After the recent developments of Iwo-Jima volcano, I’ve grown interested in whether submarine calderas are capable, or not, of producing substantial stratospheric sulfur injections, also in the effects and frequency of these injections, and eventually, as one thing led to another, in the identity of the eruptions behind them.

Large volcanic explosions emit vast amounts of sulfur dioxide gas into the higher levels of the atmosphere. The SO2 then reacts with water and oxygen to form sulfate, light colored particles that reflect sunlight back into space, which can cool the atmosphere and disrupt atmospheric circulation patterns. The most famous example was the eruption of Tambora in 1815, which triggered the so-called “year without a summer”. In the US, following the Tambora eruption, a persistent “dry fog” was observed, which reddened and dimmed sunlight. This was the volcanic sulfate veil of Tambora.

The 1815 eruption was one of the three largest volcanic sulfate spikes recorded in polar ice during the last 1,000 years. Volcanic sulfate precipitates down to the surface and can be recorded in ice cores, obtained by drilling into thick accumulations of ice, mainly in the polar regions. Some aspects of the sulfate spikes can be used to infer the eruptions that caused them, most importantly, the size, with the amount of volcanic sulfate being roughly proportional to the eruption size. There’s also the partitioning of the signal between the poles, which indicates roughly the latitude of the eruption. For example, eruptions in tropical latitudes are found in both poles, while eruptions from mid-latitudes are often only found or are far more intense in that hemisphere’s pole. I was, however, interested in finding whether there could be other clues gathered from the characteristics of the spike, mainly its chemical signature, with the objective of finding out if a submarine caldera may have been behind.

There are, in fact, many submarine calderas that collapse explosively, often in shallow waters. For example, in the submarine Tonga-Kermadec arc, almost every volcano can be considered a caldera system, and there are about 70 volcanoes in it, one of the longest ones in the world.  Particularly, the northern Tonga calderas form colossal volcanic edifices constructed in the last few million years (starting after the Lau Basin formation) and seem almost entirely built of submarine pyroclastic turbidity currents during caldera collapses. A similar thing can probably be said for the Izu-Bonin arc, and to a lesser extent, the Marianas and South Sandwich Islands. Submarine calderas are also somewhat smaller than the ones found on thick continental crust, so while their eruptions are smaller, they are also more frequent, and if they are indeed capable of injecting sulfur into the stratosphere from depths of a few hundred meters underwater, then it’s likely a lot of volcanic sulfate spikes of the poles come from submarine volcanoes.  So I was particularly interested in seeing if some characteristic in polar ice cores could reflect the nature of the eruption (a submarine eruption being very different in terms of plume chemistry…) and also to find additional criteria/s to sort the spikes into their origin eruptions. I wouldn’t have had any luck with this effort if it weren’t for the data from the SPC14.

SPC14 is an ice core extracted from the exact location of the South Pole. Located at high elevations of nearly 3000 m, far from the oceans that are an alternate source of aerosols, and also probably with little seasonal weather variation, its data has proven particularly useful due to the quality of the data compared to noisier ice cores. It also came with a second, unexpected volcanic chemical signature that was exactly what I was looking for.

The data I’m working with comes from  Dominic A. Winski, 2021, available from: https://www.usap-dc.org/view/dataset/601851

I’ve noticed three things that might be useful, which are:

Duration: The volcanic sulfate spikes related to caldera-forming events seem to always have a very similar duration. Whether it’s Cosigüina or Tambora, almost all of these obvious sulfur injections last between 2.5 and 3.5 years, regardless of the massive differences in size. So, as far as I’ve seen, the duration of the signal doesn’t seem related to the size, but it’s probably just the time that the aerosol cloud takes to fall through the atmosphere. Eruptions with an intensity similar to Pinatubo or higher, likely reach a similar range in altitude, most likely about 30-40 km above sea level, and take a similar time for their veil to be removed downwards. This duration can help spot the more significant events and distinguish them from background noise or small near-pole eruptions, which is particularly easy for the SPC14, but also works for all ice cores that have enough time resolution. There are also four sulfate spikes (since 1000 AD) that lasted 4-4.5 years, and I have reason to suspect that these were from basaltic-andesite melt eruptions, which, due to their high temperatures, sustained unusually tall plumes. One of these long spikes is a historically documented basaltic andesite eruption, and two other events have basaltic andesite eruptions as their most likely candidates based on timing/size.

Time to peak: The time it takes from the onset of the signal to the peak seems to be related to the distance to the volcano, particularly I’ve seen that pre-1000 AD eruptions from the South Sandwich Islands, near Antarctica, (notably the Vostok tephra of Candlemas Island) have very impulsive onsets rising to an immediate peak, seemingly faster than the core resolution itself. So this can help know the latitude of the eruption, as long as atmospheric circulation patterns remain consistent for a certain period, which seems to change over time. However, I’m still not too sure how reliable this is, and if it is, it’s probably only reliable for most of the time in the SPC14 core, which captures very well the shape of the sulfate signals.

Nitrate drops: Every clear sulfate spike I’ve seen in the SPC14 for the 1000-1900 AD period has a clear, and rather surprising, drop in nitrate concentration (another atmospheric aerosol). Not only that, but this seems to be a clue that can distinguish submarine from subaerial eruptions. I noticed that, in the South Pole, the sulfate signal of the 1815 Tambora eruption (subaerial) is more than twice that of Krakatau in 1883 (submarine), yet the Krakatau eruption has a nitrate drop that is twice the size of Tambora. This is promising in that there might be a way to tell apart submarine eruptions by their large nitrate throughs, but why does this happen? Well, I’ve only noticed this chemical signature in the South Pole. I’m pretty sure these drops are not present in Greenland, and if they’re present in the WAIS core near the coast of Antarctica, they’re too weak to distinguish them from the abundant noise of background nitrate variations that this core has. Nitrate can be broken down by UV light through photolysis, and volcanic eruptions have been speculated to reduce stratospheric ozone due to the injections of volcanic chlorine (which initiates reactions that deplete ozone). So I think that volcanic eruptions (more specifically, their chlorine) can produce or enlarge an ozone hole over Antarctica, which in turn destroys nitrate, causing the evident drops in its concentration on the South Pole core. The chlorine gas itself never seems to make it into the ice, even though magma has about as much chlorine as it has sulfur, probably because it remains in gaseous form after being ejected. When lava vaporizes seawater, it produces plumes of hydrogen chloride gas, potentially injecting massive amounts of chlorine into the stratosphere, maybe explaining the large low nitrate anomaly of Krakatau, and maybe being a reliable sign of submarine caldera eruptions. The subaerial Samalas eruption also has a large nitrate drop, though, maybe due to its enormous size, and possibly high magma chlorine content. Water in volcanic plumes could also have something to do with this, though

The 13th century was a very volcanically active century, characterized by the large nitrate drops of some of the volcanic eruptions, with probably as many as three major submarine eruptions this century. Blue is the sulfate concentration and orange the nitrate concentration in the SPC14.

With this said, I will try to link volcanic sulfate spikes to their origin eruptions. Originally, I was going to do a list of eruptions from 1000 AD to 1900 AD, but, because it was growing too long, I had to split the article, and I will only do the eruptions since the year 1500 to 1900.

The eruptions I will list below all have sulfate spike durations of more than 2.5 years (except for two that are just below this duration, but I considered them significant enough to include). In Greenland, they can be correlated between multiple cores, and in Antarctica between the WAIS and SPC14 cores, or have very clear nitrate drops in SPC14. The date is from the WAIS core, which likely has a better annual counting chronology than SPC14. I consider any spike where nitrate reduction (ppb) multiplied by time (years) gives >30 to be a large nitrate drop. I also give the estimated stratospheric sulfur injection from the Holocene volcanic sulfate compilation by Michael Sigl et al., 2020, in Tg (millions of tonnes).

Shorter than 2.5-year volcanic signals are also found in ice cores, particularly in Greenland. Some of these come from explosive eruptions in Iceland, from Oreafajökull and maybe Hekla. Others likely originate from VEI 5 eruptions, for example, spikes in 1646 and 1762 likely match large, deadly explosive eruptions of Makian volcano (Indonesia), which, to have been recorded all the way in Greenland, must have been larger than previously thought, maybe VEI 5 events. So are also the VEI 5 eruptions of Taal (Philippines) in 1754 and Tarumae (Japan) in 1667 found in Greenland ice. Effusive Icelandic eruptions can also be seen in Greenland, for example, those of 1477 and 1783 correspond to the fissure eruptions of Veidivötn and Lakagigar, respectively. In fact, Lakagigar is the strongest amplitude sulfate spike in Greenland for the past 1000 years, but it only lasts for ~1.5 years. A different case is that of a well-defined 5-year Greenland sulfate deposition with a distinctive plateau shape, from 1115 to 1121 AD, which I think is probably the Hallmundarhraun shield eruption, a 9 km3 eruption that took place near Langjökull. If so, the eruption rates during this shield eruption were much higher than I expected, 60 m3/s on average, making monogenetic shield eruptions in general more impactful in terms of gas emissions than I thought, but still an order of magnitude below Lakagigar in terms of intensity. These short spikes are not included since it’s hard to identify them and are probably mostly weaker eruptions from mid-high latitudes in NH.

Largest explosive eruptions 1500-1900 AD

1883, Krakatau (Indonesia), 9.1 Tg of S.

At 19 km3 DRE (dense rock equivalent) and 48 km3 bulk volume of dacite, the 1883 Krakatau eruption is the largest volcanic event for which we have detailed historical records. Despite the size, the sulfate spike was quite modest, so, with the majority of the eruption likely taking place underwater, the sea may have absorbed a substantial portion of the volcanic SO2. The large nitrate drop in SPC14 is perhaps a result of HCl gas produced when seawater was vaporized by the eruption (or alternatively, the water plume).

The eruption started on May 20 with subplinian/plinian activity followed by the effusion of an obsidian lava dome from Perboewatan volcano. Activity calmed down until late June, when the eruption intensified and two vents had become active. By August 11, there were 3 major ash plumes rising from Perboewatan and Danan cones, and many other locations were issuing steam. Just after midday, Aug 26, a major plinian eruption started with ash quickly rising to 27 km in height. During the next morning, strong detonations cracked walls, blew away plaster, and shattered glass in places like Bogor or Jakarta, 160 km from the volcano. Artillery-like sounds were so frequent that almost no one in western Java could sleep that night. At 10 AM came the last and most destructive event, probably the ignimbrite phase of the eruption, when a powerful explosion generated a pressure wave recorded all across the globe, multiple times, as it circled the planet. Around 10 AM, pyroclastic density currents travelled 40 km across the Sunda Strait, killing 2000 people on the nearest coasts of Sumatra (even 400 m above sea level), and a tsunami, likely triggered by the PDCs displacing the sea, devastated the Sunda Strait with run-ups of 50 m. The 10 AM explosion quickly formed a massive umbrella that, within 15 min had darkened the sky in Bogor (160 km upwind), forcing people to light up lamps, which was later followed by the fall of accretionary lapilli, a sort of volcanic hail formed from concentric layers of ash.

Photo of Perboewatan cone (northern end of Krakatau island) erupting on 27 May 1883 (from Volcanological Survey of Indonesia, 1883).

1835, Cosigüina (Nicaragua), 5.6 Tg of S.

The Cosigüina eruption in 1835 is another powerful historically recorded eruption. It’s visible as a small-sized sulfate peak and small nitrate drop in the SPC14 core. It started on January 20 and climaxed on the 23rd. Explosions were heard 1650 km away in Jamaica, Venezuela, and Colombia. It’s estimated to have erupted 6 km3 of material, and the caldera size would suggest about 2-4 km3 DRE. The eruption started with dacite and andesite magma with less than 10 % phenocrystals, followed by andesite with medium-high crystal contents (15-30 %).

A description of the eruption from Isla del Tigre (30 km away):

“Having gone out to investigate the cause, we saw with admiration that from the coast of Chinandega a mass rose, whose
beautiful and enormous configuration amused and frightened us at the same time (…) We had a scant light that began at four in the afternoon and ended at six, but while the rain calmed, thunder, and tremors continued. Until two in the morning of the twenty-third, when suddenly we heard a rumble so enormous that it cannot be compared to anything, followed by a very loud noise similar to the rushing of a great river as it flows between crags and rocks.”

Cosigüina. Photo by Jamie Incer: https://volcano.si.edu/gallery/ShowImage.cfm?photo=GVP-04083

1831, Zavaritskii (Kuril Volcanic Arc, Russia), 18 Tg of S

This was a mystery eruption until recently; it was surprising that such a recent colossal eruption hadn’t been found among historical records. Turns out the answer is that the eruption came from a remote uninhabited island of the Kuriles. A recent article (William Hutchison et al. 2024) has matched the spike to the Zavaritskii volcano in the Kuril Volcanic Arc through radiocarbon dating and identification of tephra in the ice. Due to its northern location (46º north), this event is not visible in the SPC14 core, at least not clearly, though it might be visible in other Antarctic cores. The caldera size, the smallest of three calderas nested within each other, suggests around 3.5-7 km3 DRE (VEI 6 likely). The melt (glass) composition of the eruption is andesite-dacite.

Zavaritskii in Google Earth

1815, Tambora (Indonesia), 37 Tg of S.

The Tambora eruption of 1815 is the second-largest sulfate spike of the last 1000 years in the SPC14 core (which may have particularly good quality information, given the lack of noise), also being a medium-sized nitrate drop. In this core, the sulfate signal is nearly identical in size to the slightly larger 1257 spike, and also very similar to the 1457-58 one, the three of them being almost twice as large as the next largest sulfate peak of the past 1000 years. Tambora is estimated to have erupted 40 km3 DRE, and 110 km3 bulk. The magma erupted was between trachyandesite and tephriphonolite composition, with usually under 15% crystal content, which is a very unusual strongly alkaline chemistry for a volcanic arc. With only 55 to 57 wt% SiO2, the magma may have been relatively fluid too.

Some minor eruptive activity had already happened in 1812, then on 5 April 1815, a major plinian eruption occurred, explosions were heard to 1400 km, and weaker activity continued until the main eruption of April 10, when the detonations were in turn heard as far as 3300 km away, in Thailand.

Below is an account of the April 10 eruption from Sanggar (distant 35 km from Tambora). The whirlwind described was likely generated by the hot co-ignimbrite cloud rising into the air. Sanggar was 5 km away from the edge of the ignimbrites but was probably affected by the air currents generated when the hot PDCs mixed with the surrounding air, maybe even firenados:

“About seven, P.M. on the 10th of April, three distinct columns of flame burst forth, near the top of Tambora mountain, all of them apparently within the verge of the crater; and after ascending separately to a very great height, their tops united in the air in a troubled confused manner. In a short time the whole mountain next Sanngar appeared like a body of liquid fire extending itself in every direction.

The fire and columns of flame continued to rage with unabated fury, until the darkness caused by the quantity of falling matter obscured it at about eight, P.M. Stones at this time fell very thick at Sanngar ; some of them as large as two fists, but generally not larger than walnuts. Between nine and ten, P.M. ashes began to fall, and soon after a violent whirlwind ensued, which blew down nearly every house in the village of Sanngar, carrying the tops and light parts along with it. In the part of Sanggar adjoining Tambora, its effects were much more violent, tearing up by the roots the largest trees, and carrying them into the air, together with men, houses, cattle, and whatever else came within its influence (this will account for the immense number of floating trees seen at sea). The sea rose nearly twelve feet higher than it had ever been known to be before, and completely spoiled the only small spots of rice-lands in Sanngar, sweeping away houses and everything within its reach.

The whirlwind lasted about an hour. No explosions were heard till the whirlwind had ceased, at about eleven, A.M. From midnight till the evening of the 11th, they continued without intermission; after that, their violence moderated, and they were only heard at intervals; but the explosions did not cease entirely until the 15th of July”

The description seems to speak of an eruption coming from multiple sides of the caldera in a two-hour-long plinian, with pyroclastic density currents sweeping down the mountain and a rain of lapilli. Then maybe 1 hour of transition when coignimbrite ash from increasingly large PDCs started to rain and 1 hour of a climactic ignimbrite phase when PDCs poured into the sea around the peninsula, carrying with them the trees that covered the flanks of Tambora, generating tsunamis and firenadoes/hurricane winds, and during which the vast amount of pyroclastic material in the atmosphere around the volcano likely absorbed the sound of the explosions.

Mount Tambora

1808, Agrigan (Mariana Islands, US)? 17 Tg of S

Paradoxically, this eruption is quite recent but has proven the hardest to assign. The peak was just under 2.5 years in duration in the SPC14 core, but I’ve included it anyway, given the large size. There are several “clues” regarding this eruption. First, this spike, despite being the 8th largest of the 1000-1900 AD period in the SPC14 core, has only a medium-sized nitrate drop. Second is that we don’t know where it happened, which is quite difficult for an eruption as recent as 1808, and this means the location was extremely remote. Third, the aerosol cloud was first spotted on December 11  in Colombia, which suggested a date of 4 December plus/minus 7 days for the eruption, and a location in the tropics (also supported by the sulfate being present roughly equally in both poles). And last, there was a small tsunami in Japan at 2 AM in the morning of December 4 that struck the Kii Peninsula and southern coast of Shikoku (thanks to questionable in Discord for digging up this previously overlooked tsunami). It was small, 2-3 m run-ups, there’s no known earthquake that caused it, and, given the date, it was most likely a volcanic tsunami related to the 1808 event (either a meteotsunami from a pressure wave or one caused by PDCs displacing sea water).

Agrigan as seen in Google Earth.

 

1694, Long Island (Papua New Guinea), 18 Tg of S

The 1694 event is the 6th largest sulfate spike, and a medium-sized nitrate drop, in the SPC14 core, it also stands out because the sulfate peak is remarkably long-lived in both the WAIS and SPC14 cores, lasting 4.4 years. There have been previous suggestions that this spike corresponds to the Tibito Tephra of Long Island.

Oral histories in Papua New Guinea speak of a “taim tudak” (time of darkness) when sand fell from the sky, crops were ruined, a few houses collapsed and people died. The tephra layer responsible for this time of darkness has been called the Tibito Tephra, and is a >10 km3 tephra airfall deposit originating from Long Island. There is also an ignimbrite in the island (the Matapun Beds) formed during this eruption, of unknown volume, but that mostly must have ended underwater. The melt (glass) composition of the Tibito Tephra is basaltic andesite with 55-57 wt% SiO2, which is quite mafic.

There is a problem with the radiocarbon dates (of carbonized logs in the ignimbrite), the average of which give a most likely calibrated range of 1631-1684 AD, but the only spike here, 1641, is already taken by a known historical eruption, and with no clear SPC14 volcanic spikes in the 18th century and the ones around 1580-1600 taken by known historical eruptions, that leaves only two events, 1620-21 and 1694 close to the radiocarbon determination. I think that the 1620 spike (one of the smallest clear volcanic signals in SPC14) is too tiny for Long Island, however, and also that radiocarbon ages overestimating the age is more likely than them underestimating the age. The overestimation could be due to most of the dated logs having carbon that is a few decades older. I cannot fully rule out Long Island being the 1808 event, given that the radiocarbon ages also graze the calibration curve around 1800 AD, and when using the south hemisphere calibration curve, SHCal20, the 1808 spike is actually closer to the radiocarbon ages (they are slightly closer to 1694 if using the north hemisphere curve), and also because there are oral stories that given the number of generations counted since the eruption suggest that it happened in the 18th century. But there is paleomagnetic evidence in lake sediments that suggest the Tibito Tephra is pre-1700 AD and historical evidence that if it was as late as 1808 the eruption changes would have been noticed by passing ships, which seems to rule out 1808.

Overall, 1694 seems like a good fitting spike. The collapse was probably relatively shallow, around 400 m deep, taking the lowest part of the rim and assuming no filling since, but this is still 30 km3 DRE given the enormous area of the caldera, so 1694 being one of the larger signals would fit the Long Island eruption. And still the volcano might have needed to be relatively degassed (maybe due to being a mafic volcano that might release gas from the magma more easily). The basaltic andesite melt I think, also explains the length of the sulfur spike from strong buoyant plumes, fed by hot mafic magma during the ignimbrite phase, reaching particularly high into the atmosphere.

There is one last problem, and it’s that there is a description of flourishing trees in 1700 AD:

“The 31st in the Forenoon we shot in between 2 Islands, lying about 4 Leagues asunder; with intention to pass between them. The southernmost is a long Island, with a high hill at each End; this I named Long Island. The Northernmost is a round high Island towering up with several heads or tops, something resembling a crown; This I named Crown-Isle, from its form. Both these islands appeared very pleasant, having spots of green savannahs mixt among the Wood-land: The trees appeared very green and flourishing, and some of them looked white and full of Blossoms.”

I should say however, that this description refers to the NW (the direction of Crown Island), and, on this side, Long Island has a tall mountain ridge, Mt Reaumur, that towers nearly 1 km over the lowest parts of the caldera rim and probably protected some of the forests on this flank. Also, maybe the areas of savannah refer to parts of the islands devastated by pyroclastic flows that, at the time, six years after the eruption, were only covered with green grass; otherwise, the land in this part of the world seems to be always covered in lush rainforest.

View of the last ignimbrite of Long Island, the Matapun Beds, beginning about 4 m above the geologists. Long Island had a major collapse shortly before the start of the Holocene, and three smaller caldera-forming eruptions during the Holocene, it’s a mafic caldera system that often erupts from two submarine rift zones extending north and south of the island. Photo by Russell Blong: https://volcano.si.edu/gallery/ShowImage.cfm?photo=GVP-00454

1641, Mélébingóy (Philippines), 22 Tg of S

The 1641 event is a recorded eruption of the Mélébingóy (Parker) volcano in the Philippines. Both the sulfate spike and nitrate drop are small in the SPC14 core. The eruption formed vast aprons of pyroclastic flows and associated lahars reaching more than 20 km from the volcano. Judging from the caldera size, the event was likely around 3-5.5 km3 DRE. The erupted lava is crystal-rich dacite, with 30-40 wt% phenocrysts, which is the same magma as the Pinatubo volcano, and overall the volcano looks very similar to Pinatubo. Description from Zamboanga (300 km away from Parker):

“On January 3rd, at 7 PM we suddenly heard some noise about 1/2 league distant, which created some concern in Zamboanga. It sounded like musketry and artillery… On the following day, January 4th by 9 AM the (supposed) artillery fire increased to such an extent that it was feared that the squadron (sent to investigate) might have run into some Dutch galleons. It lasted for about half an hour. People soon became convinced, however, that the noise originated from a volcano which had opened up, because by noon, we saw a great darkness approaching from the south which gradually spread over the entire hemisphere… By 1 PM we found ourselves in total night and at 2 PM in such profound darkness that we could not see our hands before our eyes.”

During the climactic phase of the eruption, explosions were heard as far as 2000 km away, in Cambodia.

Caldera of Mélébingóy filled with a lake. By Noriah Jane Lambayan (https://commons.wikimedia.org/wiki/File:Lake_Holon.jpg)

1620-21, Billy Mitchell (Papua New Guinea)?, 1.8 Tg of S

The 1620-21 signal shows in the SPC14 core as a sulfate peak and nitrate low, both small, and is probably tropical, given the time it takes to peak. The radiocarbon uncalibrated age given by GVP for the youngest Billy Mitchell eruption calibrates to the period 1455-1629 AD. It was a large dacitic eruption with tephra fall and ignimbrite deposits, estimated at 13.5 km3. The caldera suggests 2.2-4.5 km3 DRE, which in its upper range fits the bulk volume. Because I haven’t seen any other “available” spikes in this period, Billy Mitchell is probably the 1621 signal. There is a supposedly larger sulfur yield from a 1585 signal, but because this signal is 20 times stronger in Greenland than Antarctica (I haven’t found it to be a clear event in the SPC14 core either), and it’s a bit short-lived (2.2 years), the 1585 eruption is more likely to come from some northern hemisphere mid-latitude VEI 5; perhaps the Half Cone eruption of Aniakchak volcano, which happened around this time.

The Billy Mitchell volcano, with Bagana (right) and Reini (left) in the background. https://volcano.si.edu/volcano.cfm?vn=255020

1600, Huaynaputina (Peru), 17 Tg of S

The recorded 1600 Huaynaputina eruption shows as medium-sized sulfate and nitrate signals in the south pole. The eruption lasted from February 19 to March 20 with multiple phases of plinian eruption, vulcanian explosions, and pyroclastic density currents. Estimates vary from 4.6 to 11 km3 DRE and 11 to 26 km3 bulk volume. The eruption involved dacite magmas first with medium crystal contents (17-20 %), followed by higher crystal contents (25-35 %).

Looking into the crater. Source: http://ovs.igp.gob.pe/volcan-huaynaputina

1593, Raung (Indonesia), 9.5 Tg of S

A volcanic signal in 1593 shows in the SPC14 core as medium-sized sulfate and nitrate signals. It’s also a remarkably long-lived sulfate peak lasting ~4 years. A massive 1583 eruption of Mount Raung is described by Portuguese chronicles, though it has not received proper attention until quite recently (Firman Sauqi Nur Sabila et al. 2024). The eruption started with a small amount of dacite and andesite, followed by the main phase of the eruption, which produced basaltic-trachyandesite with pyroclastic density currents that travelled 30 km from the volcano. Brick houses have been found buried in the scoria deposits from the eruption, and it is thought that this event may have wiped out several settlements to the west of Raung. The summit caldera likely formed at this time and suggests a volume of 1.5-3 km3 DRE, while the mafic composition of most of the eruption may have generated a very strong thermal updraft high up into the stratosphere or even mesosphere that explains the unusually long 4-year duration of the volcanic sulfate in the atmosphere.

A description from Panarukan, 50 km north of Raung:

“In 1593, a terrible event occurred, which is worth remembering. The incident occurred at the top of a mountain that contains sulfur. The mountain exploded with such a huge roar that it really scared everyone in Panarukan City. For eight days, nothing was seen except roars and fiery fountains erupting from the top of the mountain, which is called ‘Mount of Panarukan’, and it rained for those eight days so much ash—I mean, so much ash fell from the air, blanketing the fields, streets, squares, public places, and roofs of houses, so covered with ashes that we can’t even walk through them. People saw that the streets were very dark because the air was filled with wisps of gray clouds that really looked like night”

The majestic silhouette of Raung volcano towering more than 3 km above the surrounding plain, with a 2 km wide caldera perched on top. Photo by Lee Siebert: https://volcano.si.edu/gallery/ShowImage.cfm?photo=GVP-01146

To be continued…

References:

Longpré, M., Stix, J., Costa, F., Espinoza, E., & Muñoz, A. (2014). Magmatic processes and associated timescales leading to the January 1835 eruption of Cosigüina volcano, Nicaragua. Journal of Petrology, 55(6), 1173–1201. https://doi.org/10.1093/petrology/egu022

Hutchison, W., Sugden, P., Burke, A., Abbott, P., Ponomareva, V. V., Dirksen, O., Portnyagin, M. V., MacInnes, B., Bourgeois, J., Fitzhugh, B., Verkerk, M., Aubry, T. J., Engwell, S. L., Svensson, A., Chellman, N. J., McConnell, J. R., Davies, S., Sigl, M., & Plunkett, G. (2024). The 1831 CE mystery eruption identified as Zavaritskii caldera, Simushir Island (Kurils). Proceedings of the National Academy of Sciences, 122(1). https://doi.org/10.1073/pnas.2416699122

Gertisser, R., Self, S., Thomas, L. E., Handley, H. K., Van Calsteren, P., & Wolff, J. A. (2011). Processes and timescales of magma genesis and differentiation leading to the Great Tambora eruption in 1815. Journal of Petrology, 53(2), 271–297. https://doi.org/10.1093/petrology/egr062

Guevara-Murua, A., Williams, C. A., Hendy, E. J., Rust, A. C., & Cashman, K. V. (2014). Observations of a stratospheric aerosol veil; from a tropical volcanic eruption in December 1808: is this the Unknown ∼ 1809 eruption? Climate of the Past, 10(5), 1707–1722. https://doi.org/10.5194/cp-10-1707-2014

Tsuji, Y. (2025). Records of a small tsunami observed on the south coasts of Kii Peninsula and Shikoku Island in the early morning of 4th December, 1808. Research Report of Tsunami Engineering Vol.42, 69-73.

Firth, C., Blong, R., Haberle, S., Davies, H., Torombe, M., Coulter, S., Wagner, T., Lee, S., Singura, L., & Denham, T. (2024). Holocene tephras in the New Guinea highlands: Explosive volcanism in the Bismarck arc produces chronostratigraphic markers for interdisciplinary study. The Holocene. https://doi.org/10.1177/09596836241297678

Delfin, francisco jr & Martinez, ML & Salonga, ND & Bayon, Francis Edward & Trimble, D & Solidum, R & Newhall, Chris. (1997). Geological, 14C, and historical evidence for a 17th century eruption of Mt. Parker, southern Philippines. Journal of the Geological Society of the Philippines. 52. 25-42.

Adams, N., De Silva, S., Self, S., Salas, G., Schubring, S., Permenter, J., & Arbesman, K. (2001). The physical volcanology of the 1600 eruption of Huaynaputina, southern Peru. Bulletin of Volcanology, 62(8), 493–518. https://doi.org/10.1007/s004450000105

Sabila, F. S. N., Abdurrachman, M., Saepuloh, A., Kurniawan, I. A., Baraas, A., Yudiantoro, D. F., & Kusdaryanto, H. (2024). Geology, Archaeology, and Historical Studies of the late 16th Century Plinian Eruption of Raung Volcano: a Potential case for Disaster Geotourism in Ijen UNESCO Global Geopark, East Java, Indonesia. Geosciences, 14(11), 284. https://doi.org/10.3390/geosciences14110284