There is a hidden continent on Earth. We have known about it for centuries, first as a ‘Terra Australis’, a large landmass required to balance the world (thank Aristotle for that idea), and later as a land underneath the ice, much smaller than had been assumed and leaving the world severely unbalanced with the southern hemisphere far more oceanic than the north.

After another continent was renamed to ‘Australia’ in the early 1800’s, the mythical southern land needed a new name. Or rather, an old one: before Terra Australis, in medieval times the name ‘Antipodes’ had been in common use, and this could be adapted to the modern world. All seas north of the Antarctic circle had been explored, and the land had not been found. Around 1890 the name ‘Antarctica’ was coined, and it stuck.

The mainland of the continent had been discovered around 1820 (in hindsight). The first landing was in 1853. Antarctica’s most famous volcanoes, Erebus and Terror, were discovered around 1840 and named after the two ships of the Ross expedition. After that, exploration of the new world was glacially slow, mainly because of the limited economic benefit. It was clear that this was a hostile, deeply-frozen place with little to offer. The polar plateau was reached only in 1909 (by Shackleton) and the south pole itself in 1911 (by Amundson). These are two famous names in polar history; while Amundson worked in both poles, Shackleton was a monopole: only Antarctica counted to him. But even their adventures of exploration and competition left most of it a Terra Incognito. Only in 1948 was it shown that West Antarctica and West Antarctica were connected and formed a single land mass.

Very little of Antartica is ice-free. There is some along the Antarctic Peninsula, some dry valleys near the Ross Sea where the wind ablates any snow or ice, a few coastal cliffs and nunataks, tops of mountains sticking out above the ice sheet, and some parts of the Transantarctic Mountains. Everything else is hidden underneath a huge glacier, which forms a large dome reaching 4 km high and stretching 1000 km across over East Antartica. The south pole which was the ultimate prize of the age of exploration lies on this dome. The ice sheet over West Antarctica is much lower.

The two ice sheets of East and West Antartica are separated by the Transantarctic Mountains, 300 km wide and stretching along a length of 3500 km. A second mountain chain runs along the Antarctic Peninsula, culminating in the Vinson Massif, the highest point in Antarctica at 4852 meters; it was first climbed in 1966. Another range runs along the coast at Queens Maud Land, called the Orvin Mountains: these are among the most scenic mountains of the continent

The West Antarctic Rift is an ice-covered basin which runs alongside the Transantarctic Mountains from (and including) the Ross Sea to close to the Weddell Sea – it almost dissects the continent. It is one of the major continental rift systems on Earth, similar in size to the East Africa Rift. The crust has thinned to 20 km in the West Antarctica Rift because of spreading, but no oceanic rift ever developed and the rift now lies quiet under the ice. Mount Erebus lies along this rift system.

Take away the ice, and a very different world appears. This has been mapped using ice-penetrating radar, done from airplane or satellite. The most recent map (2026) shows wide varieties of landscapes. There are wide plateaus, high mountains and deep ravines. Everything is deeply depressed by the weight of the ice, and much of the bedrock now lies below sea level. If the ice were gone, much of Antarctica would be covered by the ocean. But over a few thousand years the crust would slowly rise again, and eventually regain its land.

Along the coast, there are deep fjords, sometimes visible as glaciers between mountains, sometimes visible only to radar. One of those fjords is home to perhaps the most unexpected sight: a fall of red-coloured water, known colloquially as Blood Falls. It lies in-land, some 25 km from the coast at the head of one of the Dry Valleys. A glacier ends at this point, and the melt water has collected in a frozen lake. There is a seasonal waterfall that comes out of the glacier and falls unto the lake. The result is spectacular.

The water comes from a salt water lake, deep underneath the glacier, which formed when sea level was much higher than it is now. The water is draining towards the end point of the glacier, now well in-land. It obtains its colour from a high iron content. (And no, there is currently no plan to rename Blood Falls to America Falls.) (As far as I know.)

We now know there are many lakes underneath Antarctica’s glaciers. The subglacial lakes are mainly found near the coast, and form from melt water which collects at the bottom of the glacier. Some of the lakes are stable and ancient, but others are dynamic and forming outlets. One such lake drained 5 km³ of water in 2007 and 2008, leaving a big depression in the ice. There is a watery world hiding underneath the ice.

Volcanoes

Antarctica is home to one famous volcano: Mount Erebus with its semi-perpetual lava lake. At the time of discovery in 1840, lava was flowing along its flanks. The heat has sculpted the ice on the mountain into spectacular formations.

Deception Island is another volcanic feature, a large caldera with occasional eruptions which have even damaged buildings, and 4000 years ago spread ash more than 4000 km away. Otherwise, it seems a volcanically quiet land, behaving more like Greenland than Iceland.

But there are many more volcanoes here. Along the coastline from the Weddell Sea to the Ross sea lie 7 active and 2 more potentially active volcanoes, which have had eruptions within the holocene. These are James Ross Island, Deception Island, Mount Takahe, Mount Waesche, Mount Berlin, Mount Erebus, Mount Melbourne, Mount Ritmann and the Pleiades Mountains. There are extinct volcanoes as well, notably the large shield volcano of Mount Haddington on the Antarctic Peninsula. There is a strange pillow volcano on the opposite side of Antarctica, the Gaussberg, which is completely isolated from any other known volcano on the continent.

Elsewhere, Buckle Island just off the coast of Victoria Land has had two suspected eruptions in the 19^th century, when dark eruption columns were seen. The Balleny Islands, which include Buckle Island, are indeed volcanic but no indication of volcanic activity such as ground heat has been detected. On the other hand, there are a few small crater rims on the flanks of Mount Brown, Buckle Island, one of which is reportedly ice-free – but not snow-free! The volcanic eruptions here must be considered as possible but unproven.

There are other signs of volcanism. Some are ancient. A spectacular one is the large sill seen in mountains of Southern Victoria Land, 200 meters thick. These are the remnants of the break-up of Gondwana, and they belong to the Ferrar or Karroo flood basalt, at a time when Antarctica had not yet broken away into isolation. Others are recent. Especially in Mary Byrd Land, near the coast, a large number of volcanic cones stick out above the ice. They do not show signs of activity, but there are quite a few of them.

Rift volcanics

These volcanoes in Antarctica all lie around the edge of the continent. But here, the land only shows what it wants to show. What about volcanoes buried under the ice?

The majority of the known active and extinct volcanoes are on the flanks of the West Antarctic Rift. But is this just because these flanks are exposed? In the East Africa Rift, the volcanoes are mainly found along the rift axis, and flank volcanoes are less frequent. Could the same be true in the West Antarctica Rift System? Attempts to look for ‘hills’ in the ice, areas where the ice elevated indicating a buried mountain below it, found 40 such features which were identified as possible volcanoes. Ice-penetrating radar was used to look below the surface. Looking for conical elevations underneath the ice, appearing like stratovolcanoes or shield volcanoes, a total of 178 such features were found along the Rift axis. Many were tall enough to also leave an imprint on the ice sheet. For 138 of the cones, there is sufficient data to consider them as volcanoes. 80% are located along the axis of the Rift.

The density of volcanoes here rivals that in the East African Rift! It is a substantial, world-class volcanic province. It should be noted though that many of these may be extinct. Ice protects against erosion, and the cones may survive here much longer than out in Africa. The volcanic activity may therefore be much less here than that in Africa. But the existence of this volcanic activity does make Mount Erebus seem less of an outlier. It may not even be the tallest: one of the buried discoveries is listed at 3850 meters, just higher than Erebus.

However, where such discoveries are made, science looks at them with a critical eye. If this volcano is so tall, why is it not sticking out above the ice, as neighbouring but smaller volcanoes do? And can we be certain that if the shape looks like a volcano, it is one? A more recent paper raises these concerns, but also confirms 21 of the subglacial volcanoes. I cannot discuss that paper here though, because in their wisdom the authors published behind a paywall as impenetrable as the glaciers of Antarctica. Against the warmth of science, the would-be reader is left out in the cold.

The cones sticking out above the ice show what the lava is like. The volcanism in Antarctica is basaltic to alkaline. It is derived from mantle material, as might be expected in a basin environment. That is helpful, as basalt is black and contrasts nicely with the ice. Light-coloured rhyolite might be hard to see! But this is a moot point for eruptions being buried underneath the ice where only the shape gives an idea of what caused it. Buried cones from individual (monogenetic) eruptions may be similar in shape to Iceland’s tuyas, steep and flat-topped. Multiple eruptions will still produce shields and stratovolcanoes even deep under the glacier, but with a tendency for steep slopes.

Eruptions

Several of the subglacial volcanoes in Antarctica show up not as elevations, but as depressions in the ice. This may indicate ice melt, which means that the volcano is not extinct but generates heat – perhaps eruptions. Mount Berlin shows steaming fumaroles (with Erebus-like ice towers) at its caldera, direct evidence of heat in a world of ice and fire. Investigations on its summit (above the ice) have shown that it had plinian eruptionso 20,000 to 30,000 years ago with 30-km high eruption columns.

Another indicator of eruptions comes from tephra in the ice cores. They show 20 annual layers with tephra from three known volcanoes in Marie Byrd Land: Mount Takahe, Mount Siple and Mount Berlin. The youngest layer is around 7500 years old and comes from Takahe. On the other side of the Ross Sea, Mount Melbourne is a source of thick tephra layers in the Ross Sea, associated with pyroclastic flows. The youngest one is dated (very roughly) to 12,000 years ago. Nearby, the Pleiades Mountains may have erupted as recent as 3000 years ago. There is plenty of life in this frozen world!

Volcanoes can suffer from earthquakes, both when active and when considering whether to become active. The West Antarctica Rift shows some earthquake activity, although it is rather weak. The lack of consistent activity suggests that the rift underneath the ice is not or no longer extending. Two earthquake clusters have been detected. One coincides with the edge of the rift, and may be from a fault line. The other cluster of 6 earthquakes is at the centre of the rift, and the events are typical of magma movement. The cluster is at a depth of 20 km, at the location of a subglacial volcano, Mount Casertz. This volcano also shows an ice depression above it. The glacier surrounding it shows fast movement, which may be due to melt water at the bottom. The combination of these indicators suggests that Mount Casertz is an active volcano, although it is not possible to be certain it has recently been erupting. One to keep an ice-penetrating eye on.

The youngest eruption

It turns out that ice-penetrating radar can not only identify subglacial volcanoes. It can detect tephra as well. A tephra layer in the ice can reflect some of the radar signal and this shows up in the reflected signal as a boundary layer. Such a layer has been detected at the location of Hudson Mountains. It is a bright reflection which is seen over an area of 150 by 190 km, at a depth of a few hundred meters into the ice. The reflection is strongest at the centre and fades towards the outer area. This was published in Nature Geoscience by Hugh Corr and David Vaughan. (The original paper is not open access but the authors have also made it available on researchgate.)

The Hudson Mountains are located in Ellsworth Land (see the map at the top of this post), not far from the coast. They lie within the Bellingshausen volcanic province. The mountains are mostly covered by ice, but a few nunataks stick out and these include a few younger-looking cones. The Hudson Mountains are listed as a volcanic field with 17 stratovolcanoes and parasitic cones, dated to around 5 million years ago. The field is dominated by three stratovolcanoes: Mount Moses, Mount Manthe and Teeters Nunatak; all three are badly eroded.

The GVP mentions a satellite observation of the possible eruption at Webber Nunatak in 1985. This report is unsubstantiated. It is not mentioned in any publication I could find. Recent investigations on this nunatak have not reported any evidence of volcanic ejecta. The nunatak is volcanic but old. There is a seasonal meltwater pond on one side and it is perhaps possible this was (wrongly) taken as evidence for an eruption.

The tephra layer lies at a depth in the ice which corresponds to an age of approximately 2200 years. The age is consistent across the region where the tephra is found. The tephra is missing from the area of the Pine Glacier, but this is understandable because this is a fast flowing glacier. The age corresponds to a date around 200 BC, with an uncertainty of 250 years either way.

The eruption would have penetrated through the ice, ejection ash into the atmosphere and forming the tephra layer through air fall, later buried by snow. It seems likely that there was little wind at the time, as indicated by the elliptical shape of the tephra area. That would favour an eruption in the summer (Nov – Feb) as the winters here are windier.

This region in the highlands of Ellsworth has slow ice movement. The typical flow rate at the surface may be 1 to a few meter per year. Over 2000 years, the ice would have moved by no more than a few kilometers.

The image above shows the measured surface flow rate of ice. A similar figure is presented in the paper. The tephra layer follows the dark blue area where the ice flow velocities are lowest. But further away the ice speeds up considerably and the velocities are measured in meters per day rather than per year. At 1 meter/day, the tephra could have moved by hundreds of kilometers since the eruption, and be lost to the measurements. The elliptical shape and extent of the tephra may just reflect the area where the ice is stable enough to retain this memory.

It is difficult to pin down the precise volcano that eject the tephra. The authors suggest it sits near the centre of the ellipse at a topographic high. They name it HMSV (not the most catching name, one must admit), which stands for Hudson Mountains Subglacial Volcano (which is not much better). The tephra volume is estimated at around 0.1 km³ but with a very large uncertainty (and some of the tephra may be lost.) Think VEI 4, operating very similar to Grimsvotn. This could be the largest Antarctic eruption since Deception Island blew up.

Antarctic ice cores are used to detect sulphate from eruptions around the world. Could they show such a local eruption? The two nearest ice cores were taken 1000 km away, so for a smallish eruption this is not as easy as it sounds. But both ice cores have a strong peak around 320 BC. This peak is not seen at ice cores elsewhere in Antarctica, and it is therefore a candidate for the Hudson Mountains eruption. Full evidence would require tephra and this has not yet been reported from these two ice cores.

The Hudson Mountains are adjacent to the Pine Island Glacier, a fast glacier which is one of two which could destabilize the West Antarctic Ice Sheet. (The other one is Thwaites Glacier) Active volcanism close to these glaciers was not known before. The eruptions themselves won’t damage the glacier. However, melt water from an eruption could cause the Pine Island Glacier to slide even faster. The much larger Thwaites glacier is safe from Hudson Mountains melt.

But perhaps this has already happened. Sea water near the the end of the Pine Glacier shows an excess of 3He, which has to come from melt water underneath the glacier. It indicates that the glacier is receiving this water from a volcanic source underneath the glacier or its tributaries. It requires a volcanic source not dissimilar to Grimsvötn. The Hudson Mountain Volcano would be a plausible source for this.

Past and Future

Once global warming removes the West Antarctic Ice Sheet, a lot of volcanoes will see the light of day. The pressure will be off – but this pressure from the weight of the ice may be suppressing eruptions at the moment. When the ice age ended in the north, Iceland’s volcanoes became very active. When it ends in the south, will the same happen?

Antarctica is in so many way the antipode to the Arctic. Land versus sea, Ice versus water, continent versus ocean. And now we can add volcanoes: whereas the Arctic is a volcanic desert, Antarctica is an oasis. One we didn’t know existed.

Every story about Antarctica has to mention Shackleton. His expeditions became legendary: they were about battles of survival, knowing when to put people before ambition and bring them back safely. At his first run to the South Pole, Shackleton came to within 200 km of the pole (and was the first to reach the high polar plateau) before deciding that supplies were insufficient to go further. At his second run he never even set foot on the continent and lost his ship – and still got everyone of his companions safely home. That is the kind of failure that the English (and perhaps the Irish, as he was of half Irish descent) could value. His stories include a volcano: his men were the first to ascend Erebus. But there is another link to the story here. While Shackleton’s ship was caught in the ice, floating to destruction, he wrote seeing an iceberg with a dark band embedded in it. Exactly what this was is not known. A volcanic eruption is one possibility. Could this have been the tephra layer from the Hudson Mountains eruption? The Weddell Sea may be a bit too far (and Deception Island is rathe closer) – but it is nice to think so.

Albert, September 2026