Iwo Jima in 45 eruptions

Iwo Jima has long been of particular interest to VC. It is an enigmatic volcano with a very human history. One of the pivotal battles of the second world war took place here, at immense human cost which arguably changed the way the war ended. The island remains home to the Japanese navy, and visits are restricted to those coming for memorials – of both sides. The people who visit find many reminders of the past. But in spite of the memories, the island has changed. The beach where the US marines first landed is now more than 10 meters above sea level. The island is also larger than it used to be. The volcano below is stirring.

Early September there was a report of increasing earthquake activity, which was followed by a minor eruption on September 12. It happened just off the south coast, and witnesses reported seeing water fountaining 10 meters high. But no pictures have emerged of the eruption, and after the event things quickly calmed down again. In itself, this was a normal event. Iwo Jima has such minor explosions every few years, sometimes on land, sometimes off the coast, although they have become much more frequent over the last decade.

We have found one (and only one) image which caught the effect of the eruption. A SkySat satellite photographed half the island on Sept 13. It uncovered strong discolouration in the water, on the south side of the island, in the general region where the fountaining was said to have occurred. Older images show nothing similar in this location, and it is likely volcanic ejecta. But the exact location of the outburst is not clear from the image: it will be either inside the coloured patch or just to the right in the region that was missed.

This post aims to bring together what is known about Iwo Jima, which perhaps is less than we should know. It is an extraordinary place, where the long-term inflation is among the largest seen anywhere in the world. The inflation is clearly volcanic: since 1889, 45 separate eruptions have been recorded. It deserves a closer look.


But first some background. The Volcano Islands (Kazan-rettō in Japanese) are a group of three small volcanic islands,of which Iwo Jima is the middle one. North of it is Kita-Iō (San Alexander Island), and south of it Minami-Iō (San Augustino Island).

Iwo Jima was (probably) discovered by the Spanish sailor Bernardo de la Torre, in 1543. On September 25, ‘they had sight of certain islands which they named Mal abrigos [‘bad anchorage’]. Beyond them they discovered Las Dos Hermanos (Two Sisters). And beyond them they also saw four [or three] islands more which they called Los Volcanes’. (Edited from Thomas Suarez, early mapping of the Pacific, 2013. For a more detailed report of the expedition, see the bottom of this post.) The first two islands are believed to be the Daito islands, and the last four or three islands are assumed to include Iwo Jima. Various sources state that he landed on Iwo Jima and named it Sufre (or sulfur) Island, but this appears not to be true: no landing is mentioned and the name was given by Cook’s expedition, two centuries later. In either case, the name has stuck: ‘iwo’ is sulfur, and ‘jima’ is island. (There are a number of ‘sulfur islands’ with a similar name, and it is not uncommon that an image of the wrong island accompanies a report on Iwo Jima. Anything showing a smoking volcano should immediately be discarded.) The official name was changed in 2007 to Ioto, and this should reduce the confusion. The new name has the same Japanese characters and meaning, but a different pronunciation and it is argued to be more historically correct. It will be hard though to remove the name Iwo Jima from memory. Too much has happened here under that label.

Captain James Kirk Cook is reported to have investigated the island in 1779. His journal contains a rather discouraging description of the (re-)discovery of the island:

Between 9 & 10 o’clock we passed by the island within a mile from the shore, and as it appeared to be barren and uninhabited we kept on our course without making any stay on it. At the southwest point of it stands a round hummock which appeared to be a volcano as we saw some brimstone on it; at the north east point stood a number of remarkable rocks close to the beach not unlike in their appearance to Stonehenge; the island is about 6 or 7 leagues in circumference & the land low & much of it covered with low shrubbery but no trees of any kind.

A remarkable detail is that this was on 15 November 1779, which is 9 months after Cook had died; there must be some doubt about his involvement! This discovery from beyond the grave was in fact done by James King and John Gore, who had taken joint charge of the expedition. The discoverers were sailing north to south, so the fact that they picked out the southwestern point as the notable feature means it really stood out. The expedition did not land on the island, in contrast to what is claimed on wikipedia. The journal states that Iwo Jima is one of three islands they saw. De La Torre may have seen four.

The journal of Cook’s voyages (authored by William Bligh & James Cook) contains drawings of the island. They show the view towards the east-northeast. The ‘hummock’ dominates, with a low neck and a rocky dome. At the far end there are some rock plugs on a small dome, only just above sea level, with presumably formed the ‘Stonehenge’ feature he refers to.

The map of the island shows the thin neck, the two high areas, and the surrounding shallows. It also indicates three areas further from the coast which presumably are rock reefs.

The basic features that Cook found still exist. They have acquired names since. The southwest ‘hummock’ is called Suribachiyama (or Mount Suribachi). The round dome is Motoyama, and the sandy neck in between is called Chidorigahara.

The risen land

But even though those features are still there, much has changed. On Cook’s profile of the island, Suribachiyama stands around 2.5 times higher than the highest point of Motoyama, and Chidorigahara is only just above the water line.

The recent situation is shown below, using profiles re-published by Kenneth LaJoie, in 1986. They show Motoyama almost as high as Suribachi, and Chidorigahara 40 meters above sea. Surabachi is currently about 160 meters tall. If it was this high also in Cook’s days, the old map suggests that at that time Motoyama was 60-70 meters tall. Nowadays it rises to 120 meters. Chidorigahara has gone from near-zero to 40 meters above sea. The conclusion is that much of the island has risen by 40-60 meters since 1779, where Motoyama has probably risen a bit more than Chidorigahara.

Source: Kenneth R.LaJoie, Coastal Tectonics (1986, published in “Active Tectonics: Impact on Society”.

The recent data shows the beach lines of the past. The one labeled ‘1779’ is suggested to be the beach seen by Cook, assuming that Motoyama and Chidorigahara have risen by the same amount (in reality, Motoyama is likely to have risen more). The line labeled ‘0.5-0.7 ka’ is for a layer which contains carbon-dated coral. Clearly, this layer was below water when the coral formed, somewhere around 1450+-100 AD. Nowadays, it is 100 meters above water. But the entire top of Motoyama shows coral fragments and rounded pebbles, showing it was entirely under water until quite recently.

So Motoyama has been coming up at an average rate of 20 cm per year, over at least 500 years! If Bernardo de la Torre did indeed see the same island, it must have looked a very different place, with Suribachi and Motoyama forming two separate islands. Perhaps this is why he may have reported four islands where later explorers found three (but see the bottom of the post).

Growing up

Iwo Jima. August 2018

The island has not only grown up. It has also grown out. The most current image comes from a satellite view obtained in August 2018. I have attempted to put the old map of Cook on top of the current view. This was done by assuming that Suribachi and Motoyama haven’t moved, and using these to align the two images. The overlay may not be perfect!

The comparison shows how much Chidorigahara has widened. Much of the dotted parts of the old map, indicating shallows, are now land. The east and west side of Motoyama have especially grown. The northeast coast, where the old ‘Stonehenge’ must have been located, has not moved as much. The coast here is quite steep.

The rock reefs are an interesting problem. One of the reefs on Cook’s maps, Kamaiwa, has become incorporated in the beach (this happened around 1968). His two other reefs have disappeared. Two new features have taken their place. On the west, there is a new island (called Kangoki-iwa) which is close to merging with the main island, and on the east side there is a new reef (Higashi-iwa) where the rocks come up to just above the water line. It is a little suspicious that two reefs have gone missing while two new ones have come up. It is possible that both were misplaced on the map.

The rocks of Higashi-iwa

The growth of the island has been followed now for over a century. Below are a series of maps compiled by Norio Oyagi and Takashi Inokuchi, and published in Geology of Iwo Jima. The maps depict the changes between 1911 and 1983. Until 1952, the coast line was fairly stable, apart from a slight extension on the west side. After 1952, growth took off. The changes on the west became spectacular. Cook’s old reef which since had acquired the name of Kamaiwa used to be 1 kilometer off the coast. The channel between it and the coast was 36meter deep and navigable. After 1952, the beach extended into the channel by as much as 50 meters per year, and in 1968, Kamaiwa became connected to the coast. Beyond it lies the island of Kangoki-iwa. It used to be more than 1.5 kilometers from the coast, far enough that it was used to house prisoners. The shark infested waters provided an escape proof barrier. By 1983 Kamaiwa formed the closest part of the coast. The beach has continued to grow since, and in the 2018 image above, is only a few hundred meters from Kangoki-iwa. Escape finally beckons for the prisoners, albeit too late to be of any use.

Why did the beach extend so quickly and by so much? This was more than just uplift. It turns out that it was related to the uplift of the coast to the north. The steep coast line here eroded while it was uplifted, perhaps hastened by the regular typhoons (several pass though here each year). The eroded sand was moved by the current and waves, and deposited on the west side. Sometimes the sea just can’t win. But it does try.

This image (from iwojima.com) shows the typical lineation of raised beaches. Look above the cliff, and the lineation continues in the rocks, suggestive of earlier periods of wave erosion.


Iwo Jima is doing more than just grow up and grow out. It is also frequently volcanically active. The eruptions are small, short-lived explosions, and come from a number of different locations. The same location can erupt again, even after decades. Interestingly, although the activity is centered on Motoyama, the outbursts are rarely there. They seem to follow the edge of the dome. But some eruptions have happened on Motoyama itself, in particular in December 2016 and in late 1969. The larger explosions can leave holes tens of meters deep and wide.

Fumarole activity is widespread, mainly from the crater on Suribachiyama and its steep western slope, and from the northern part of the island. The fumaroles deposit sulphur which was commercially mined until the war. At the top of Motoyama there are even some boiling mud pits. The vents move around, old ones go extinct and new ones appear.

On the map above, from Ueda et al., 2018, the numbered circles indicate the known locations of eruptions since 1890. The most frequently erupting location is ‘1’, with 13 outbursts, and ‘7’ with 10. These two locations are next to each other.

The 12 September 2018 eruption is close to location ‘10’ which previously erupted in 2001. It is a repeat offender.

The map also indicates four seismic stations. One, 0605, is located at Suribachi; the others are at and around Motoyami. They have provided continuous GPS measurements since the late 1990’s. Before that, measurements were taken every two years. Between 1980 and 2000, the GPS data showed relative little change: the rapid changes in the 1950’s and 60’s had given way to a quiet few decades. But after 2000, renewed uplift began to show, and since 2010 it has accelerated further. This was accompanied by increasing earthquake activity. Suribachi remained relatively unaffected, with minor uplift beginning to show only in the last few years. But the movement near Motoyami has been extreme.

Station 0604 has risen by 5.5 meters since 2000, most of which happened after 2011. Not many volcanoes in the world have shown such large movements! This continued from the pre-1980 events: between 1920 and 1980, the maximum uplift on the island was 11 meters. The peak at that time was north and west of Motoyama, while Motoyami itself rose by 6 meters. This time, the uplift seems centred on the south side.

Volcanic activity also accelerated. Over the past 7 years, there have been 20 separate volcanic explosions. During the 7 years before that, there was only a single event, in December 2007. There is more here than meets the eye: there are developments below ground.

Below the water

Iwo Jima forms the top of a strato-volcano. Remove the sea, and a large mountain appears, 2 km tall and 40 km across.

The plots below (from Sohei Kaizuka, 1992, Quaternary International, Vol. 15/16, pp. 7-16, 1992) show the detailed bathymetry, Iwo Jima is surrounded by a smooth under-water plateau at a depth of some 15 meters. At the outer edge of the plateau, there is a drop to a depth of about 100 meters, followed by a patchy rise which in a few places sticks out above the water. Outside of this rim, the mountain steeply falls to the sea floor, 1500-2000 meters below. It is a sizable volcano!

The patchy rim runs roughly from Kaimawa and the island of Kangoku-iwa, around the north side, to Higashi-iwa, 2 kilometers off the east coast. The rim is about 1 km wide. This is wider than Kangoku-iwa: there are four rocks which just break sea level a few hundred meters on the seaward side of the small island, which also are located on this rim. Two of these are visible as white spots on the August 2018 image above, west and northwest of Kangoku-iwa. On the north side, the raised rim is located some 100 meter below sea level. But to the south, the plateau ends with a sharp fall, without a clear raised rim. The profiles shown above have rims at positions A, B, D and F, but there is no indication at positions C and E.

The rim is argued to form a 10-km wide caldera. Iwo Jima and its plateau covers much of the caldera; Motoyama is near the centre. But the classification as a caldera may not be fully certain. The raised rim only extends halfway around the island, with Suribachi placed outside it.

The inner part is unusually shallow for a caldera. But this flat top, 10-km wide, on a mountain 40-km wide, unusual for a volcano, makes sense from the point of view of wave erosion. This has kept removing the top, down to the depth where waves can reach, some 15-20 meters. A caldera would imply a massive explosion removing the entire top of the stratovolcano down to exactly sea level. Erosion can do the same thing, but takes longer.

The rim has had volcanic activity. Kongaku-iwa consists of lava. Higashi-iwa looks like a volcanic cone, similar to Suribachi. Suribachi itself is just outside the rim. But the volcanic outbursts over the past century have been minor crater-forming explosions, rather than cone building, and have taken place on the plateau and the island, and not on the rim. Clearly Iwo Jima can do much more than what we have seen in recent years.


In fact, the island has in the past blown its top, and it has done so more than once. The evidence is everywhere. Motoyama is covered by pyroclastic deposits. Suribachi has its own pyroclastic layer, with some lava. Kangoku-iwa consists of lava. Bore holes at Motoyama going 150 meters have found alternating layers of lava and pyroclastic deposits. This adds a warning from history to the on-going inflation. It would be useful to know the full volcanic history of Iwo Jima.

Geological map, Ueda et al. 2018. The map shows the various parts of the island, with the local names. The colours indicate various deposits, including pyroclastics and some lava.

Pyroclastic deposits surround both Motoyama and Suribachi, but they form separate layers and do not come from the same eruption. It appears that Motoyama erupted first. Carbon dating has been done on the lowest layer of ignimbrite and on the rocks of Kaiwama beach. Lava is itself of course not date-able: the carbon is obtained from whatever organism was buried by it. At Kaiwama these were shells, which were dated to 131+-20BC and 31+-20 BC. At Motoyama, interestingly, it was carbonized wood and twigs, dated to 761+-20BC and 762+-20BC. This means that Motoyama, at the time of the major explosion, was partly above water and was forested.

Around 2700 BP, this forested island was destroyed by an explosion. The island disappeared below water, were the ignimbrite rained down. Next, lava came and build up a new dome, but it remained submerged. The lava had an unusual composition called trachyandesitic. More eruptions followed, with layers of pyroclastics. How long this lasted is not known. Based on the carbon dates of the shells, it may have continued for over 500 years. It left a yellow, soft tuff, used in the war to dig shelters.

Suribachi formed afterwards, but how much later is not really known. It existed in the time of Cook, and probably also in 1543, and must therefore have formed earlier. That still leaves us a 2000 year window. Trachytic magma was erupted at Suribachi three times, first as marine pyroclastic eruption, second effusive in shallow water, and finally form a pyroclastic cone on land.

Now the third phase started, where Motoyama calmed down but inflated, at an average rate of 20 centimeter per year. Precisely when the inflation started is not known. At has lasted at least 1000 years, but may also have begun shortly after the Motoyama eruption. And it is still continuing, with frequent volcanic activity. The fact that the top of Motoyama is littered with old coral shows that there was no lava associated with the rise. It was pushed up from below, not build up from above.

In this history, Motoyama is the sum of the 2700-2100 BP eruptions pushed up by the subsequent inflation. The under-water plateau surrounding the island may be the result of the volcanic ejecta, less affected by inflation and continuously decapitated by the waves. Suribachi is a separate eruption, perhaps 1000 years ago, perhaps older.

What about the caldera rim? If this is a caldera, it existed already before the eruption at 2700 BC. There must have been a massive eruption, but all we can say is that it was earlier than 2700 BC. There is no evidence for this other than the partial ring. However, there is another layer of lava underneath the Motoyama lava, called the Hanareiwa lava, and this shows that there were earlier eruptions. The age is unknown.

Seismology indicates that below the Hanareiwa lava is another lava layer. Together, the three layers are 200 meters thick. Between 200 and 500 meter deep is a high velocity layer, presumed to be tuff. This must be from an even older eruption and perhaps this was the caldera forming eruption. Below this is even more lava.

Pyroclastics in a Motoyama cliff face, surrounding a large tuffaceous block

Magma on the move

But what is causing this extreme, long lasting inflation? Intermittent inflation and deflation in large calderas is common, and normally caused by moving hydrothermal water. But the inflation at Iwo Jima is far too large for that. It is caused by accumulating magma. The geodesic measurements show that there are shallow sills underneath Motoyama. They grow, erupt and deflate, pushing the summit of Motoyama up and let it come down again. But at the same time a much larger area is continuing to inflate. That indicates a deeper, and growing, magma reservoir.

And what is causing the eruptions? They are phreatic, meaning caused by water. It appears that there is water inside the deep tuff layer, 200-500 meter under ground. Rising heat brings pockets of this water to the boil, but it can’t get out because of the seal of solid lava above. The lava is pushed up but the agitated water, causing rapid inflation and earthquakes at ground level. After a few days a crack develops in the lava seal. Now we get a sudden explosion, as happened on 12 September this year. The eruptions are primarily in the region where the rise of the central region has caused faults, around Motoyama but within the caldera rim. Two such faults became active shortly after the war.

So where is the magma? The shallow sills are thought to be 800 meter to 1 kilometer deep. It is fed from a deeper magma chamber, where the main inflation occurs. A pulse of magma into the deep chambers brings with it heat, which percolates up into the tuff water. Phreatic eruptions follow. Iwo Jima is currently in such a phase.

Future history

In the long term, Iwo Jima seems to repeat itself. The inflation will continue, but eventually a major eruption will break through and destroy the island. Pyroclastics and lava will build a new dome. After a few hundred years, the eruptions cease. Inflation resumes and over time the island reforms. Until the next eruption.

Where are we in the cycle of Iwo Jima? That is not easy to know. However, it is unlikely that the island was much larger than it is now, at the time of the 2700 BC eruption. The build up to the next one is well under way. But when exactly is impossible to say. How large would the eruption be? We can make a maximum guess by assuming that a 100 meter high island, 23 km2 in size, is replaced by a 200-meter deep hole. That suggest something in the range 5-10 km3. This the DRE value: the tephra will be several times larger. At the top end, a Krakatoa-size eruption might be possible.

But volcanoes do not like being predictable. Sometime before 2700 BP, perhaps long before, it may have done a significantly larger eruption. Just to keep us on our toes.

Iwo Jima has played a part in shaping the post war world. The memories of that event run deep and have left scars, but scars are also signs of healing. We should not forget what happened here, but it belongs in the past. While Iwo Jima keeps rising, the volcanic heritage will become more and more important. One day, it will happen again.

Albert, October 2018

Recent papers on Iwo Jima, extensively used for this post:

Volcanic History of Ogasawara Ioto(Iwo-jima), Izu-Bonin Arc, Japan Masashi Nagai* and Tetsuo Kobayashi, 2015, Journal of Geography(Chigaku Zasshi), 124, 65–99 (with many images of rocks formations on the island)

Phreatic eruptions and deformationof Ioto Island (Iwo-jima), Japan, triggered by deep magma injection, Hideki Ueda*, Masashi Nagai and Toshikazu Tanada, 2018, Earth, Planets and Space, 70, 38

source: Nagai and Kobayashi, 2015

Appendix: The discovery of Iwo Jima

In 1542, Ruy Lo´pez de Villalobos sailed a Spanish expedition from Mexico to the Philippines. After arrival, and meeting a mixed reception from the locals, Bernardo De La Torre was tasked to find a route back, taking one ship from the six of the expedition: San Juan de Letra´n, in order to ask for reinforcements. This task failed: the first successful eastward crossing of the Pacific happened only in 1565. De La Torre sailed a route roughly north/northeast, reaching 30 degree north and finding a number of new islands, but eventually was forced back by storm and lack of water. The chronicles of his journey appear to be lost, but the trip is mentioned in several documents from the 1540’s and 1550’s. These reports are second-hand and in some places are contradictory. The most likely actual journey was pieced together by Bernhard Welsch, in 2004, and I am following his arguments here.

De La Torre departed from the central Philippines on 26 August 1543 (the dates in use at that time were the Julian calendar. The Gregorian calendar was not adopted until 1582). He sailed east for a few days before turning north, eventually crossing the tropic of cancer. (Typically, reported latitudes are reliable, but longitude could not yet be measured.) Several islands and groups of islands were discovered on this leg, before he was forced to turn back.

On 25 September 1543 they sighted a small island at 26°N which they called Mal abrigo (bad anchorage) because the sea was breaking against it. They sighted two more islands 26 leagues further which they called Duas yrmaas (Las dos hermanas, meaning The two sisters), but didn’t land there. (A spanish nautical league at this time was not perfectly defined, but in practice there were about 15 leagues to a degree, so 26 leagues was a bit less than 2 degrees.) Later they saw three (one of the reports says four) more islands, at 24 and 25°N. One of these was volcanic with fire in three places. These islands were called Balcones (Volcanoes).

On 2 October 1543, they sighted an island they called Forfana (the orphan), beyond which there was a high mountain or rock, with fire at five places. But the story here appears confused, because the description ‘fire at five places’ sound very similar to the previous ‘fire at three places’. The first is reported in one of the second-hand sources, and the second in another. It seems likely the two sources were reporting the same observation, but attributed this to different islands. In reality, only one active volcano was seen.

On 18 October 1543, between 29° and 30°N, after hitting a northerly storm, they became worried about their supply of drinking water. They turned around and after 13 days arrived back at the departure point. On the way back, they came across some smaller islands ranging in a north–south direction from the 15th to the 16th degree N latitude, which were the Ladrones (the Marianas), but they did not anchor there.

So which islands did they find? The first ones, Mal abrigo and Los Dos Hermanos, are accepted to be the Daito islands. Mal abrigo appears to be Oki Daito, 24.5°N and surrounded by coral reefs, and the two sisters are Minami-Daito and Kita Daito, about 1.5 degree due north of Oki Daito and less than 10 km apart. One of the old but second-hand reports places the first island at 16°N, and this is therefore often identified with Farallon, but this appears unlikely: the Daito islands, 10 degrees further north, fit much better with the descriptions.

The three ‘Volcano’ islands are identified with the Kazan Retto group, 10 degrees due east of the Daito islands; Iwo Jima is the middle one of the three islands of this group, and it is located at 24.8°N. That identification was first suggested in 1803 by Burney, a member of the Cook expedition. Burney wrote :

it will appear very probable, that the Sulphur Island, with the North and South Islands, seen by the Resolution in her return from the last voyage of Captain Cook, are the islands which were called the Volcanes, discovered by the San Juan. Their agreement in number, their spreading nearly a degree in latitude, and in the same parallels, and their appearance so well corresponding to the name, form a combination of circumstances that amount to very little short of conviction.’

The next island that was discovered was called ‘Forfana’, and is said to be an uninhabited island 30 leagues (150-200 km) from ‘Volcanes’. It was sighted on 2 October 1543. The direction is somewhat problematic. One source says it was east to northeast from the Volcanes, where only empty ocean exists. But this source also states that the second Daito island discovered is northeast of the first, while in reality it lies due north. If the same mistake was made here, the final island really lies in a north-northeasterly direction. And there are islands there: the southernmost part of the Osagawara archipelago, where the nearest island is 150 km from the Volcano Islands. This makes ‘Forfana’ to be Hahajima, the southernmost of the major islands, or one of the small islands which surround it at 2-3 km away. This fits well with the description of an ‘orphan’ (Forfana) beyond which was a high mountain. The highest mountain on Hahajima is over 400 meters.

The route from the Daito islands to Iwo Jima and on to Hahjima is not straight. It runs a bit south, and turns northward. If De la Torre passed through the Iwo Jima groups east-west, he could well have missed either the southern or northern island of the group. If Iwo Jima consisted of two separate islands at the time, as seems likely, it is perhaps possible that of the three Volcano islands he saw, Iwo Jima accounted for two. But this is speculative and mainly illustrates the uncertainties around the story.

After Hahajima, De La Torre sailed on for two more weeks into the open ocean, but eventually was forced to turn around, found Vila Lobos gone, and finally met up with the expedition around the Moluccas (not far from Sulawesi) in early 1544.

There are some lessons from this story. The San Juan was nowhere near where the sailors thought it was. It sailed a complex route, first east, than due north, east-southeast again, and finally north-northeast, putting the ship far to the west of where it though it was going. If you can’t measure longitude, the course can be a bit of a guess. Some of the deviations can be explained by tacking against a predominantly northeasterly wind, and others by the effect of the west-flowing north equatorial current, which becomes the northeast-flowing Kuroshio current in the region of the journey of De La Torre. Some may have been deliberate. He may have been aiming to catch the returning North Pacific current across the Pacific, at 30N, and therefore trying to gain latitude before turning east. Or he may have been deliberately looking for islands, in order to obtain water. We can only guess.

We have further learned that Iwo Jima was discovered after 25 Sept but before 2 Oct 1543 (Julian dates). Often, the first date is given, but that is for the Daito Islands. Looking at the distances involved, the most likely discovery date is 30 September 1543. We have found that the name ‘Sulphur Island’ comes from the Cook expedition: De La Torre’s name for the islands was just ‘Volcanes’. What we do not know is what Iwo Jima looked like, however as it was clearly labeled as ‘volcano’, the peak of Suribachi must have existed, as this is the only obvious volcanic feature seen from a distance.

That fire-belching volcano

There is one final point to solve. What was that erupting volcano that so impressed the sailors that two second-hand accounts tell the story, albeit attributed to different islands? One account places it at the Forfana group, i.e. the eruption was at or near Hahajima. This group of islands is indeed of volcanic origin, but far from recent. The eruptions were several million years ago, the chain has gone quiet and the eruptions are nowadays at a separate chain of islands further west, where for instance Nishinoshima is located. In spite of the statement placing ‘a high mountain or rock, which was belching forth fire at five places‘ at Forfana, this event cannot have been here.

The other account places it at the Volcano islands: ‘‘they sighted three more islands; one being a volcano belching fire at three places’. This narrows it down to one of three possible islands. Assuming both accounts are for the same event, we are looking for a high mountain with multiple eruption sites.

The three Volcano islands are, from north to south, North Iwo Jima (Kita-ioto), Iwo Jima (Ioto), and South Iwo Jima (Minami-ioto): together they make up the Kazan Retto group. The southern island is a single, 900-meter tall cone, about 2-km wide at sea level. There was a marine eruption a few kilometers off the coast in 2005, where lava floated to the surface. The northern island, Kita-ioto is larger but slightly less tall, at 800 meters: it appears more eroded. The summit of Kita-ioto is extinct, but there have been several eruptions off the coast over the past two centuries. Neither island is known to have had historical activity on the main mountain on-land. But the description clearly refers to an eruption on land. Although neither can be excluded, given that Iwo Jima itself is known to erupt on-land, it is a likely candidate for the erupting mountain.

But Iwo Jima also shows little or no evidence for recent lava. The most recent ‘fire’ eruption was probably the one which caused the top pyroclastic layer around Suribachiyama, and this could be as young as 1543. But the description gives no mention of explosive activity, just multiple fires. Phraetomagmatic explosions, similar to the current activity, can also be excluded, because it lacks associated light. Was the entire description made up, or greatly overstated?

There is an alternative. The crater of Suribachiyama, and other places on Iwo Jima, can show heavy fumarole activity. Looking at the dates and distances, De La Torre probably passed here around 30 September. Two days later, Oct 2 1543, was a full moon. Did he perhaps pass at night, and see very active fumaroles reflecting the light of the nearly full moon? That might have looked like fire to people unfamiliar with fumaroles, and it would explain the multiple locations where fire was seen. A night vision could even explain the confused reports about the actual location of the event.

So, event though the circumstances of its discovery are now clarified, the volcanic fire that so impressed the sailors remains a mystery. If only we could recover the lost chronicles of the discovery! Second-hand reports, even from within a few years of the events, leave too much unclear.

Main source: (2004) Was Marcus Island discovered by Bernardo de la Torre in 1543?, The
Journal of Pacific History, 39:1, 109-122, DOI: 10.1080/00223340410001684886

344 thoughts on “Iwo Jima in 45 eruptions

    • Tell me about it! Jesper’s in full magmageddon meltdown, I may have to start administering the electronic valium…

  1. Hehehe … hmhehe
    Just me that wants Ontong Java Plateau on land 2.0. Its my Birthday 29 october
    I Hopes the planet makes one for me on my day
    😏 I will get a Siberian Traps chocolate – icecream Igenous cake on my birthday
    With VEI 8 effusive sugar glacing on
    There is NO mantle plume today Thats strong enough for magmageddon today I think.

    • Hawaii is, half of its cenozoic erupted volume is less than 2 million years old, and 1/5 is in mauna loa and kilauea.
      Otong java was not as sudden or rapid as many other flood basalts, it was maybe like Iceland but faster, rather than a submarine Deccan traps.

      Also having a flood basalt entirely in 1 day would turn the earth into a hell planet that makes venus look nice, as I said before if the Deccan traps was a near instant event it would heat the atmosphere to over 1000 C worldwide. The amount of energy released by the Deccan traps lava cooling to ambient temperature would be equivalent to something like 100 million tsar bombas, 3 orders of magnitude more than holuhraun or leilani, every day the Deccan traps erupted the same amount of lava as kilauea erupts in a year… Siberian traps is the same but every number is doubled…

      A curious thing about these flood basalts is that the really deep plumes can dredge up magma from deep enough in the mantle that it erupts as ultramafic lava. Most ultramafic lavas are komatiites that are over a billion years old, but there are 90 million year old komatiites in the Caribbean LIP which was formed by the Galapagos hotspot first surfacing, nyiragongo might erupt some of this stuff in the future too with how extreme its lava already is.
      Komatiite lava would have basically been like iron ore slag, 1600 C and with a fluidity probably similar to oil or even water. In a big eruption this stuff would have been just as scary as any pyroclastic flow.

  2. LoL … sorry, just me enjoying volcanism
    Yea sometimes .. one gets geeky

  3. Őa in iceland is inflating. Magma at 5 km… jon freeman is reporting. Is this a case of consern

  4. Öræfajökull Is a Classic closed conduit system.
    Cold and brittle in its upper parts
    There will be a major earthquake swarm: if it really wants to erupt. The magma haves to cut its way through the old solidifyed magma in the upper conduits. The volcano been sleeping since 1700 s

    The Next eruption coud be a small to strong VEI 5
    It will likley get very very very explosive.
    This volcano makes Ryholite.. its a magma with a very high sillica content and makes it extremely viscous. If the gas content is high it will give a very explosive power and potential that blows the magma into ash. Ryholite ash is often light in colour and extremely fine.
    This volcano is east of the mar rift zone
    and may become magma starved and inactive in the future as the spreading push it away from the hotspot and mar ridge supply

  5. For anyone looking at readings on the big Island over the next couple of days…

    * SURF…Building to 8 to 12 feet tonight through Friday along
    south facing shores.

    * TIMING…Through Friday afternoon.

    * IMPACTS…Moderate…Expect strong breaking waves, shore
    break, and rip currents making swimming difficult and


    Beachgoers, swimmers, and surfers should heed all advice given by
    ocean safety officials and exercise caution. Know your limits and
    seek ocean recreation areas best suited for your abilities

  6. Can anyone here give me the link for cumulative seismicity plots for Grimsvötn?
    I have not looked at these for over 2 years and wants estimate the conditions of the volcano.
    I cannot find the link

      • Nice .. looks like its even picking up a bit in the end. Still lots of magma was lost in 2011. But Grimsvötn haves a huge magma supply and its rapidly refilling

      • What is not really obvious in the CSM-plot is the lack of M>2 quakes for the current (red) curve. I think that is what is causing the current curve to lag behind the previous two curves. If you ignore the size of the quakes and instead only count the number of M1+ quakes, the current curve follows the two previous curves much better.

        Here is a graph I did by counting quakes: I only included M>1, since the detection capabilities have improved a lot and a large number of smaller quakes are now detected which would bias the result for the current curve.

        Even if the seismic moment release is slow, it seems like the quake frequency is increasing steadily. I wonder if the lack of M>2 is due to the ground still being hot and ductile since the large eruption in 2011? Or maybe the Holuhraun eruption had an effect on the local stress field, enough to affect the behavior of Grímsvötn?

        • There is a bit of a uptake in the plots over recent months. But no indication for the sudden acceleration that leads up to an eruption. I think is at the least 1 years away, and more likely several years. Öræfajökull will go first, I think.

        • Grimsvötn haves the highest magma supply in entire Iceland
          The whole arera there is hot and ductile and there is huge ammounts of geothermal activity under the ice
          This hot ground forms Grimsvötns subglacial lake and Skaftarkatlar cauldrons. If it was ice free .. it woud be a pretty impressive sulfur and fumarole fields in that arera
          This is the heart of the Iceland hotspot. It cannot sleep for very long given How high its magma supply is. In my lifetime its been 5 eruptions in Vatnajökull
          1996, 1998, 2004, 2011, 2014
          The 1996, 2011 and 2014 was very volumious and impressive.
          Holuhrauns 1185 c to 1195 C its one of the hotter basaltic eruptions measured so far.

          • And Gjalp 1996
            was pretty much a subglacial tuya eruption that never got above the pillow hydroclastic ridge .. pheratomagmatic Surtseyan phase.

            Had Gjalp lasted longer .. the lava woud break through the ice and meltwater lake and gone effusive with lava fountains and rivers.
            A dark volcanic cone woud Then grow there.

            It never got to that phase

  7. https://www.facebook.com/photo.php?fbid=246525602883255&set=pcb.1112085198957297&type=3&theater&ifg=1

    Some nice pictures of a few of the vents on kilauea. Apparently the fissure 22 cone is actually not that big, only about 18 meters tall from its base, but fissure 8 cone is much bigger, 35 meters high from its base and about 50 meters above the pre-eruption ground, and probably 200-300 meters wide, no idea how big fissure 17 cone is but it is probably closer to fissure 8 cone than fissure 22, but only one person has apparently seen it since May and they gave no landmarks to go off so I have no idea how big it is. You would think that being observed forming in action, the first major vent of the eruption and the first not-basalt erupted in Hawaii in recorded history would make it more interesting but I guess not.
    Still wondering what fissure 8 has been named, apparently a name has been decided but not been made public yet.

    Once pohoiki road is rebuilt across the rift (might take a while still) I think fissure 22 cone will become quite popular with tourists, being near the road and very quintessentially ‘volcano shaped’.
    Also once the road across the lava flow to pohoiki from mackenzie is finished I think a lot more people will go exploring around the earlier vents too, especially fissure 22-17 which were largely cut off by the big flow from fissure 8. Someone will probably take a picture from near the livestream house to show how much the view changed from mid May when it started.

  8. New burst of activity Iceland. Hekla shakes twice and Habunga, Godabunga, Askja…

    • This one’s an interesting one, team Katla just got a boost:

      25.10.2018 19:00:47 63.634 -19.072 18.8 km 2.1 99.0 6.1 km NNE of Hábunga

    • Don’t care, still probably going to ruin the cookout I had planned for Saturday. 🙁

    • Well, if you are demented as I am, I would still grill the sausages while the storm is raging…. but then I keep my grill on a screened in back porch with a roof. (The roof is handy in keeping the screen walls attached.)

      BTW, I literally lived in this room in the weeks after Ivan, it was the only place it was cool enough to sleep. (fan powered by the generator keeping my freezer and fridge running)

      BTW, my wife just told me it’s raining outside now, so the moisture blob has gotten this far.

      And for pure entertainment value, the sad tale of the USS William D Porter. Though likely embellished, most of this is actually true.

  9. Somewhat off topic – I live near the bristol channel which has a 15m/45foot tidal range -when the tide here goes out it goes out a long long way (in the region of a mile/kilometer I think)

    So now imagine a tsunami hits the bristol channel when the tide is out – and lets say it’s a 10m tall tsunami – I know it would get funneled somewhat – but given that the beach normally handles 15m what would it look like locally – would it just appear as an unexpected extra tide in the middle of the normal tidal sequence?

    Obviously if the tide was already in the wave would be on top of that and so be devastating, but if the tide were out would no damage occur, what about boats moored on the mud with long enough anchor chains to handle 15m tides ?

    • First, the Bristol channel does not get strong tsunamis: it is protected by the continental shelf which would take energy out of the incoming wave. It s very susceptible to storm surges: https://www.volcanocafe.org/the-bristol-tsunami-of-1607/ These take much longer to come in.

      The height of the tsunami is measured as the wave height. But at the coast, it runs much higher: this is called the run-up height. A ten meter tsunami would probably still overwhelm the flood defences, especially in the bays along the channel. The boats may be able to handle the height but the anchors won’t hold against the flow. Boats resting on the mud will be destroyed. When the water retreats, it is full of debris and anything that survived the incoming wave will get hit by this.

      Further up, the bore would become utterly destructive.

      • My pet subject/ specialism.
        Yes to everything Albert said.
        But to take it further, there are some important locations to be aware of.
        First of all, Avonmouth. In the area around Avonmouth the estuary suddenly becomes very much shallower, so this is a point at which any incoming surge would slow down radically, allowing the water behind it to catch up, and the event front/ wave would grow significantly.

        Next is Beachley. As a former member of the bore surfing crew, we have scoped out the estuary in great detail… We have access to flat bottomed RIBs and a Master Mariner among our number, for whom the estuary is his home water. On a normal bore tide this is the most seaward location we have ever witnessed the tide forming a wave, and although it was tiny, it probably becomes visible here due to a further sudden narrowing. Ergo the size of any incoming surge would become very much amplified at this point.

        Sharpness – Lydney. There is a further narrowing here, and I will censor some of the information about the wave, as I know there are surfers out there who trawl the internet looking for information to guide their quests for the kind of waves they’re looking for. If any are reading this, just don’t. Don’t even go looking. Attempting to access this region of the estuary without very detailed knowledge accumulated over many years, and a very advanced and specific skill set will result in death. Not “Risk of”, but the real thing. Venture onto these mudflats, or even just getting into the river and letting the current of the river and retreating tide will kill most who don’t have the right skills and knowledge.

        That being said, there is a further narrowing, and if it weren’t for the prevalence of sinking sand/ quicksands in the area, you could wade across from one shore to the other at low tide. The wave would again intensify.

        Sharpness sill is the datum point for the tides used by the bore riders. The maximum tidal range on a very big peak tide here is just a little shy of 11 metres. And the lock gates where the Sharpness canal meets the estuary are huge. The drop from the harbour wall to the mud at low tide is around 50 feet ! I have personally seen a 9.8 metre tide augmented by a half metre storm surge in this region. The difference that half metre made to the wave (datum point for that was Avomouth, where the estuary is very much wider) was very significant. A 9.8 metre tide is a decent size to surf and have a lot of fun with. I rode that tide, picking it up several miles further inland. It was a wild and untamed beast. Extraordinarily powerful. I am i no doubt whatsoever that as big as the defences are, they would be utterly swamped by a very large wave and an even bigger following surge. The sea defences on the Northern shore are less substantial.

        Lastly, Hock Cliff near Frampton is relevant. Here the river-proper begins with a sharp Northward bend. The bend itself is up to a mile wide and is notorious both for quicksand and intense currents. A very large whirlpool often forms there due to the geography and the competing currents of river and tide. Hock Cliff is on the outside of this bend, and seems to be made of some pretty fragmented material. This could be very significantly gouged by an event of this magnitude.

        Beyond that point lie a few miles of mudflats up to about half a kilometer wide in places , with a low tide channel up to about 100m in width . This is the area I’ve surfed (between Arlingham and Maisemore Weir). Any surge that reached this point would again be massively increased in size, easily overwhelming all current flood defences

        I could only foresee a very large loss of life and probably complete destruction of any settlement near the river on either side in this region, with water travelling many miles inland in low lying areas, and along the courses of tributaries certainly as far up as Maisemore (which would include rivers like the Leadon, which runs up through Ledbury near the Malvern Hills.

        Damage and loss of life would be catastrophic.

        I would guess that the end point of the surge on the Severn itself would be the very substantial weir at Tewkesbury.

      • sorry forgot to say this was just a hypothetical – and that I was never in doubt that tsunami would probably not happen there (I guess massive landslide – part of devon falling to the water or similar, might possibly generate something if we ever had a really substantial quake, but again the tide would need to be in for the landslide to have water to displace) – and the answers below are interesting and I’d forgotten that I’d read albert’s 1607 article so that was a good refresher – thanks 🙂

        but tidal bore comes in with the tide I think, and storm surge is only significant if the tide is in

        looks like (from albert’s comment) the term I was looking for was ‘runup height’ – what sort of run up height would you get from a 1m/2m/5m/10m tsunami in the bristol channel ?

        and would it matter if the tide was in/out for the gain in runup height ?

        and assuming it wouldn’t matter and we only get 5m runup height – if it arrived when the tide was out – what would it ‘look like’ – I know the classic kanagawa tsunami wave image is wrong, and that it is more like a tide coming in, but I’ve only ever seen tsunami footage where it has been devastating, and that doesn’t quite look like the tide, and I’m trying to imagine what I might see if I were on the beach.

        first sign is apparently the tide goes out quickly (even though it is already out – so further again, presumably just down to the trickle from the river itself) and then comes in quickly with tidalwave, does it then ebb back out again to the ‘normal’ tide out level – or does it slosh around up and down for a while before eventually calming down ?

        • I was assuming this was hypothetical !

          Storm surges have a tendency to come in with the tide, so with storm surges it’s just as well to add them on to the height of the tide. If you were to look at water level over time for somewhere like Sharpness, the chart produced by the data would look somewhat akin to a saw blade. From low tide, the rise to high tide would have a very rapid onset ( about 35-40 mins) , then almost no slack tide period (a couple of minutes) followed by a fairly rapid initial drop in level, slowing down as it goes.

          I think in the case of the Bristol Channel this may well be very similar to the onset of a tsunami. We are both aware of how much the water retreats, and that it does so quite rapidly, so I reckon it would be very difficult to actually see the drawback, unless you were watching from somewhere like May Hill near Newent (great view of the estuary) in clear weather conditions.

          The stage of the tide would almost certainly be critical, but bear in mind that as far as I am aware most tsunamis do indeed involve several waves, with the water retreating as Albert said, violently in between them. This increases the likelihood of an interaction with an incoming tide. If it did so, this would indeed be devastating, because with the estuary being shaped as it is, retreating water from an earlier wave would have nowhere to go but back in… with whatever debris it picked up on its first visit.

          Conjecture: This would be further complicated by all sorts of rebounds and wave harmonics, up to and including rebound effects from English and Welsh coasts, and as far afield as Ireland.

          As to what it might look like as one such wave were arriving? Much as the bore does when it crosses mudflats I would guess? There’s a very low altitude paramotor video of that happening somewhere near Saul , south of Gloucester about half way through this video. Just scale it up (considerably ).

          Height of the event front would depend on very localised conditions. There are places in the river which can produce overhead waves, in circumstances in which everywhere else it might only be thigh high. That is dictated by the shape of the bottom, the direction of flow, and so on.

          So I don’t think it is REALLY possible to call it on height, except to say that the further inland it travels ( at least up to about Minsterworth) the bigger it will get. A wave showing 1 foot of surge as it passes beneath the inner of the two Severn bridges may well show over 10 feet between Newnham and Minsterworth, based on previous observations (it’s usually only about an inch under the bridge at Beachley, and yet produces a wave of 2-4 ft further inland).

          Around 2 min 50 seconds in, scaled up is what it would look like.

          • that’s grand, thanks, that’s cleared it up in my mind nicely, thankyou 🙂

  10. Hawaii have the capacity of ultramafic temperatures.
    At Kilaueas and Mauna Loas roots temperatures are above 1500 C
    Hawaiian hotspot is really really hot. When the magma first forms in Kilauea it haves a temperature of around 1520 C or more. I takes 8 years for it to travel upwards from the astenosphere and then on its way up it have cooled to about 1210 c as it emerges into Halemaumau. As it reached Puu Oo and Leilani it have cooled around 1150 Cin the east rift system storage.

    If magma supply in Kilauea increases and gets faster and rise more faster, it wont cool so mouch.
    Kilauea have capability of doing really really hot lavas if it really wants.
    Maybe Kilauea will erupt 1400 C lavas one day? The Greenland LIP did that when it was going.
    If magma starts rising very quickly in Kilauea lavas will emerge hotter and hotter.
    Hawaii is a really deep and hot plume and maybe capable of erupting superhot basalts.
    Hawaii already is the worlds hottest basalt

  11. Hi Folks,
    Possible discussion item/post. There has been a spattering of quakes in the Alban Hills outside Rome and the Avernge in France over the last number of years. Can these be be deemed truely extint volcanic systems? From research i understand that the last eurptions of the Avernge was about 6000 years ago but longer for the Alban Hills. Thoughts?

    • Auvergne has been active for several million years and eruptions extend into holocene in Le Chaine des Puys so it is not likely to be extinct, it just doesn’t erupt very often.

      • There are many volcanoes like that in the most unexpected places, close to where I live there is a volcanic field, Campo de Calatrava that had been inactive for the last ~700000 years, and then 5500 years ago it decided to wake up and produce a small eruption.

        • I read that fumerols were active there unti lthe 18th century, certainly a sign that there may be plenty of energy left underneath the surface. Im sure that if you went looking activity could be found at the surface in some form or other.

          • The fumarole is still active, you don’t see much there though, dead animals can be found in it sometimes. There are cold-geysers driven by CO² of magmatic origin every few years, the one known as Chorro de Granatula which formed in year 2000 was higher than 50 m. There are also apparently some craters from phreatic explosions that are younger than the last eruption of the volcanic field.

  12. Alban Hills are in deep dormancy but… they are almost certainly still alive
    Volcanoes can sleep for very long without being concidered dead

    • I am a big fan of our european volcanoes but most of the chat is usually dominated by Vesuvius, Etna and the Canary’s. The dangers of Campi Flegrei are now more well understood however little explored are the many other long dormant European volcanic features. The French and Spanish mainland come to mind, the volcanoes of the central spine of Italy and the Alban hills with the Eiffel volcanic field. Are we blinkered to the big names, the Vesuvius’s of Europe? How well are the overlooked volcanoes monitored or do they need to be?

      • And don’t forget the Greek contingent -Thera, Nisyros, Methana

  13. Campi Flegeri is problematic dense populated
    even a small lava dome eruption in that huge city woud cause enromous material disaster and costs!

  14. “Suspended over the western slope of Arsia Mons, an enormous volcano near the red planet’s equator, the elongated cloud stretches for about 930 miles, and was first spotted by the European Space Agency’s (ESA) Mars Express orbiter on September 13. Though the cloud looks like the kind of volcanic plumes huffed out by Earth’s active volcanoes, Arsia Mons is long extinct—its last eruption is estimated to have occurred around 50 million years ago.”


  15. “Quantification of ash sedimentation dynamics through depolarisation imaging with AshCam”

    Abstract: “Even modest ash-rich volcanic eruptions can severely impact a range of human activities, especially air travel. The dispersal of ash in these eruptions depends critically on aggregation and sedimentation processes – however these are difficult to quantify in volcanic plumes. Here, we image ash dynamics from mild explosive activity at Santiaguito Volcano, Guatemala, by measuring the depolarisation of scattered sunlight by non-spherical ash particles, allowing the dynamics of diffuse ash plumes to be investigated with high temporal resolution (>1 Hz). We measure the ash settling velocity downwind from the main plume, and compare it directly with ground sampled ash particles, finding good agreement with a sedimentation model based on particle size. Our new, cost-effective technique leverages existing technology, opening a new frontier of integrated ash visualisation and ground collection studies which could test models of ash coagulation and sedimentation, leading to improved ash dispersion forecasts. This will provide risk managers with improved data quality on ash location, reducing the economic and societal impacts of future ash-rich eruptions.”


  16. Fernanadina island will likley merge with Isabela on Fernandinas eastern side.
    As it continues to grow larger.
    Galapagos is a pretty powerful hotspot,but its nothing like Hawaii.
    Fernandina completely lacks rift zones and short lives eruptions around the calderas gives it these steep slopes and round shapes. Happy there was 3 eruptions in less than 360 days now in Galapagos.
    Still Sierra Negra is a very underrated volcano.. the 2018 eruption made a kilometer long 300 meter high curtain of fire the first day.

  17. I’ve just noticed that some of the major Galápagos volcanoes have a startling resemblance to the giant shield volcanoes of Mars, just nowhere near as big in scale as their Martian counterparts. This also includes the smaller trio (but still enormous by Earth standards) SE of Olympus Mons, namely Ascraeus, Pavonis and Arsia Mons in order from north to south. Also, the same goes for a number of much smaller shields. Many of those shields have disproportinately large calderas and lack rift zones. Olympus Mons somewhat less so, but a region of flatter slopes surrounding the present nested caldera hints to me that it may have had even bigger calderas in the distant past as it grew over milions and millions of years.

  18. And I freaking wants a time machine
    I wants to see Siberian Traps in action
    What a wonderful sight it must have been… what an awsome sight.
    I imagines fissures… many hundreds of kilometers ( 600 km ) long with huge lava fountains all across its lenght. Lava flows… bigger than UK .. and 50 meters thick swallowing up everything in their path.
    IF I time machine gets invented… I knows just where I wants to go lol
    Dark dreams of volcanic armageddon

  19. Of course… IF I gets what I wants… I will regret that
    This stuff is much much worse than laki haze

Comments are closed.