appear to be linked to ?resize=700%2C461&ssl=1″ rel=”nofollow ugc”>the intermediate composition volcanics in Héctor’s previous fine article.

    So why don’t such hydrothermal systems form in these rhyolite/rhyodacite zones? I have no idea. But Chile is the biggest copper mining country on Earth due to their volcanoes.

    • Ooh, Héctor’s excellent graphic was rendered by WordPress! I wasn’t expecting that. But the copper mining areas in the Andes tend to follow the green blobs, Mines like Escondida and Spence.

      Let me try another graphic, this one from the wiki… 🙂

    • Hehe that didn’t work. Here’s the graphic of copper mines in Chile: ” rel=”nofollow ugc”>Map.

      • So there you have it. The copper mines follow the Cordillera not the felsic and mafic zones.

        I’m a chemist not a geo guy, so I don’t know why. But the correlation with the volcanic and metallic chemistry springs out of the map. Which is always fun, if you are a scientist.

        • If you need hydrothermal systems, it is better to be on the western side of the Andes. The eastern side is bone dry. The west gets the winter rains and snows, and the snow melt keeps some river valleys moist in the summer.

          • With the exception of Atacama, right?

            In the context of copper it might be interesting to know that the name is derived from Cyprus where the Romans discovered it in the Troodos mountains that are also famous for ophiolites indicating the subduction of Tethys.

          • There’s is some more to it though which is younger (around 2Ma) whereas the oceanic slab is stematid to be from 90 Ma:
            “While there is widespread consensus on the role the Eratosthenes Seamount played in the uplift of the Troodos Mountains, the tectonic architecture and the kinematics of the tectonic processes as well as the mechanisms that caused uplift are less well constrained. A popular model is that the uplift of the Troodos Massif was caused by serpentinization of mantle peridotite and that the massif represents a huge diapir”…

        • That’s an interesting question. The intermediate volcanoes are stratovolcanoes which have a distinctive composition and eruption style. The porphyritic diorite dikes of copper deposits are probably the crystal-rich andesite magmas that are erupted by many stratovolcanoes. One particularity of subduction-zone stratovolcanoes is that they often feature vigorous fumarolic activity, say Kawah Ijen or White Island. And the interior of stratovolcanoes is often altered into a yellowish clay. So I’d guess the formation of copper deposits has to do with this fumarolic activity. This is all, in the end, probably a consequence of water from the subducting ocean crust.

          • Is it as simple as higher temperatures in the subducting arc compared to the back-arc system?
            Worth noting that copper occurs primarily in crust that has been pushed upwards, albeit it is more commonly associated with acidic volcanism.

          • It could be related to the crust being more saturated with water/salts at this point?

    • The Central Volcanic Zone of the Andes is special in that it features a very evolved stratovolcanism. Highly crystal rich andesite and dacite volcanoes are very common. There are some types of volcanoes that you only find in the Central Andes, like crystal-rich dacite supervolcanoes, ignimbrite domes (like the Purico Complex), or massive fields of crystal-rich andesitic coulees. The copper deposits seem to have something to do with this family of crystal-rich andesites and dacites. The South Volcanic Zone has mostly basaltic-andesite magmas, so it’s probably not very proficient at making those copper deposits.

      • Is this similar though?:
        ” It is generally accepted that these andesites and dacites in SW Japan are attributed to melting of the subducted plate in late Cenozoic time (Feineman et al., 2013; Kimura et al., 2014; Morris, 1995; Shibata et al., 2014).”

        The Tada Silver and Copper Mine is close to the Chugoku Mountains, precisely in the Hokusetsu mountain range southeast.

        • Yes, those andesites and dacites from Japan are probably similar to the ones from the Central Volcanic Zone of the Andes. In fact, the volcanoes in Honshu remind slightly of the CVZ, with those thick lava flows and dome complexes everywhere. Although the CVZ is more “extreme” and has unique or almost unique volcanic types related to the highly crystal-rich, very high viscosity andesite and dacites there.

  8. A pleasant and professionell study! In this area volcanoes don’t look much different to “innocent” non-volcanic mountains. It is hard to imagine how magma can flow through the thick orogenic mountain chain of the Andes.

    Concerning magma types in my mind “Dacite” is tied to “volcanic TNT” because both Mount St. Helens and Pinatubo were Dacite volcanoes which exploded like volcanic dynamite. Is it usually really the most explosive and dangerous magma?

    • Thanks Volcanophil.

      I would create the term bazooka volcano for some stratovolcano systems of crystal-rich andesites and dacites that explode regularly in a plinian manner with VEI 4-6 eruptions. Crystal rich andesites and dacites are often a slurry of rhyolite melt and mafic crystals. So they have the same viscosity as rhyolite, if not higher due to being colder and probably more water-rich. Quilotoa and Pinatubo would represent the typical dacite bazooka volcano. But I’m pretty sure there must be more contributing factors to these volcanoes being more violent, since not all crystal-rich andesite and dacite volcanoes behave so violently. Bagana is a purely effusive crystal-rich andesite for example and is very different from an andesite bazooka volcano like Kasatochi.

      • Bazooka: Mount Pelée-eruption Martinique 1902?
        Andesites, subduction zone, Stratovolcano.
        Considered worst volcanic eruption of the 20th century due to death toll, basically all of St. Pierre.

        • A lot of andesitic eruptions are related to dome building due to high viscosity followed by dome collapse, not sure if this was the case with Pelée but it certainly has been with other Caribbean arc volcanoes. Soufriere Hills & Soufriere St. Vincent were two good examples, but also Semeru.

          • Semeru isnt really this sort of volcano though, it makes domes in a way but not the sort of massive spiny domes that Pelee or Shevluch do. Its magma seems to be actual andesite, probably relatively crystal rich but the melt isnt felsic like it probably is in the other examples. Andesite melt liquid is apparently not very viscous at all, if erupted crystal free, but most andesites are rhyolites which mafic crystals that average out. Probably the majority of basaltic andesite lavas are andesite melt with mafic crystals, really any melt 60% SiO2 or lower seems to be free flowing more or less, even some more silicic examples too. Semeru is also not very explosive, maybe somewhat gas poor after being open for the last 60 years.

            The last two ‘eruptions’ were from repeated lava flows filling a steep valley that then eventually collapsed into a pyroclastic flow. Quite a lot of eruptions that are identified as ‘dome building’ are probably this sort of activity instead. It happens frequently on Fuego and Stromboli, and sometimes on Etna, none of which are viscous volcanoes, nor known for peleean eruptions, yet Fuego is probably the volcano most prone to pyroclastic flows right now.

      • Does the crystal-rich magma show a tendency of magmatic intrusions to stay below the mountains completely? Orogencies often have plutons (granite …). Volcanism of the Andes might only be the tip of the plutonic iceberg.

        How thick is the crust below the Andes compared with Himalaya? How can magma find a path through this giant mass of rocks? Himalaya may also have had some magma intrusions which become plutons (granite). The early Alpine orogency had only few intrusions. More volcanism happened during Cretacous before the Alpes became mainland, when they were an island arc. https://en.wikipedia.org/wiki/Geology_of_the_Alps#Intrusions

        • Basically the Andes and the Himalaya are different as the latter represents a collisional orogeny. The Andes stand for a non-collisional orogeny. With collision continentental crust is meant.

          The crust should therefore be thicker in the area of the India collision as continental crust is six times as thick on an average.
          Only before India arrived the situation was comparable to South-America’s west coast because of subduction of an oceanic plate.

          Concerning the Alps we have a similar situation to India as there was subduction of an oceanic plate first and collision with the African Plate later.

          To find a comparison for the Andes we have to travel the Ring of Fire I’d say and find numerous examples like Japan or maybe New Zealand, although New Zealand again is special due to the Chatham Rise and some other submarine igneous provinces. Kamtchatka though comes to mind, also Alaska.

          • So basically I would draw the conclusion that volcanism tends to get to an end when the water disappears which would then be the case in the European Alps and also the Himalaya.
            But that is so far an idea. There is certainly some research about this question.

          • Yes, I also think that the continental collision zone shuts the paths for possible magma movement. But the granite of old orogencies show that there still may happen intrusions which stay deep below the surface.

        • It’s very thick crust under parts of the Andes, particularly the Central Volcanic Zone.

          I think, based on morphology alone, that I would put the Cerro Chao lava flow, in the Altiplano-Puna (Central Volcanic Zone), and by extension some of the nearby domes with the same composition, as some of the most viscous, if not the most viscous, lavas in the world right now. They are crystal rich rhyodacites. The sheet of lava ranges in thickness from 100 to 300 meters along the sides, and the main lava channel is almost 3 km wide. It comes from a supervolcanic-sized system that hides in plain sight, I talked a little about it in the 1st part of the 10 volcanoes with supereruption potential series, where it was referred to as El Tatio volcano.

          • But it also depends a lot on the individual’s volcano plumbing. While most systems in the CVZ are highly viscous, there are also very fluid monogenetic volcanoes towards the southern part of the CVZ, near Antofagasta de la Sierra. The volcanoes around Antofagasta erupt basalts, trachybasalts, basaltic-andesites, and basaltic-trachyandesites, through some of the thickest crust in the planet, the flows are probably crystal poor given that they are fluid, making pahoehoe flows that are down to tens of centimetres thick.

          • The thick crust also makes it difficult to find plutones there. But I’d assume that the crystal rich rhyodacite magma is an indicator for large production of granite/dolerite/gabbro/syenite/…
            The volume of a plutonic rock chamber can be as high as VEI8, but remain completely in the earth until it’s solid. Yosemite Nationalpark is a pretty example of granite with the Halfdome: https://en.wikipedia.org/wiki/Half_Dome

  9. What a great series, Hector!
    So much have I learned.
    Ever since Chaiten went off (where a good friend’s wife was born/raised then got trapped there by the eruption while visiting family), I became fascinated with the region…and it was then that I first found Ralph Harrington’s then Erik Klementi’s then Jon Freiman’s blogs (which as we know was where many of us future VC members first met each other) that were providing invaluable updates on the eruption which my friend was then able to forward to his wife and her cut-off community.
    Anyway, of all the great info/analysis you’ve provided, the post-mega-thrust of 2010 causing inland deflation(s) was the most interesting…since it’s so counterintuitive and “apparently” somewhat in conflict with long term eruption observations.
    Since 1900, almost the entire west coast of Chile has experienced a series of mega-thrust events at one time or another, and I have often wondered if during the decades that followed many of these events if there would be an increase in magma production (and volcanism) as new material gets thrust down to the “melt zone” beneath the inland volcanoes’ magma chambers. But your analysis and data shows deflation post 2010, yet I wonder if the localized deflation is leading up to a greater inflation rate later on, such as what’s now occurring under Laguna del Maule?
    In general are there any data that you know of that shows any kind of possible correlation (either more or less) of active SA volcanism being related/correlated to prior earthquake activity? One would think that due to the long latency that would be expected (quake vs. eruption), trying to find any direct causative linkage might be real difficult to draw any conclusions from, but I wonder what your thoughts might be?
    For instance, in the 67 yrs prior to the ~ 1960 M9.5 Valdivia earthquake, there were 7 eruptions in Chile (that we know of)…but all were significant being >VEI 2.
    However, in the now-63 yrs after 1960, there have been 13 Chilean eruptions, with three being VEI0 and three others going semi-large (including a VEI5 at Cerro Hudson and VEI4 at Puyehue). So at first glance, it would appear that there have been approx. 2x as many eruptions after the 1960 mega-thrust event (plus many other subsequent mega-thrust events) than before. Is there a trend here?
    If we look further back in time, there have been at least 21 separate ~M8+ mega-thrust events since 1906 along/near the Nazca Plate, so why is there a “relative” spike in eruptions since 1960…and especially since 1988?
    So inanutshell, will the “recent” widespread tectonic unrest that SA has been experiencing (much of it in the last 60yrs) heightened or lowered the chances of more frequent and/or powerful eruptions in the not-so-distant future as new water-soaked land gets subducted then melts and feeds into the various lower Magma reservoirs?
    Only time will tell…but it’s an amazing mystery nonetheless to speculate on.

    • Thanks Craig Heden!

      I think there are good chances for some kind of cyclicity in activity linked to megathrust earthquake cycles, but many it’s weak enough to be overpowered by other processes, and there could be multiple different mechanisms for these cycles, some favouring low activity after the earthquake, and others favouring high activity, there is simply a lot we don’t know. But it’s something interesting to consider.

      I’m not aware of any obvious cases that would show a link. Laguna del Maule was already inflating at full speed before the 2010 megathrust in the area. Its inflation had gradually accelerated between 2007 and 2009. I’m not sure if data with sufficient temporal definition exists for deformation at the time the earthquake happened to know how it was affected. Planchon-Peteroa, the volcano immediately north of Calabozos, did produce ash explosions in 2010 and 2011 that may have been triggered by the earthquake, which were probably gas-driven (I don’t think any fresh magma was erupted).

      • Thanks Hector.
        I’m thinking though, that the quake-eruption linkage would have a long latency in between…possibly on the order of decades….which is why I was curious as far as whether the current unrest can possibly be traced to recently past mega-thrust earthquakes such as the 1960 M9.5.
        But overall, virtually the entire SA coast and the Andes has received a slug of new land getting thrust under it within the last 100yrs…and especially since 1960…and since the linkage between subducting land and volcanism is well known, common sense dictates that the entire Andean chain has been given a shot of geological steroids.

  10. @hector,

    Where does Caldera del Atuel figure in all of this? There is precious little information about it, and I have to imagine if it actually was a 30×45 as the GVP states, the caldera eruption would be rather ancient (likely over 1 million years ago at the minimum).

    According to http://geode.colorado.edu/~geolsci/facultyweb/sternpdf/Bosch%20et%20al%20Abstract%20for%20the%20XIV%20CGCH%20first%20draft.pdf , there are some rather large unsourced rhyolitic eruptions in the northern end of the southern Andes zone, enough to have 1-3 meter thick rhyolite tuff deposits in the city region of Mendoza. And these are not from the Diamante caldera, which has been sourced and identified.

    • I personally don’t think Caldera del Atuel is a 30×45 km caldera. But there is a possible ignimbrite, to the east of Cerro Overo, the volcanic complex in the northern part of Caldera del Atuel, which consists of a deposit about 300 meters thick or more that runs for a length of 28 km through the floor of a valley. However, the presumed ignimbrite can be found both inside and outside Caldera del Atuel, and there is no break or scarp where it crosses the border of the basin, so it probably came from a smaller caldera buried under the Cerro Overo volcanic complex.

      Caldera del Atuel/Cerro Overo is a very slightly back-arc volcano, and part of a chain of mostly inactive stratovolcanoes in the present-day back-arc that reaches to the Llancanelo/Palaoco complex, a series of ancient central volcanoes now exposed as radiating dike swarms and laccoliths, and younger, possibly active, alkali basalt shield volcanoes. So the location of Caldera del Atuel/Cerro Overo in relation to surrounding volcanoes is similar to Calabozos.

      As for those rhyolite deposits, the most likely candidate is Tupungatito and its 4-km caldera, which is the closest to Mendoza.

      • I can see a hint of 7×6 km caldera in the Cerro Overo volcano, but it could be imagination. I don’t think this volcano has been studied at all. This is a consequence of the entire area being mostly uninhabited.

    • Actually, those rhyolite layers in Mendoza could have come from Caldera Diamante. Ages for the caldera forming event of Caldera Diamante are controversial. There are three dates in zircon crystals from the Diamante ignimbrite. One article gives ages of 0,47 ± 0,07 Ma and 0,44 ± 0,08 Ma, but another gives 0,150 Ma. The problem with zircons (and other minerals) in silicic eruptions is that they are often much older than the eruption, since they have been sitting in the magma chambers for hundreds of thousands of years. For example, zircons from the last eruption of Valles caldera ~70,000 years ago, are up to 1 million years old. They date all the way back to the destruction of the earlier magma chamber. To get a good date of an ignimbrite, you have to date the rims of numerous crystals, ideally tens of them. What we do know is that Maipo stratovolcano started to grow inside the caldera around 87,000 years ago, which are the oldest dated lavas. So caldera-collapse happened before 87 ka, but that is about as much as can be said. Those rhyolites of Mendoza could actually come from the Caldera Diamante system, maybe from precursory eruptions with a different chemistry. Of course, they could also come from other volcanoes.


      • Thanks Hector. I would be curious where and why the GVP is sourcing their information that Atuel is in fact a ginormous caldera. It’s referenced in there no only from Atuel itself, but also mentioned in other nearby volcanic systems.

  11. The Beaufort Gyre is stabilizing and it looks poised to release over 23.300 km3 of freshwater into the North Atlantic, which would have significant ramifications for the AMO. In my opinion, this is likely to be the most significant climate event in decades, if things go wrong, this could easily surpass the Pinatubo eruption and 2016 el nino.

    • It’s only a theory from Andrey Proshutinsky. It may not happen, and I’ve not seen any drift to alarm from Woods Hole and the community. Apart from which the Gyre is largely dependent on northern Russian river flows, and if they have a drier year the Gyre could substantially weaken.
      Nonetheless, an interesting topic and something to watch. Thanks!

      • Well, thank you for the piece, but it is no extremely good missing all sorts of dates. Last volcanism on Bermuda estimated at 30Ma.

        Bermuda is sitting on the other side of the MAR compared to the Azores. So now every thinking person would argue that Iceland will also sit and is already sitting on two opposing sides of the MAR.
        However, the Azores have an ongoing volcanism and are sitting on top of a triple junction being very lively indeed. Bermuda is an obviously extinct volcano that started as a seamount and will be submarine one day in the deep future. According to that fate it already has its burial decoration consisting of coral reefs.

        So, a connection here is more than unlikely the Azores being active.

        Now as there is not an awful lot of research about Bermuda I am free to speculate that it could also be a remnant of CAMP. This could only be proven by geochemistry.

        • That is one possibility. I suspect there is a pre-Triassic rift structure here, imaging shows the lithosphere is thinned in this region. The volcanism began roughly 47Ma and definitely isn’t the same wandering hotspot from Mississippi etc. as has been postulated – it just doesn’t add up.

          This was a summary from the Vogt paper:
          1) The Bermuda Rise is elongated at right angles to the direction of plate motion;
          2) There has been little or no subsidence of the Rise and volcanic edifice since its formation—in fact, Rise uplift continued at the same site from the late Middle Eocene into the Miocene; and 3) The Bermuda Rise lacks a clear, age-progressive chain. We infer that the Bermuda Rise and other Atlantic midplate rises are supported by anomalous asthenosphere, upwelling or not, that penetrates the thermal boundary layer and travels with the overlying North America plate.

  12. Héctor,
    That area you have described in your trilogy is altogether adorable, not only for its volcanoes, but also for its long history of subduction and then for these:

    Dreadnoughtus, reconstruction.

    We can only guess why east of your area the most abundant collection of fossilized bones from Titanosaurs, today als known as Lognkosaurs (first half of the expression deducted from language of first people there) are found. In a piece from 2014 about them the authors described bones from 40 different species from Patagonia, Argentina, only one species from Chile and ten different species from Brazil.

    That invites the imagination to wander around.
    1. Huge volcanism in the west, caused by the subduction of the Phoenix Plate.
    2. Island arcs plus mountain building.
    3. A fluvial plain in the east with possibly fumaroles and ideal climatic conditions for breeding.
    4. Tall trees, that favoured the larger species, so the smaller ones went extinct, plus beginning of blossoming plants.

    I also read about another species in India, fossilized around the Deccan Traps which might mean that when they developped there would still have been a connection between South America and Antarctica and also India. So this was the core of Gondwana with huge absolutely awsome herbivores and accordingly, abundant trees and enough rain, all bordered by volcanoes.

    Altogether awsome area, very wonderful to see today and to imagine in the deep past. Those amazing Chilean glaciers wouldn’t have existed.

    • Today we admire a mammal with similarly solid legs, plant-eater as well: The elephant. We wonder how it can walk so lightly and even run with this body mass. It helps us to imagine the creatures with ten times and more the weight of an elephant.

      Now the big or silly question to Albert:
      When creatures go so big could there have been more radiation?

  13. One questian may be allowed, Héctor, that you may answer or not:

    Are you from Argentina or Chile?

    • Spain, actually, though it seems I have my volcanoes a little abandoned.

      • Lots of Spanish speaking volcanoes in the world! Even New Mexico has Spanish as on official language

        • What’s great is that I can access a lot of volcanic literature by knowing English and Spanish. The only ones that still evade me are Asian volcanoes.

          • Time for Mandarin.
            Or Japanese. Better. More volcanism.
            Thank you for answering. I believe though that you like the Andes very much, but also Big Island.
            As to your own volcanoes, Gran Canaria might be interesting. The question whether it could start another eruption cycle or not.

          • Gran Canaria has a very interesting past, with a complicated history, although a lot of Canary Island volcanoes have complicated histories. It is possibly the most outstanding example of ocean shield volcano felsic volcanism, with tens of ignimbrites ~14-10 million years ago. It had the epic rebirth of Roque Nublo 4 million years ago that plundered the whole island in pyroclastic flows, again. And it has possibly the most alkaline lavas in the Canary Islands, having some nephelinites and melilitites, if I remember right, as well as the only rhyolites in the archipelago. Plus, it is still active and has erupted in the Holocene. It is typical of Canary Islands volcanoes to reactivate and become powerful central volcanoes again, like it has happened with Tenerife.

            So there are certainly fascinating volcanoes in Spain too. Although I think I prefer Hawaii Island.

          • If you know English then Spanish is about half readable, although certainly mispronounced internally. Same for French, where about 1/3 of English words originate.. It is also surprisingly not too hard to read the Icelandic texts if you can remember a few words.

            Mandarin and Japanese are beyond me though, and I have tried, I even live with a native speaker of the former for several years now… Mandarin is simple gramatically, and Japanese is easy to say and has a lot of English loanwords, but everything else is a nightmare… But maybe speaking two languages will make it a bit easier for you. You could give Indonesian a try too, it is supposedly quite easy to understand for those who speak English and well, lots of volcanoes 🙂

          • Sounds like Gran Canaria was a sort of bimodal pyroclastic shield volcano, probably like Emi Koussi. Both of them are mostly ignimbrite but on Emi Koussi there are also lots of small mafic vents, some of which look extremely recent, being jet black and the tephra not yet buried in sand or disturbed by the wind.
            Gran Canaria is a lot older and has lost its shape to erosion but the mafic vents are still there, so it is still alive and another 3rd round of major volcanism might well be possible.
            It is fascinating how long some volcanoes live, with multiple generations of central volcanoes. And Lanzarote is still alive, so none of the Canaries have actually died out yet…
            I wonder what volcano that is presently active is the oldest. I imagine the above might be a good contender.

            Seems there might be a different origin to the Canary volcanoes than a plume too, given the only thing that supports that is age progression but the strength of volcanism is quite unrelated to this it seems. Actually, this might apply to all of the Atlantic islands, and including the Cameroon line too, there is no age progression like there is in the Pacific. The volcanism on these islands also seems far more prone to forming ignimbrite, and even mafic volcanoes make stratocones, like Pico or Fogo.

          • I think Fuerteventura is over 25 Ma old, and is still active. Lanzarote is very old too, but the age is not known, bouguer gravity anomaly maps show the central shield volcano is buried under the Quaternary lavas, where the strongest concentration of young vents is, but doesn’t outcrop anywhere. The only outcrops are from later shield volcanoes to the north and south of the central shield.

            Gran Canaria’s activity was monstruous, it was erupting ignimbrites every 30,000-40,000 years immediately after its basaltic stage ended. As far as I know it was not bimodal, it went from full basaltic, to full rhyolitic/trachytic:

            The subaerial Miocene evolution started with the rapid formation (<0.5 Ma) of the exposed, mildly alkalic shield basalts. The basaltic shield phase ended between 14.04 ± 0.10 Ma and 13.95 ± 0.02 Ma
            and was followed by a 0.6-m.y. magmatism of trachytic to rhyolitic composition (Mogán Group). Single-crystal 40Ar/39Ar laser dating shows that the ash flows erupted at intervals of 0.03−0.04 m.y., with peak eruption rates as much as 2000 km3/m.y. during the initial stages of silicic magma production (Lower Mogán Formation). After the rhyolitic stage, >500 km3 of silica-undersaturated nepheline trachyphonolitic ash flows, lava flows, and fallout
            tephra, as well as rare basanite and nephelinite dikes and lavas were erupted between 13.29 Ma and 13.04 Ma (Montaña Horno Formation) and 12.43 Ma and 9.85 Ma (Fataga Group). This stage was accompanied and followed by intrusive syenites and a large cone sheet swarm in the central caldera complex, lasting until at least 8.28 Ma. Following a major, nearly nonvolcanic hiatus lasting ~4.7 m.y. (Las Palmas Formation), eruptions resumed with the local emplacement of small volumes of nephelinites, basanites, and tholeiites at ~5 to 4.5 Ma, with peak activity and eruptions of highly evolved phonolite magma between 4.15 and 3.78 Ma (Roque Nublo Group).


          • Btw, it seems interesting to me that Spain tackled the whole mountain chain from North to South whereas the weaker 😉
            Britons, French 🙂 and Portuguese only colonized relatively flat land.
            The reason me thinks might be gold.

            As the volcanic area as a whole was Spanish speaking or using different languages of tribes, reports about volcanic eruptions then might be lost. So there should be several volcanoes there who might be culprits for mystery eruptions.

            Decent science and globalized knowledge certainly needed a lingua franca, which was English.

          • The lingua franca of science was greek, becoming latin at the time.

            Agriculture can be a major factor in who colonizes where. English crops require moist summers. Spanish are used to hot and dry summers. The vikings needed summer grass. That may explain a lot of what you see. South America was divided between the Spanish and Portuguese by the pope. That was pretty effective, in hindsight

  14. A couple of decent quakes in my neckadawoods in northern California in the last 18hrs.
    The very shallow (near surface) hypocenters of the M5.5 and M5.2 shocks are located directly under Lake Almanor about 15 miles south of Mt. Lassen.
    When I checked the nearby Lassen seismos and GPS, it appears that Lassen has been undergoing a rapid rise since early January…GPS is indicating maybe as much as a 6 inch rise. That a lot for just ~4 mos.
    ATTM, it’s difficult to say if there is any volcanic component to the rate of location change…it could be just tectonic in origin, or possibly a sign of hydrothermal activity, but all the GPS stations surrounding Lassen are showing variable amounts of elevation change and a slight shift to the NW.
    Anway, something interesting to monitor in the coming months…especially if the Lassen area pops off a few jolts (normally the mountain is rather quiet).


    • There are eruptions there quite often it looks like. Lassen didnt erupt for 22000 years until 1915 but there was an eruption creating Cinder Cone in the 17th century and at Chaos Crags about 1000 years ago. A second eruption at Chaos Crags might have happened in the 17th century too.

      The eruptions makign Chaos Crags also were a series of several over about 300 years so perhaps a similar situation might be occurring if this uplift means anything. Lassen might well get a lot taller if that is the case, or we get a new mountain to add to the state map 🙂

  15. I only just finished this article, Hector. That’s how slow and methodically I read and absorbed the info, similar to the other articles.

    Brilliant, bravo! Spectacular stuff!

Comments are closed.