Volcanoes can hide themselves in clear sight. We may see a rugged hill or a lake, but may not recognize the lurking danger. A volcano that lacks a conical shape can hide its nature from us.
Volcanoes are deceptive by nature. That cone with the steam plume at the top may look friendly, but it has a dark secret. The name was Tambora. The rugged normal-looking hill may not show any indication of a volcanic heart, until it suddenly explodes. It was called Pinatubo. That beautiful circular lake is a popular place to spend the summer, away from the heat of nearby Rome. But how did it form? And that island with frequent small, eminently viewable (from a distance) eruptions suddenly disintegrates and sends a tsunami halfway across the world. No, appearances can deceive. We cannot always trust what our eyes tell us.
Do you believe what you see? The cubes below appear to be moving. They are not.
One volcano in particular is so deceptive that it is called Deception Island. It is near the Antarctic Peninsula, and not easy to reach. On approach it shows as a typical island. After a while a narrow channel appears giving access to the interior. Pass through the channel (beware of a submerged rock in the middle of the channel) and the island disappears, as in a mirage. A sea appears, almost completely surrounded by an irregular steep rim with black beaches. Chinstrap penguins meander on the black sands. The rim created the illusion of a solid island, while hiding the large hole in the centre. It is a very useful hole: the enclosed sea provided a safe harbour for the whalers of centuries past. Remains of their buildings can be seen. They were also attracted by the volcanic activity which warms the soil, very welcome so close to Antarctica. But this island harbour is not all it seems. The 10-km wide ring is the remnant of a massive volcanic eruption which wasn’t all that long ago. There is danger here.
The video below shows it well. It was provided by contributor Virtual, and dates from a 2012 visit.
With illusory islands and deceptive volcanoes, it seemed worth a try writing an illusory post. So I approached ChatGPT: the new chatbot which may put an end to essays as a form of assessment. The bot was set leading questions, in order to get a post on Deception Island. Here is the result. My questions are in italic. The order has been slightly changed.
ChatGPT on Deception
Me: How many active volcanoes are there?
ChatGPT: It is difficult to give an exact number of active volcanoes, as new ones can be discovered and the level of activity of existing ones can change. However, a commonly cited estimate is that there are around 1,500 potentially active volcanoes on Earth. Of these, approximately 50-60 are erupting at any given time.
Can we predict volcanic eruptions?
Volcanic eruptions can be predicted to some extent using various methods such as monitoring changes in seismic activity, ground deformation, gas emissions, and temperature changes. However, the accuracy and lead time of these predictions vary and there is still a significant level of uncertainty involved, especially for less predictable or less monitored volcanoes. It’s important to note that the science of volcano forecasting is still evolving, and the best way to stay safe during an eruption is to stay informed of current conditions and follow evacuation orders from local authorities.
When was Deception Island discovered?
Deception Island, located in the South Shetland Islands of Antarctica, was discovered in 1820 by a group of sealers led by American seal hunter Nathaniel Palmer. (Warning: error of fact.)
Tell me about the bases at Deception Island
Deception Island, located in the South Shetland Islands of Antarctica, was used as a base by various countries in the early 20th century, including Britain, Chile, and Argentina. The first permanent scientific base on the island, called Whalers Bay Base, was established by Britain in 1944. It was used for various scientific and meteorological purposes. However, the base was destroyed in 1967 following a volcanic eruption. Currently, the island is a popular tourist destination, with only a Chilean refuge station, which serves as a research facility, still operating on the island.
Wasn’t there an Argentinian base as well?
Yes, you are correct. There was also an Argentine base, called “Carlini Base”, located on Deception Island. The base was established in 1952 and was primarily used for scientific research and meteorological observations. It was one of several Argentine bases in Antarctica and served as an important hub for the country’s scientific community. (Warning: incomplete fact.)
What kind of science is done on the island?
Deception Island, located in the South Shetland Islands of Antarctica, is primarily used for earth and atmospheric sciences research. The following are some examples of scientific studies that have been conducted on the island:
- Volcanology: Deception Island is an active volcanic caldera, making it an important site for the study of volcanic processes, including magma formation and eruption dynamics.
- Geology: The island’s unique geological features, including its geysers, hot springs, and fumaroles, make it an important location for geological research.
- Meteorology: The island’s location and weather conditions make it an ideal site for meteorological observations, including studies of the atmosphere, oceanography, and climate.
- Biology: Deception Island is also a site for biological research, including studies of the local flora and fauna, as well as their interactions with the changing environment.
These are just a few examples of the scientific research that is conducted on Deception Island. The island remains an important hub for scientific investigation, as it provides a unique opportunity to study the Earth’s systems in a remote and challenging environment.
Do you have a reference for this information?
I’m sorry, but as a language model AI, I don’t have any specific references. However, you can find more information about scientific research on Deception Island through various sources such as scientific journals, research papers, and academic databases. Additionally, information about the history and current status of scientific research on Deception Island can be found on the websites of organizations such as the Antarctic Treaty System and the Council of Managers of National Antarctic Programs.
(So it did actually tell me where the information came from.)
How active is the Deception Island volcano?
Deception Island is a potentially active volcano located in the South Shetland Islands, Antarctica. It has a history of frequent eruptions, with the most recent one occurring in 1970. Since then, there have been numerous seismic events and fumarolic activity observed, indicating that the volcano is still active and capable of erupting again in the future.
The level of activity at Deception Island can vary and is monitored by scientific agencies to assess the risk of future eruptions. At present, the volcano is considered to be at a low level of activity. However, it’s important to note that this can change quickly and unexpectedly, so visitors to the island are advised to follow any safety recommendations from local authorities.
How did the 1970 volcanic eruption of Deception Island occur?
The 1970 eruption of Deception Island was a phreatic (steam-driven) eruption that occurred on the island in Antarctica. It was triggered by the interaction of magmatic heat with shallow water or sediment, causing a steam explosion.
The eruption caused significant damage to scientific stations and facilities on the island, but no human casualties were reported. Ash and pumice deposits were spread over a large area, and the shape of the island was altered by the formation of a new caldera (a large, bowl-shaped depression formed by the collapse of a volcanic cone).
After the 1970 eruption, the island has continued to show signs of volcanic activity, including steam emissions, earthquakes, and changes in sea level. These suggest that Deception Island remains a potentially active volcano, and scientists continue to monitor the island for signs of future eruptions.
How did the 1967 eruption happen?
The 1967 eruption of Deception Island, located in the South Shetland Islands of Antarctica, was caused by a volcanic eruption. The eruption occurred on the 12th of February, 1967 and was one of several eruptions that have taken place on the island throughout its history.
During the 1967 eruption, magma rose to the surface and broke through the caldera, which is a large crater formed by a previous volcanic eruption. The eruption produced ash, steam, and volcanic gases, and also caused significant damage to the British research station located on the island. The station was evacuated, and several buildings were destroyed or severely damaged.
The 1967 eruption was considered a moderate-sized eruption, with an estimated volcanic explosivity index (VEI) of 2. Despite the damage caused to the research station, no casualties were reported.
The Deception Island volcanic system is considered active, and ongoing monitoring is necessary to understand its behavior and potential for future eruptions. The island remains a popular site for scientific research due to its unique geological and biological features, and the lessons that can be learned from its volcanic history.
When did the caldera of deception island form?
The caldera of Deception Island formed during a volcanic eruption that occurred in the late Holocene epoch, around 7,000 years ago. The eruption was one of several large explosive eruptions that have occurred on the island in the last 10,000 years, and it caused the collapse of the central part of the volcanic cone, creating the large, circular depression that is now known as the caldera. (Warning: outdated fact.)
The caldera of Deception Island is now filled with seawater, forming a sheltered bay that is known as Port Foster. The caldera and its surrounding areas are a popular destination for tourists and scientific researchers, who come to study the island’s unique geology and wildlife, including penguins, seals, and whales.
How often has Deception Island erupted?
That model is currently overloaded with other requests.
How often has Deception Island erupted?
Deception Island is a potentially active volcano located in the South Shetland Islands, Antarctica. It has a history of frequent eruptions, with several known eruptions occurring in the past 10,000 years. The most recent eruption at Deception Island took place in 1970, when a phreatic (steam-driven) eruption caused significant damage to scientific stations and facilities on the island, but no human casualties were reported.
Since then, the island has shown signs of volcanic activity, including steam emissions, earthquakes, and changes in sea level, suggesting that it remains a potentially active volcano. The frequency of eruptions at Deception Island is difficult to predict, as it is influenced by a complex interplay of geological and environmental factors.
Scientific monitoring of Deception Island is ongoing, and scientists are using a variety of techniques, such as seismology, gas monitoring, and satellite imagery, to track changes in the level of activity and assess the risk of future eruptions.
I decided that ChatGPT writes well and gives relevant information but it has a tendency to digress, uses stock phrases, suffers from repetition and inaccuracies. The content is variable, sometimes quite good (as in the overview of science), sometimes more superficial and on a number of occasions wrong. Palmer did not discover Deception Island: he was the first to visit it. The British base was destroyed in 1969, not 1967. The Argentinian base it lists was on nearby King George island. Their base on Deception Island was a different one and was called Deception Station. (The bot didn’t actually say it was on the island!) The inaccuracies reflect where it gets its information from: as it told me, it is a language model, not a science database. You get what is available. It will produce mistakes if that is what it finds or if it doesn’t fully understand what is written. A similar experiment asking physics questions (that wasn’t me) gave very dubious results. I noticed it prioritized the British base over the Argentinian or Chilean one, suggesting the language model takes input mainly (or only) from english-language sites, and it preferred the American discovery of Deception Island over the earlier British one.
ChatGPT is probably best used as as a front-end to a search engine or to wikipedia. Rather than giving a list of web sites, it presents the relevant material in a clear way. The writing is very good but it is obvious there is no real understanding of the material behind it: the critical analysis is a task for the reader. In fact it told me so! It will need a fact checker attachment. The stock phrases give the impression of knowledge without understanding. In a student essay such phrases would be guaranteed to lead to questions about their meaning, in order to test the underlying understanding.
ChatGPT would be ideal for writing a strategy document: these typically consist of stock phrases, are repetitive and lack meaning. A basic popular website on volcanoes might also work but you won’t get more than what is commonly known. It is wisywig: what you see it what you get. Writing a post for VC will take more training for the bot. But I would not rule it out.
(I can see why google is worried. This bot could pull the rug from underneath their income stream. I would expect google to respond by inserting adverts into the information stream in such a way that the bot does not recognize them as such. Once that happens, ChatGPT could include answers such as: ‘ Excursions are very popular and can be booked at …’ . My prediction for the future of artificial intelligence is that there will be a battle between truth and adverts. VC favours the former.)
Let’s get back to the real world. Deception Island measures 14 by 13 km. It is the tip of a mountain which measures 25 kilometers across: it is a substantial volcano. The central bay which takes up the interior of the island is called Port Foster. The bay is 10 by 7 km wide; the entrance channel is only 500 meters across. The highest points on the island are on the east (Mount Pond) and southwest (Mount Kirkwood), at 550 and 450 meters. Glaciers have formed on these hills. Elsewhere the island is ice-free but often snow covered. Temperatures are cold but moderately so, a bit below freezing in winter, barely above in summer, with little change between day and night. Port Foster has a depth of 180 meter at the centre, but there is a shallower platform around the coast which is less than 50 meters deep.
Everything on the island is volcanic. The composition is basaltic-andesitic. Dikes (there are not many) are mainly basaltic. There are active fumaroles and hot ponds. There are also many maars, remnants of phreatic explosions, often at the edge of Port Foster.
There are ruins of Norwegian whaling stations on Deception Island; these were occupied intermittently from 1911 to 1929. They were abandoned when whaling no longer required shore facilities. The British scientific base was opened in 1944, the Argentinian one in 1948 and Chile followed in 1955. They were abandoned after the eruption in 1967.
Faults of deception
The tectonics here are complicated. Deception Island is located in between the Antarctic peninsula and the South Shetland Islands. The strait between them is called the Bransfield Strait. The Strait is around a million years old and formed as a back-arc basin. The arc is the volcanic arc of the South Shetland Islands, and the cause was the subduction of the Phoenix plate, a small plate in between the Antarctic plate and the Scota plate. This subduction pulled down the back-arc. But the subduction ceased some three million years ago when the Phoenix plate stopped spreading. The Phoenix is now dead. In response, a new spreading centre formed in the Bransfield Strait, perhaps because of the relaxation after the Phoenix died. The South Shetland Islands now lie on their own microplate. Deception Island lies exactly on the new spreading centre that separates this microplate from the Antarctic peninsula.
The region is fault ridden. A belt of faults runs southwest to northeast underneath the island, forming a half graben. The few dikes that have been found on Deception Island are mainly aligned with this SW-NE fault which appears an axis of extension. A significant fault (called the Maraconi fault) runs SSE-NNW along the side of the island, causing the straight cliff that forms the western coast of Deception Island.
The island was shaped by a caldera forming eruption. After this, Deception Island did not go quietly into the dark. It continued to be a very active volcano. The fumaroles and geysers have been present ever since the discovery and presumably much longer. An eruption was reported in 1842 from the slopes of Mount Kirkwood. This eruption was discredited by later reports, which argued there had been no other evidence of eruptions. However, debris from this eruption was later found in the glacier ice showing that the doubters were wrong. But this was only proven after the events of 1967.
The island first proved its volcanic doubters wrong on 4 December, 1967. A pyroclastic eruption occurred in Telefon Bay, in 23-m deep water. After three days the eruption ceased: by that time the new island was 62 meters high, 1 kilometer long and 350 meters wide. There were three craters on the island, aligned northeast to southwest. This orientation suggests a dike was involved, as this is the axis of least resistance due to the graben. The new island quickly began to erode on the southeast side facing Port Foster, and the craters became water filled. The opposite side of the new island was better protected.
A second eruption occurred about 2 kilometer to the east. This happened on the coast of Port Foster where a new bay was cut in the coast here. There may also have been a submarine eruption in the same region. A damaging debris flood came downhill just north of the eruption. This may have been ice melting from a third eruption point.
The eruption lasted only three days. But a little over a year later, on 21 February 1969, strong tremors appeared. There were five people present on the island at the time. They were from the British Antarctic Survey and were staying at the British station at Whalers Bay on the east side of the island. The five men fled the station huts but while trying to reach an evacuation point, an eruption began at Pendulum Cove on the side of Mount Pond, near to where they were. Two eruption plumes were seen and rocks began to rain down. The men sheltered underneath corrugated iron sheets from the abandoned whaling huts. Carrying these sheets above their heads, they went back to the station for better shelter. There they discovered that a debris flood carrying large ice blocks had swept through the site in their absence. It was a lucky escape. Helicopters from a Chilean vessel rescued the men later that day. This was the same vessel that had evacuated the bases on 5 December 1967! This was the time the base was actually destroyed, not 1967. (ChatGPT, take note.)
The eruption had occurred along a 5-km fissure with 20 separate eruption vents. This had triggered the debris flood, which reached 15 meter high in places. Both the Chilean and British base were badly damaged.
The 12 August 1970 eruption happened in the same location as in 1967 but it was much larger. It was not directly observed: no one was present as it was still winter, a good time not to be there. There was an earthquake at around 5.40pm on that day which may have been the onset or the run-up of the eruption. The next morning, ash fell on nearby islands. It was later found that the eruption had produced 13 vents on land and in shallow water. Some of the vents coincided with the vents from 1967 and had enlarged those craters. The eruption plume was 10 kilometer high. The eruption likely reached VEI 3.
It is notable that all three eruptions came from linear fissures. Even though the eruptions were along the inner edge of the caldera, it does not appear that a ring fault was involved.
Other recent eruptions
The ice on the island contains a record of the recent eruptions: these produced layers of ash. Between 1780 and 1972, there are indications for 23 separate explosive eruptions. There may have been twice as many: in summer ash is blown around by the wind, and it may be hard to distinguish these from new eruptions. In winter the island is covered in snow and all ash is fresh. The oldest reported eruption was in 1842, when ‘a wall of fire’ was seen on the side of Mount Kirkwood. The ice shows that this may have been the largest eruption in 350 years. Ice deposits also show eruptions in 1912 and 1917 which were not recorded.
What will happen next? It is plausible that the next eruption will be in the northeast of Port Foster. The sea floor here is showing inflation, with a 5 meter rise between 1990 and 1999. We’ll see.
Caldera formation: Deception in the making
The faulting in the region has led to the suggestion that the caldera collapse was caused by the faults, and was largely a tectonic affair. The opposing view is that the crater formed in a large eruption driven from a shallow magma chamber.
The evidence points at the latter. The eruptive layers are exposed best in the coastal cliff but are also seen elsewhere. The bottom of the stratigraphy shows a basaltic shield. These were often strombolian eruptions which build up the large edifice. Above this is the so-called ‘Outer Coast Tuff Formation’ consisting of yellow tuff. It is present everywhere on the island with a very uniform thickness of around 45-50 meters on the coast and 90 meters at the centre. It shows individual layers of some 10-15 meters thick. Above that are the more recent ejecta. These are different in different parts of the island, as expected from small individual vents, but are typical of similar thickness to the Tuff layer.
The yellow tuff clearly comes from a major explosive eruption, and it is considered as the origin of the caldera. It consists of ignimbrite and co-ignimbrite.
The internal strcuture of the caldera has been measured with seismic waves. This has shown a staggering of 1.2 km depth of infill by pyroclastic materials similar to (and probably identical with) the yellow tuff. Above that is up to 100 meters of post-caldera eruptions. The caldera once was much deeper than it is now.
The caldera cuts the yellow tuff. However, it also cuts the early post-caldera material. This suggests that the current caldera did not all form in the original eruption. Part of the caldera formed later, either as a true collapse or by erosion. The original crater was smaller than the current size.
One can estimate the original size by assuming that the later volcanic vents are located on the edge of the original crater, which is where the structural weakness is. This gives a size of 50km2, versus 82 km2 of the current caldera. It has an irregular shape. The edges of the original caldera seem to coincide with structures or faults that are still visible in the landscape. Thus, although the caldera formed in a large explosion, the shape of the hole was largely determined by pre-existing faults. There was no obvious ring fault. It is perhaps not surprising that a volcano located in such a faulty region will be affected by those faults.
How large was the eruption? The size of the hole may have been as much as 89km3. Of this 60km3 fell back into the crater, albeit as lower density material. The remainder was thrown out. So the eruption was at least 29km3 DRE. This makes it a high end VEI-6 at minimum.
What caused the eruption? The yellow tuff is of uniform composition, but also contains basaltic inclusions. This may suggest that the eruption was triggered by fresh basalt entering and destabilizing a more evolved magma chamber. If this is the case, then the eruption happened fairly quickly after the intrusion as there is no evidence for mixed magmas, only the two separate magmas. The composition suggests that the magma was stored at around 10 km deep, well below the current caldera. The magma chamber was not particularly shallow. The magma moved up very quickly from this depth.
Some of the yellow tuff shows effects from water while other doesn’t. This suggests that water may have been involved with part of the eruption. A natural suggestion is that the eruption started on dry land, but that sea water intruded once the vent had dropped below sea level. This could make the eruption similar to Krakatoa – or Hunga Tonga. It is quite speculative, but whenever a large eruption takes place on an oceanic island not far from sea level, the influence of sea water must be considered.
When did the eruption happen? This has only recently been discovered. It had not been conclusively identified with a dust layer in Antarctic ice cores. It could have been in the late Pleistocene (ice age) or the early Holocene. The co-ignimbrite shows that the eruption was above sea level. The nature of the eruption may suggest that the sea level was similar to current levels. However, we don’t know that for certain. But in 2018, three tephra layers were found in lakes on Livingstone island, about 40 km from Deception Island (Antoniades et al. Science Reports, 8: 17279). The tephra had composition similar to Deception Island, and the largest layer showed the presence of two separate magma types, as was seen in the yellow tuff. This layer was carbon dated to 3860-4110 BP. The other two tephra layers are younger, at 2370 ± 100 and 1890 ± 140 BP, and were much smaller eruptions.
Thus, the caldera-forming eruption appears to be only 4000 years old. ChatGPT thought it was 7000 years ago. It also suggested that later eruptions were also large. Both claims were inaccurate. The yellow tuff is the only major layer here. Later eruptions were much smaller whilst earlier eruptions were probably not explosive. (In fact, the description of the 1970 eruption by ChatGPT seems a bit dubious. It is a language machine, not a library.)
These new data provide an intriguing piece of information. They show that the tephra was dry: the isotopic ratios were not affected by sea water. Thus, the major eruption took place on land, or before sea water could interact with the magma. That means the caldera had not yet formed. This has similarity with Krakatoa where only the last of the four major explosions on the day of destruction produced wet ash: the first three were dry. This leaves the precise role of sea water unclear. The team finds indications for a large earthquake shortly after the deposition of the ash, and suggest that this was due to the collapse of the caldera. But if that was true, how did 2/3rd of the ejecta end up inside the caldera? One wonders whether the ‘earthquake’ was just the thick tephra blanket in Livingstone Island sliding into the lakes.
Deception Island is a unique place, an active caldera masquerading as a typical antarctic island. It is neither a typical fault-driven collapse nor a typical ring-fault caldera. It was a singular VEI-6 or 7 eruption which was shaped by pre-existing faults.
The volcano sits on top of a spreading rift which was caused by the cessation of the spreading of the neighbouring Phoenix plate. We can speculate that the local oceanic plate began to be pulled towards the stalled edge of the Phoenix and this triggered a new spreading centre to accommodate the pull. Deception Island formed at the end of this spreading centre, in a region of extension, where perhaps heat welled up past the subduction edge. It grew out of the Phoenix plate.
The phoenix is a well-known story which even played a role in Harry Potter. It dies in fire to be resurrected. But it is an illusion. The pPhoenix did not rise from the ashes. Instead, from the ashes of the Phoenix rose Deception Island.
Albert, February 2023