Sahara, Scotland

The conifers stand tall, straight. They look old but there are patches where trees have been cut, and there is replanting elsewhere, evidence of tree harvesting. The evergreen forest is popular with tourists. This is in spite of the latitude: there are more northerly places in Scotland, but not many. The climate is not as bad as may be expected, as the mountains shelter the coast from the westerly storms and rains. But summer may not happen every year. The forest is sandwiched between the land around Brodie Castle and Forres to the south, and the shores of the Moray Firth to the north. “Sandwiched’ somehow seems the right word: scrape the fallen needles off the ground, and sand appears, as if the trees were growing on a beach. It is a forest out of place, the kind you might find growing in a wardrobe. If these trees could talk, what story would they tell?

Nothing in Culbin Forest is natural. The trees are all planted, in a battle with the environment that lasted more than 200 years. Once this was thriving farm land. But one year, a northerly gale came and whipped up a sand storm. Hour after hour the wind blew and the sand kept coming. Nothing could be seen: people left their tools on the fields and fled. Eventually the storm ended, but it was too late for 16 farms and their laird. Over an area of 3800 hectares (38 km2), nothing was left. In places the sand lay 30 meters high. This was the year when the Sahara came to Scotland. It was in 1694.

The story is re-told in many places including the BBC. It sounds unbelievable: can there really be a sand desert in Scotland? The answer is complex: there was indeed a disaster and it did involve migrating sands, but the official history should be taken with a grain of salt. The disaster had roots: it did not come out of the blue. And while the sand storm was obliterating the Culbin estate, elsewhere a volcano was altering history.

Background, human

Scotland after the Middle Ages was a feudal society, where the land was owned by the estates, and awarded by the king. If the estate had a castle or manor house, it could be called a ‘barony’ and it’s owner a baron. The baron was considered ‘nobility’ but was not inherited: the title could be bought by buying the manor. This was unique to Scotland since elsewhere in Europe ‘baron’ was an inherited title which was not tied to specific real estate. In Scotland, the system of a barony awarded by the king was stronger in the east than in the west, as the authority of the king traditionally was stronger in the more hospitable east.

Another group were the tenant farmers. The most powerful of these had tenancies that could be passed on to their children. Over time, this group became as powerful as the barons, and together they became known as the ‘Lairds’, the ‘Lord’ (owner) of a large estate. This title brought voting rights in the Scottish parliament of the day (there was a minimum size to the estate to qualify). Whoever inherited (or bought) the estate would also get the title and the rights.

Below the barons and lairds, there were by-and-large three groups: the (non-tenured) tenant farmers, the crofters and the cotters. Tenant farmers were nominally independent but tenures were often short, typically for 5-10 years. This wasn’t as insecure as it may seem, as renewal was nearly automatic. The rent might be increased at renewal: some land owners were more supportive, other less so. There could also be special rents, payable for things like peat-cutting, collecting seaweed (used as fertilizer), etc. The size of the tenured farms varied a lot, and many were so small that several farmers would pool resources, for instance sharing a plough with a team (8) of oxes. For many, it was a life of poverty. Rents were often paid in-kind, especially with grain. During the late 17th century, there was a slow shift to rents paid as money, where the most recently developed farms would pay in money and older farms in-kind. Rent arrears were common, especially among those who paid in-kind. However, eviction was relatively uncommon. Rent arrears were often seen as a normal part of the ups and downs of the farming life. A fraction of the tenancies were held by women, often as widows.

A cotter lived in a cottage provided by the land owner, and would work the land for the owner in lieu of paying rent. This might be for something like three days per week, leaving him the rest of the time to earn his own living. Cotters could be villeins: these were nominally free but were not allowed to leave the land. This was almost a form of slavery, but had the advantage that they could not be evicted and so it gave some security. In Scotland the arrangement was normally a bit different, and obligations appear to have been limited to occasional labour service, seasonal renders of food, hospitality and money rents.

Finally, a crofter would own his (or her!) cottage, and rent a small piece of land, often as part of a community. Crofters play no part in our story.

A compilation of rent arrears shows the good years of 1665-1695, followed by the climate disaster. source:

In the coastal lowlands the agriculture was focussed on growing crops, especially barley, and wheat in the best locations. The yields were not high, 2 to 3 times the seed sown, which is near the limit of viability for farming and explains why rent arrears were common. A bad year could spell disaster, not just for the farmers but for everyone requiring food on the table. To mitigate, after a crop failure food would be imported from the Baltic. This was so common that there were Scottish colonies in the Baltic! Famines were regular until around 1650. After that, famines ceased, in part because of better farming methods introduced by those most hated of enemies, the English. The improved tillage methods were first introduced in the south of Scotland, and were very slow to reach the north. By the time the story begins, agriculture at Culbin was probably still traditional. And the improvements failed to prevent the last of the Scottish famines – the awful years of 1695-1700.

Story, official

The story is often republished, and can be readily found. The emphasis is invariably on the suddenness of the disaster that befell a prosperous community.

River Findhorn

The barony of Culbin was run by the Kinnaird family, and dated back to the 13th century. Their manor house was a stone-built square building, with a large garden and orchard. The accompanying land lay between the rivers Findhorn and Nairn. It was was rented to some 16 farms who did well, both because of the fertile soils and because of the river Findhorn which provided excellent salmon fishing: the barony held the fishing rights on the west bank of the river. It was a well-to-do community and the baron became wealthy.

The storm came suddenly, in 1694. It arrived so fast that the reapers and a ploughman had to leave their work and tools on the land and flee from the sand drifts. The sand attacked everything, from the cotter’s hut to the baron’s mansion. The weather calmed in the morning but the sand was so deep that the people had to break their way out of the houses where they had sheltered. Now the storm restarted, and everyone fled. The sand continued to flood in. It even blocked the Findhorn, which flooded the land until it found a new path to the sea, directly north.

When the people returned, they could not find their houses. Everything was covered under deep sand. Even the manor house and chapel had disappeared. The people left; the baron (and his wife, child and nanny) never returned. The land and manor were abandoned to a desert of sand. Over the years, the sands kept moving and tantalizingly would reveal a glimpse of the buried barony: a ploughed field, part of a ruin, a dove cote, or the chapel; once the chimney of the manor house appeared. Legends tell off the fright it caused among the people exploring the ruins. However, a bit of fear never stopped a Scotsman. Whatever the sand revealed was quickly stripped of any stones that could be used for building. The sand always quickly returned and covered the past up again. A hundred years ago, before the forest took root, Culbin still remained a sand desert.

The baron had lost his income. He petitioned the Scottish parliament not to have to pay the land tax, because there was no income. The barony was sold, but the family had lost its wealth. The last surviving member of the family died in 1743.

Map of the area, where the large green area is the Culbin forest (previously sand)

This is the official, i.e. oft-published story. There are some minor issues with it which I’ll come back to. We need go a bit deeper to explore the sea, the sand, and the people.

Background, Scotland


The sea in Scotland is a funny thing. During the depth of the ice age, sea level was a hundred meters or more lower than it is now, because of all the water locked up in ice. So you might expect Culbin to have been far from the sea. You’d be wrong. Scotland had grown some glaciers itself, effectively transferring the water from the North Sea to the land. And the glaciers were much thicker than the 100 meters lost from the sea. The glaciers weighed down the land, and pushed Culbin down by more than the amount by which the sea had fallen. Culbin was 50 meters below water, although the glacier itself may have kept the sea at bay.

When the ice melted the sea quickly rose and for a while Culbin became even wetter. Different ice sheets melted at different times, so that the sea level rise was not always regular. But of course, the removal of the glaciers meant that Scotland shed a lot of excess weight (the so-called ‘water diet’), and the land rebounded, in isostatic rebound. The lithosphere isn’t as quick as the sea: one acts like treacle, the other like (ahum) water. So initially the water rose faster than the land, but after a while the land caught up and eventually it began to rise faster than the water. Whether the sea level rose or fell around Culbin depended on which one rose faster at any particular moment.

The plot shows the results for northeastern Scotland. 15,000 years ago, at the location of Culbin, the land was 50 meters below sea level. However, it was rising fast. You can’t keep the Scots down. 13,000 years ago Culbin rose above sea level and it continued to move up in the world. It peaked 10,000 years ago, at 15 meters above sea. Now the sea began a new assault. 7,000 years ago, Culbin found itself again at sea level and shortly after, a few meters below water. Since that time, the land recovered, and rose a little, and the sea withdrew to its current location.

Sea level in meters with respect to the land, versus time in thousand years. Note that a positive number for sea level means that the current coast is under water. The situation in Culbin was similar to Inverness. Source: Kurt Lambeck,

The higher sea levels 6000 years ago left raised shorelines in many places around the Moray Forth. In low-lying areas, these can be a considerable distance in-land. The sea formed gravel beaches, and the ancient shorelines can be recognized by these gravel ridges. Many are now hidden underneath Culbin forest.


White Sands, New Mexico

The retreating sea left sand behind. Sand is as funny as the Scottish seas. It comes in an incredible variety, from the black sand of Hawai’i (which burns the careless feet) to the dazzlingly white White Sands in New Mexico (which, I can tell you, also can get very hot). Sand is defined as small particles, up to a few mm, but it can have almost any composition. Some is biogenic, consisting of marine-life’s skeletons. Colour doesn’t always tell what it is: the white beaches of tropical islands are made from the lime in the reefs, whilst the white of White Sands is gypsum and the whiteness of Florida’s beaches is quartz. The most common type are mineral sands, often quartz or granite.

Sand forms by battering rock into submission, through weathering, wind, waves, or ice. The sand contains whatever the rocks contained: even some uranium is not impossible. Wind and water moves the newly formed sand around. The North Sea is full of it, partly from the scrapings of Norwegian granite by the ice age glaciers. The waves deposit it near the shore, and once it has dried, the wind can pick it up. The dry grains blow low over the surface and if trapped by some obstruction, form sand dunes. If a large area is exposed during low tide, more sand can dry and more land can be covered by dunes. Scotland has 50,000 hectares of these sand dunes.

The character of the dunes depends on distance from the sea. The shore has no vegetation and lots of blowing sand. In-land are the mobile dunes, with a covering of marram grass which captures the blowing sand but which does not prevent the dunes from moving. Further in-land there is more vegetation and here the sand is fixed in place. But in general, sand shows that water is (or used to be) near.

The Moray Firth acts as a funnel: sea currents bring sand in, but the currents are too weak to take it out again, and the sand collects along the southern shore. When the sea rose, sand dunes formed on land and sand banks in the sea. When the sea retreated, the sand was left behind. And because the sea shore moved around so much, sand has been left everywhere around Culbin. The wind formed a large dunescape, covering the old gravel beaches. The coast here has long been prone to moving dunes: Boethius in 1097 AD already mentioned inundation of parts of Moray by sand thrown up during storms in the North Sea.

Locations of past and present gravel beach ridges. Source:

On-going ridge development at Culbin. Source:

Landsat image (google maps) showing a close-up of Culbin forest. The apparent, slightly curved lines are the gravel ridges, left behind by the retreating sea over the past 6000 years. They are visible because they affected the tree planting.

In this area, there was more going on. During the high sea level of 6000 years ago, the river Findhorn created a wide estuary, terminated to the north by a gravel bank. The bank diverted to river to the southwest, where it dropped its sediments. When sea level dropped again, the estuary became dry land, where soil and peat formed. When people came, they found good pasture. The barony of Culbin was located on this old estuary.

People, lost

The land around Culbin had originally been given to Richard de Moravia in 1234, founder of the Murray family. Later, the Kinnaird family joined by marriage: they were an equally old family, given land in 1184. Both families were originally Flemish. By the way, the Moray Firth is not named after the family: the name was adopted from the Pictish kingdom of Moray. The families did well, and the estate could support a wealthy life style. But this changed in the 17th century, and by 1660 financial issues appeared when the baron had to borrow money.

The area around 1680

The tenants and farm workers did not share in the wealth. Theirs was a life of poverty, with a total lack of luxuries. One of the problems was that the sandy coastal land lacked stone for building. The workers therefore build their houses from turf, whilst the mansion and the chapel were reportedly build from sand stone, perhaps sourced from Caithness. The turf huts didn’t last long, and often had to be repaired or rebuild. The people took the turf from the carse: the ancient gravel ridges where farming was not viable. The roofs were made from grass, and for this the marram grass on the coastal dunes was used. This harvesting was fine, as long as the soil, turf and grass could reform and regrow at the same rate as it was taken. As a form of recycling, the old walls and roofs were dumped at the place where the new was taken. But in the deteriorating climate of the 1600’s, regrowth no longer kept up with demand. The dunes here always had had a tendency to move around, using the plentiful sand. Around Culbin, it appears that the dunes had been stabilised by the vegetation, even at a time when dunes elsewhere were increasingly activating: this has been inferred from soil profiles. But the turf removal stripped out both the vegetation and the soil and opened up the sand, whilst the marram harvesting activated the mobile dune system. The dunes began to move.

According to the stories, the sand storm of 1694 came without warning. In reality, the warnings were there and they were recognized. Two previous storms had already brought sand, in April 1663 and in the autumn of 1676. From fragmented records, it appears that the peat supply on the Culbin estate had been lost some years before, perhaps covered in these earlier storms, and the laird (a bit of a scoundrel) had resorted to pillaging the peat of his neighbour. This, of course, caused a fair amount of tension, not helped by the fact that the Kinnairds were catholic and the neighbours protestant, at a time this difference meant something. But the sand didn’t care about beliefs and morals. To the west, the inhabitants of the town of Findhorn were already preparing to relocate their village. To the east, the town of Nairn was being threatened by drifting sand. Here, the local council ordered that the stripping of turf should be stopped. But the rulings were not enforced, the dunes were on the prowl, and in the big storm of 1694 a large dune covered most of the Culbin estate. The following year, when Culbin was lost and the baron petitioned not to have to pay property tax on his lost property, an act of parliament was passed to prohibit the removal of grass from the dunes. This was too late: the horse had bolted.

The land never recovered and the sand remained without vegetation. Up to the early 20th century Culbin remained a sand desert, a Scottish Sahara, the largest such area in the UK. Early attempts to stabilize the dunes by planting failed. But clearly, it was unacceptable to have migrants dunes wandering around stealing jobsloitering with intent threatening surrounding areas. Tree planting began in 1922, with the added economic advantage of providing 20,000 m3 wood per year. The planting was completed only in the 1960’s. Nowadays, we would have considered the sand dunes a unique landscape, and (perhaps) protected it. But in those days, people saw no value in sand – it was just a nuisance.

Facts, false

in the popular stories about Culbin it is said to have covered an area of 3600 acres (about 1500 hectares), with 16 tenant farms: a large and populous estate. Both numbers are incorrect. The number of 16 farms comes from a document that actually lists 16 tenants, over only 6 holdings.

The size of 3600 acres comes from a document from 1866, but it refers to the area of the sand dunes, and these covered more than just Culbin. For some of the holdings we know the size from older documents. Putting these on the map, and adding the rough ‘carse’ of the gravel ridges, shows that the Culbin estate was around 750 acres, of which less than half was useful for farming.

The second false fact is that it was lost in a single, dramatic storm. But in reality the sand had been encroaching for some time. The petition made in 1695 for tax relief stated that over half the estate had been covered with sand. It was this petition that started the legend of the sand storm and the buried barony. And the text of the petition was dramatic: “where the best two parts of the estate of Culbin, by an inevitable fatality, was quite ruined and destroyed, occasioned by great and vast heaps of sand, which had overblown the same, so that there was not a vestige to be seen of his manor place of Culbin, yards, orchards and mains thereof, and which within these twenty years was as considerable as many in the County of Moray, and the small remainder of his estate which yet remains uncovered was exposed to the like hazard, and the sand daily gaining ground thereon, wherethrough he is like to run the hazard of losing the whole.” The plea convinced parliament and the baron was let off for two thirds of the tax. A few years later, shortly before his death, he again petitioned parliament, stating “having the three parts of his own and his predecessors’ estate overrun with sand, and the fourth part yet remaining is sold for payment of his creditors, as far as it will extend, and having nothing but to recur to his frugality and industry for a living in time coming.” This shows that part of the estate had survived the storm. The 1694 took a big bite out of the area, but it had already lost some before and would lose more later. The 1694 onslaught was probably when the manor house was lost, and which tipped the finances over the edge. But these finances had already been deteriorating since the 1660’s. Was that because of the sand and the loss of peat? Because of wasteful Lairds? Or was there an impact from the improved agriculture further south which had ended the famines and made grain more affordable – i.e. less profitable? We don’t know.

Did the storm actually happen? There are other reports of a major storm around the North Sea basin in July 1694. However, some of the modern stories give late October as the date (it is not clear on what basis) and we have no other reports of storm damage from that time. But going back a bit in time, in the late 19th century a subtly different view of the events ( existed. This told that the sand had first come in 1686 when a large dune approached. Over the next ten years this dune marched over the estate and swallowed farm after farm. This version makes some sense knowing how dunes move, but none of the old documents that we have mention it. The estate was taken over by a new (younger) Kinnaird in 1691, and at that time it must still have been (or appeared to be) viable. We just do not know all the details and some story telling is inevitable.

Winter, volcano

But this is not the end of the story and the storm of 1694 did not stand on its own. Events elsewhere in the world would shortly be making any recovery for Culbin difficult. After half a century of good times, bad times were coming. The period 1695-1702 would become known as the “seven ill years”. After decades without famines, now there were four failed harvests in quick succession (1695, 1696, 1698 and 1699). In the past, Scotland could obtain grain from the Baltic, but now that area too suffered from failed harvests. Famines came, and the people were unprepared. As many as 20% of the population of Scotland died. This was the last major famine to hit Scotland. (The European Drochshaol (potato famine) 50 years later, which so utterly devastated Ireland, also affected Scotland but to a much lesser degree). It hit hard.

And the problems were not limited to Scotland and the Baltic. The winter of 1695 was one of the coldest and snowiest on record, not only in the UK but also elsewhere in Europe. In France, the big complaint was that in the Palace of Versailles the wine froze in the glasses. Several very cold winters followed, with sea ice reported 20 cm thick around the UK. Estonia was hit hard: the summer of 1695 was too wet for the crops, and an unseasonably early frost in autumn ruined the harvest. Here also the winter was extremely cold and lasted until May, and another wet summer followed. In the great famine of 1695-1697, 20% of the population of Estonia died. In Finland too it became known as a year of many deaths. Further afield, Iceland was hit, with sea ice surrounding the entire country and remnants of the ice staying throughout the following year. Asia did not fully escape: in 1695 the southeast monsoon was drier than usual and this became a pattern that lasted a decade.

An ice-cold Atlantic may have contributed to the damaging frost of September 1695. In the UK, it began with a storm, after which the wind turned north: this suggest a low pressure system in the North Sea. The wind borrowed from unusually frigid air further north: the northern Atlantic ocean was ice-cold. The traditional explanation is that the 1690’s was when the Little Ice Age hit its peak. But that does not explain the runs of extreme weather – it just gives it a name. Did something amplify the changing climate? What had made the North Atlantic so cold?

There have been other times when the weather became exceptionally cold for a few years. If you plot average temperatures over the past 2000 years these show up well. Some care should be taken with this: there were no actual measurements done (thermometers being a recent invention) so indirect indicators are used, such as the composition of snow (amount of the heavy isotope of oxygen) as measured in ice cores, or the width of tree rings. The plot below shows an example, where the long-term variations (a cooling trend in the middle ages and the pronounced recent warming) have been removed to make the fast fluctuations stand out. It turns out that these fast fluctuations are always of colder weather, not warmer. And the largest drops are associated with major volcanic eruptions: 536, 1257, 1453, 1815 (note that the cold weather lags by a year). How about the smaller fluctuations? That is not known, but it is notable that the 1695-1700 drop is one the largest of these less major drops. Could it be ..?

Ice holds the answer. It faithfully records large eruptions, in the form of sulfur spikes in the annual layers. I made the plot below from the data in Michael Sigl et al, 2015 ( Red (lower line) is for Greenland (northern hemisphere), and blue (upper line) is for Antarctica. I have shifted the Antarctica line upwards, by the way, to avoid confusion. You will recognize the volcanic spikes, and in some cases will know the responsible volcano. The huge spike in Greenland in 1783 is of course the eruption that (I remember Henrik telling us) caused the French revolution. And indeed, there is a clear spike at the start of the cold period at the end of the 17th century. It wasn’t just the Little Ice Age: it had help.

There seems to have been a northern eruption which shows up in the ice from early 1693, which would have happened in the winter. Was it in Iceland? Jon Friman, in his marvellous compilation of Iceland’s historic eruptions (, mentions a damaging, ashy eruption from Hekla beginning Feb. 13, 1693, with ash going northwest – towards Greenland. The real firework came a year later. The sulphate in the Greenland ice cores began to increase at the end of 1694, and it peaked in the summer of 1695. In Antarctica the increase began a few months earlier, in the autumn of 1694, with the largest peak in the autumn of 1695; here it took until 1697 before the sulphate was down to its usual level. The fact that it is seen in both hemispheres shows that it was from a tropical eruption, and although not a Tambora, it was significant.

The evidence for the temperature drop and for the tropical eruption are strong. But are the two related? That is always difficult to proof! It wasn’t a massive eruption, and no one complained about the Sun being faint or red. The typical signpost of a volcanic winter is snow in summer, and that indeed happened in 1695 (counting September as ‘summer’ – in Scotland, the first month of autumn is usually August). I would argue that in itself, the eruption would have had a notable but fairly minor effect. However, the north had already cooled to dangerously low temperatures as a result of the Little Ice Age, which appears to have been the onset of the real ice age. In northern Canada, permanent snow cover had begun and everywhere glaciers were growing. In this precarious situation, the volcano erupted. It pushed things over the edge. But not irredeemable: after a few years, the climate did recover.

What about the sandstorm of October 1694? This appears to have come before, or at best simultaneous with, the eruption, so it is hard to blame it on that. The 1693 Hekla eruption was also too small to provide a helping hand. The storm just happened. We don’t know how severe the storm really was. Storms in Scotland get their energy from the temperature contrast between the warm side and the cold side of the polar front. A cold northern Atlantic increases this contrast and should strengthen storms. Indeed, storms were more frequent and stronger during the Little Ice Age (and conversely, they have diminished in the last 30 years as global warming took hold). (This is not true for the subtropics where the storms get their strength from the warm sea water – here the number of storms has not increased, but their strength has.) Anything that would have cooled the northern regions could have increased the chance of such a storm. And that is what the Little Ice Age had done. But either way, it was inevitable that eventually, a bad storm would come. It is like that friendly neighbourhood volcano: you should be prepared for what every bone in your body tells you won’t happen. One day, it will. The most dangerous factor in disaster prevention is the human tendency to take the future for granted. Even though people had seen the warnings of the encroaching sand, and knew they needed to stop taking the cover off the dunes, in the end the choice between a living and a high risk of failure was easy. People needed houses and fuel, and the future would have to take care of itself. It did.

And so the sand came, and the volcano came, and the people lost their homes, land and living just at a time when famine was about to hit.

Trust me – it won’t ha..

Words, final

The story of the Culbin sand disaster is legendary. It may not quite have happened as legend has it, and we will never know all the details, but it is well worth remembering how this part of Scotland was lost. In itself, it is a local story about human vulnerability and loss. But there was a much wider context. As Culbin was lost, the Little Ice Age and an unknown volcano came together to bind northern Europe into years of winter, and cause major famine in much of the region. The people of Culbin lived a small story within a much larger event. Isn’t it always that way? Nowadays, we would also point at the power of nature, and would see the formation of the largest area of sand dunes in Great Britain as a treasure coming out of a disaster. But in the past, people did not see it that way. They just saw sand, and they planted a forest instead. And so the treasure that kept the memory of 1694 was lost. And that, in the end, is the story of the trees.

Albert, April 1695 2019

212 thoughts on “Sahara, Scotland

  1. Excellent reading Albert.
    I have some limited knowledge of the area, having visited the locality twice back in the 80s. I had noted (being a keen horticulturalist) that the soil was pretty much 100% sand, and I wondered at the time how the pines could flourish there, as I imagined that the water table would be of brackish water, being so low lying and close to the shoreline.

    I know of another area which has a claim to a similar history and fate as Culbin. That is Kenfig, Between Bridgend and Neath on the South Wales coast. Apparently the dune system there conceals a medieval town, although I don’t know any further details.

    • Believe Time Team may have dug at that village … or one with a similar tale.

      Thanks for the post Albert! Great read 😀

      • Possibly so.
        I know Kenfig because there are good waves to be had there when conditions are right. I heard about the (possible) archaeology from local surfers.
        I knew Mick Aston briefly… Friend of a friend… Unfortunately that was before I discovered the waves in that part of Wales.

  2. That part of the world is extremely milded by the Gulf Stream
    I seen palms and other subtropical plants planted in gardens in Scotland in Google earth at latitude 57.

    Similar latitudes in Sourthen Hemisphere is South Sandwich Islands
    Without the Gulf Stream they haves an oceanic polar climate
    Glaciated down the sealevel at similar latitudes just in sourthen hemisphere.

    Örnsköldsvik in Sweden is similar latitude as Antarticas Northen Penninsula IF I was in Sourthen Hemispehere

  3. Douglas fir forest? Some years ago I heard a lecture by a man from Kew gardens who complained that our beautiful Pacific northwest forests looked just like a Scottish forestry commission plantation. I like the evidence that a mid level volcanic event coinciding with a cooling trend can precipitate disaster.

  4. NW UK, between the isolated red squirrels of Formby and the wide sands of Southport, there’s a very vulnerable coast. Formby Point was badly affected by Victorian and later ‘training walls’ in the Mersey approaches, but its exposure plus a nine (9) metre tidal range means it gets hammered. Beyond beach losses of several metres per storm, the dunes naturally migrate inland at approx four (4) metres per year. The ‘National Trust’, who maintain the fragile inland habitats, are resigned to losing ~400 metres this century, while dealing with storm damage and erosion revealing foundations, Victorian rubbish dumps etc etc. And, yes, a lot of sand gets blown onto those inland habitats…

    Just North of Formby, there’s a very handy ‘relief’ road just behind the big dunes, by-passing Southport’s sprawling outer suburbs. Been a few years since the last big ‘blow out’ when, despite grassing, bushes, fencing etc etc, several hundred yards of road was deeply buried. Heroic re-planting, re-fencing, mesh etc seems to have stabilised those dunes, many with their toe creeping onto the road…

    There was near-panic after a grass fire there, as losing the ‘stabilisers’ could have let all those big dunes walk. Happily, the water table in the dunes stood high enough to prevent a ‘deep burn’, avert catastrophe…

    Long term, the sea level is rising. If that part of the coast does a ‘Southport’, a wide, wide beach will accrete. If a ‘Formby’, with erosion tipping the balance, the road would need to be built up as a sea-wall, the scrubby area between it and the inland developments’ low levée left as brackish marsh.

    Very long term, it is a low, soft coast, with little ‘terra firma’ above ~50 metres for too many miles inland…

    • Really enjoyed the article, Albert. Regarding the Oregon dunes, I imagine a BIG difference between them and the Scottish Sahara is the(many?) tsunami’s that must have had an influence on the sand distribution in Oregon.

      • Yep the dunes area bit more resilient than you might think . It is the inlets, valleys and bays around the dunes that are the problem.
        Erosion can undermien thewhe teh water ro a Tsunami is running in and back out…
        There are valleys in between the dunes that have a very warm microclimate.
        However you are constantly dealing wit the sand-it is always there..

  5. Ahhhh! a lovely post. Read part of it early this morning as i was chasing sleep and losing. (old people have trouble sleeping) and i love the way Albert writes human history backed by outside sources. Makes one look around and wonder what could happen locally. Thanks, Albert! Best!motsfo

    • “old people have trouble sleeping”

      Yer kidding right? My issue is getting into actual “sleep.” Once I’m there, I’m a log. I’ve even slept through missile launches before, and my “pit” was directly in front of the aft missile magazine.

      And a distraction for anyone interested.

      18 Jun 1178

      • I have narcolepsy.
        Somebody tell me again how hard it is for older people have trouble sleeping?
        It’s just taken me 2 hours of ongoing sleep attacks before I could actually stand up 😀

  6. Nice… when people become bothersome with end time stuff i just tell them to start worrying about real problems… like meteroites… shuts them right up and they wander off with a dazed gaze…. (bad motsfo, bad girl)… Well Easter Confession is coming up and i always like to have something concrete and specific to confess. 🙂 note: don’t get behind little old ladies in line; they always take the loooooongest. 🙂 😉 Best!motsfo

  7. A very distracting and easy to read article Albert. Thanks! I had never heard of the storm and dune tale, though have visited Nairn a few times in ancient days when I was a youth.

    • Good programme on BBC four last night . Natural world where they led a small expedition to Mount Nyiragongo. Stunning scenery and yet such a potential for devastation in the near city. But the city needs the volcano to thrive on.

  8. Nebraska was covered in Sand Dunes very recently in middle ages according to climate data
    The climate was drier.
    The Sand Dunes are still visible in the landscape covered in grass as it rains more now these days
    Nebraska grassy Dune seas are famous

    • That would be the Sandhills region of Nebraska. I’ve been through there. Very beautiful and interesting landscape. Grass covered dunes with many small lakes and wetlands in the low areas.

  9. What’s an “Other Event”

    2.5 Other Event 6km ESE of Pahala, Hawaii 2019-04-02 19:32:42 (UTC) 20.2 km
    2.4 Other Event 10km SSE of Pahala, Hawaii 2019-04-02 19:29:03 (UTC) 43.5 km
    2.3 Other Event 12km SE of Pahala, Hawaii 2019-04-02 19:27:22 (UTC) 38.3 km
    2.3 Other Event 14km SSW of Pahala, Hawaii 2019-04-02 19:24:19 (UTC) 46.6 km
    2.2 Other Event 12km SSE of Pahala, Hawaii 2019-04-02 19:21:17 (UTC) 39.1 km
    1.9 Other Event 9km SSW of Pahala, Hawaii 2019-04-02 19:14:40 (UTC) 41.4 km

      • Speaking of Mauna Loa, Tina Neal did a radio interview and it sounds like the alert level might go back up fairly soon and HVO is considering the small possibility of it taking over the hotspot again- at least somewhat plausible with Kilauea being relieved of stress and will be rather empty for a while (probably. Or more likely, it could loosen ML’s summit just enough for this renewed unrest to actually lead to an eruption this time, I would guess probably 2020 if it does go based on final buildup to the last two eruptions. I do get the feeling we might have a 1950-like scenario which would certainly dominate the headlines…

        • I watched that too, very interesting and a lot of things to study. Also I think maybe the most important part to working out how the events of last year is when she said some of the magma from fissure 8 was from really deep, which would fit my hypothesis that the may 4 quake was what allowed a deep dike to move at the base of the mountain, maybe as far as heiheiahulu where it rose and joined the dike from pu’u o’o that was shallower and happened fast and with a lot of quakes.

        • I would also suspect that the current alert level situation will be a fleeting and infrequent one too, with 0.25 km3 of magma coming from the hotspot every year it is basically impossible for the big island to be non-eruptive for more than a few years. At 0.2 km3/year it will take only a few years to fill up any spaces within kilauea created last year and that us just the base rate if a big new pocket of magma arrives like in 1959 or after the previous caldera collapse (in the 1780s some time, the date may well be revised in future studies) that rate it will probably briefly go up to several times higher and that definitely wont go unnoticed. It is only in the past 2 years that it was discovered most of the southwest rift has been covered by lava in a 30 year period, and in that same publication last week there is a quick but important mention of an observed eruption on the middle east rift in 1794 and apparently this is in agreement with oral tradition, which throws a spanner in the current working model…

          This information as well as the general assumption that the keanakako’i tephra was erupted in the year of 1790 (to my knowledge this has never been completely confirmed nor has any of the LERZ before 1840) suggests that even in the so called historical period the volcanic history of hawaii is generally poorly known until the start of the 20th century when HVO was built. Only since about 1840 has the history been decently well recorded and mostly for mauna loa. Western science hasnt seen anywhere close to kilaueas full potential and that is a problem because it is the hotspots preferred outlet 90% of the time…

          • With 0,2km3 every year for Kilauea and Mauna Loa likley more 50 000 years ago
            Its possible to substain slow long lived tube feed pahoehoe eruptions
            That coud last 100 s of years each
            Mauna Loa had slow lake tube feed pahoehoes that lasted 100 s of years each as most of Mauna Loa is covered by ancient pahoehoe in Google Earth

            High volume fissure open Aa channel eruptions are the current phase of mood for Mauna Loa

          • There is some very recent research on the keanakako’i tephra after 1790 and before 1823 by Mr. Don Swanson.



            Personally, I think that the “Mystery Unit” above the “Footprints Ash” may also be post-1790 and pre-1823.

          • I want to know more accurate dates of eruptions in the lower half of tge east rift, beyond puu oo. At the moment only eruptions from 1840 onwards are known well. There is the 1790 eruption too, whuch followed on from the heiheiahulu eruption in about 1750, and also the kapoho crater and puu honuaula eruptions, all less than 400 years old. Problem is that none of these is well dated, despite ease of access and risk assessment.
            Heiheiahulu was like puu oo, a long lived pahoehoe shield eruption, and 1790 was a fast eruption, very similar to and almost in an identical location to last years eruption. And also apparently with an entire other fissure line on the north side of the rift. Puu honuaula was also another huge fast eruption that happened probably in the early 1700s, covering an area comparable to last years as well as a large cinder cone. I find it hard to believe this wouldnt have made a caldera. Likely it nade a caldera that was similar to the one now but maybe a bit shallower and wider, and this persisted as the rift activity continued in preference to the summit, with further eruptions creating kapoho crater probably some time after puu honuaula maybe 20 or so years later. Eruptions also occurred in the middle and upper east rift, now entirely buried by puu oo and other recent eruptions, but some of these are quite large and cover significant areas of the volcano. Then some years later again, likely in the latter part of the 18th century, heiheiahulu erupted, initially as a widespread voluminous a’a flow field then a pahoehoe shield that probably erupted for at least a few years. The northern rift of the 1790 eruption goes right up next to heiheiahulu, and ages seem to overlap enough to be variable as to which eruption the flow actually belonged to, so this us pretty good evidence these eruptions were close in time, so heiheiahulu was very likely terminated by the northern branch of the 1790 eruption, which probably actually happened in the 1770s to mid 1780s. This might have also had summit collapse, which in the already deep caldera subsided at least part of it below the water table despite the erupted volume being quite small compated to last year, probably well under 0.2 km3. The 1790 explosive eruption was likely an eruptive event like 1959, a high fountaining eruption fed by a huge surge of deep magma and when this erupted initially it was not big but after a vent erupted in the deep part where a lake formed all hell broke loose. This is supported by observations too, apparently kilauea was already erupting before Keoua and his army made their camp near the rim, but it got very violent soon after and then the main stage happened. Other small eruptions happened around the caldera too, and eventually after the lake was gone the eruption probably continued. By analogy to 1959, the second southern part of the 1790 LERZ eruption might have been like the 1960 eruption, following this big summit eruption, and then maybe another small eruption in the middle rift in 1794, after which activity began filling the caldera rapidly and also erupted out of the southwest rift numerous times before 1823 when things are better recorded.

            Dont know if this interpretation will stand the test of time at all but it is interesting to think about nonetheless. All it really highlights in that the dates of eruptions on the east rift really need to be refined, as that could give a better idea of what to expect in the near future.

          • Turtlebirdman
            I and a few friends runns a Kilauea group on FB its my fav volcano too.

            we love kilauea volcano
            Is the only hardcore Kilauea group on FB you are very welcome to join it. It was created in 2016 and is highly active

    • Yeah, Squonk, I have the same question about the “other event” identification that U,S.G.S. ascribes to the quakes south of Hawaii.

    • Squonk, found this on the website.
      Data Type
      Typical Values
      “earthquake”, “quarry”
      Type of seismic event.

      It does not list other, but if I remember they used other for the flank collapse events during the eruption. Need to find out if tremor is one of the “others”


      • I see they have gone back and re-categorised earlier events as “Other”. Starting with

        M 2.6 Other Event – 18km SSE of Pahala, Hawaii
        2019-03-27 19:54:34 (UTC) 19.049°N 155.431°W 43.0 km depth

        • I did a different search and added volcanic eruption and volcanic explosion to the chart and make the range in magnitude 2.5-6.0. Date range from Jan 1 2018 till today and it does show the “flank collapse events” now categorized as “volcanic eruptions.” Now show 85 events but still only 23 of the tremor. Maybe they were not differentiating the tremor signals in the past?


          • Well, unless they go back further again USGS/HVO seem to be suggesting something new occurring. Here’s the 27th March which you posted on Carl’s “Deep Quacks” currently matching the earliest “Other Event”.

          • This just posted on Hawaii tracker Facebook page –

            Deep Earthquakes Around Pahala This Week, What They Mean

            This past week Tracker has received dozens of questions regarding the recent earthquakes deep underneath the Pahala region of Ka’u, and asking why on USGS’s earthquake map are some of the events marked differently than others? Deep quakes around Pahala are common, but this last week there was a small influx in seismic activity for the area, but the confusing part for me was the diamond markers placed on USGS earthquake maps, each denoted as an “other event.”

            To answer this question I reached out to geophysicist and seismologist Brian Shiro from HVO to explain dynamics of the recent earthquakes, the insights he provided are worth sharing;

            “These are deep volcanic tremor events and are common in the deep (~40 km) area near Pahala, Hawaii. This is where scientists think the mantle plume comes up, which ultimately supplies magma to the volcanoes. It’s an area long-known for this type of seismicity. Nothing unusual is happening now.

            “The reason these events show up as “Other Event” on the USGS Earthquakes website is due to a database schema limitation. The QuakeML schema, which the world’s seismologists have agreed upon, doesn’t currently include a “volcanic tremor” event type, so when HVO submits events with with that type, these get transposed into “other event”. The USGS Earthquakes website plots any non-earthquake event with a diamond symbol. You might recall that the near-daily M5.3 Kīlauea caldera “volcanic eruption” collapse events from May-August 2018 were similarly plotting as diamonds.

            “These events are also plotted on the HVO website, but that website doesn’t distinguish symbols based on event type. All are circles there.” ~ Brian Shiro

            HVO earthquake map:…/…/hvo_earthquakes.html

            Brian also explained that historically there have been plenty of examples of similar, low frequency events including tremors present in the area deep under Pahala which get mixed in with more normal high frequency quakes. The endurance of the deep tremor and earthquakes beneath the Pahala region point to a persistent source for magma supply to Kīlauea, “though other pathways may, and likely do, exist.” Wech and Thelen (2015).

            For more information, see Wech and Thelen (2015, or Aki and Koyanagi (1981,

    • Found the other types of “other”, but no tremor is listed. If you do a custom search for events on the USGS site you linked from it gives you the option of looking for any of these “type events”. I guess volcanic flank collapse and tremor are not yet on the list. I did a search for the last month and saw 20 events, since the new year only 23.
      Event Type
      Acoustic Noise
      Acoustic Noise
      Anthropogenic Event
      Building Collapse
      Chemical Explosion
      Chemical Explosion
      Experimental Explosion
      Ice Quake
      Induced or Triggered Event
      Mine Collapse
      Mine Collapse
      Mining Explosion
      Mining Explosion
      Not Reported
      Not Reported
      Nuclear Explosion
      Nuclear Explosion
      Other Event
      Other Event
      Quarry Blast
      Quarry Blast
      Rock Burst
      Rock Slide
      Rock Burst
      Snow Avalanche
      Sonic Boom
      Sonic Boom
      Volcanic Eruption
      Volcanic Explosion

  10. Nice article Albert!
    Main reasons things went wrong in dry sand conditions in the Netherlands (coastal, but also inland) were the combination of grazing cows, horses and/or sheep in high numbers and sod cutting (farmers used traditionally sods mixed with animal droppings to fertilize fields). Once sandy soils get caught by wind, organized action is needed to stop fast extension of a dune landscape. In fact, it was the reason our State Forest Department was established in 1899.
    You need not more than 4-5 beaufort, so far less than a stormwind (9 bft), to let wind transport sands.

    I am wondering though how the turf digging could affect the dune forming. Were these bogs formed in an era before isostatic rebound took place? Or did the sand already cover bogs and were ‘dug up’ to use the turf? Actually this did happen in the Drenthe province, doing this contributed to a large dunecomplex!
    And ehm, I couldn’t find a decent translation of the word ‘carse’. Could you please clarify?

    • There was sand under everything, Carse is a scottish word, which I had assume stood for ‘coarse’ but seems to have meant grass growing in wet places. The grass must have been good but it was not useful for wheat. That is how it was used in the documents I read. If someone would like enlighten us, please do!

  11. Amazing how quickly isostatic rebound can happen. And is still happening in the Norwegian, and Swedish areas.

    • Don’t forget also Canada and the northern US, including Alaska! Canada has some of the highest rates of post-glacial rebound, especially around Hudson and James Bays.

      • IIRC, the future of Hudson Bay is quietly, but bitterly disputed. Also, IIRC, Hudson Bay is rising unevenly, compounding modelling complexity…

        I’ve seen suggestions that the current uplift rate will NOT taper off further, as deep inflow will now provide ample material for continued isostasis .

        I’ve seen suggestions that it may not only become ‘high and dry’, like Northern ‘Gulf of Bothnia’, but second order effects will spill the Great Lakes South into the Mississippi Catchment, reversing the Chicago River.

        ( Or, if you wish to pitch it thus, stealing its head-waters, as the Amazon is doing to a river near the Andes. There, there’s already a link, and the flow is gradually increasing…)

    • Where I live it is still 4 mm per year. Enough to make us update sea charts every 10-20years..

  12. Grimsvötn is also showing signs been more quakes there recently
    This volcano wont stay quiet for long

      • Turtlebirdman whats your musings on Grimsvötns future?
        Im soure Vatnajökull will become intresting as the Iceland Hotspot grows in strenght and also the North Rift Zone becomes dominant.

        What I think will happen:
        Grimsvötn will continue to rift build itself and form New calderas and as its done previously
        The arera may remain highly active for a very long time indeed and numerous calderas will come and go and plinian events and flank lava floods comes and goes. If magma supply continoues to be this profilic as it is today and if it gets even larger in the future.. Grimsvötn may enter phases of long lived activity.
        Large basaltic plinian eruptions like 2011 and 1870 s will likley be more frequent.
        The most crazy woud be a Laki sized or Thjorsahraun like event inside Vatnajökull itself
        With long lines of 2011 ash colums in icesheet and later huge lava fountains in Glacier

        I Imagines large eruptions growing out lava shields in the icesheet very much like Surtsey but grows out of the icesheet after the initial subglacial pheratomagmatic phase.
        No recent eruptions in Grimsvötn have grown beyond glacial surtseyan phase all stopped before.
        It have certainly happened before in Grimsvötn subglacial tuyas growing out of the Icesheet and lava flows Pour down into the icesheet.
        A large Grimsvötn eruption can become effusive as the cone grows out of the icesheet.
        That woud likley look like Veniaminof eruption.
        But more like a Puu Oo thats grown out of the icesheet.

        • I dont think it will do lava shields the way it is now. Bardarbunga does shields but it is also much less active and I think that is important. Grimsvotn is basically icelands kilauea, a young and extremely active volcano that receives magma at a rate about an order of magnitude faster than other areas. Grimsvotn will most likely become something more like a flood basalt than a lava shield after the ice melts, a flat and gently sloping lava plateau that has a caldera in the middle and large rifts, with relatively infrequent but really big lava eruptions. Shields appear to form mostly when the volcanoes dont have much of a shallow system, either when the volcano is new or when it is dying, otherwise the magma just collects in a magma chamber. A lava shield might form next to katla where that cryptodome is, or maybe just northeast of grimsvotn but not on grimsvotn proper.

          Grimsvotn will probably be a very scary volcano in the future, the dead zone rift seems to have only become active along its southeast side where laki and eldgja are during the middle holocene onwards, so this area might only just be getting started. Even though 1783 was part of the grimsvotn swarm neither katla or bardarbunga look capable of doing anything close to a big rifting eruption of that scale this time around and grimsvotn seems to be rather a lot more active in general.

          • I don’t think it can do another Laki. The crater is still too deep (although, I expect, not as deep as it was in 1800.)

            The precse location of the MAR can move around a bit. Once a volcano forms it will slowly drift away from it, and after a million years it may be too far to benefit from MAR magma. Grimsvotn is still close to it, possibly even closer than Bardarbunga. Speculation: Bardarbunga operates mainly from its own magma chamber, grimsvotn operates from a small chamber but with an almost (the ‘almost’ being very important) open connectiob to the deeper reservoirs of the MAR.

          • Is the exact location of those deeper reservoirs known? under Vatnajökull? under the Dead Zone? or a particular volcano?

          • Well i dont necessarily mean that grimsvotn will do another laki, more that I think it will do eruptions of that style instead of shields. Things around 0.1 km3 up to maybe 5 km3 probably about 1-5 times a century with the bigger events being more rare, but probably on average the eruptions woukd have to be quite big to happen in this setting. I guess this would probably make a shield shaped volcano anyway but it is not a pahoehoe shield like puu oo or skjaldbreidur it would be a sort of ‘flood shield’ and I dont know if there are any good examples that are well studied, I think the volcano next to hekla might be similar though. The very biggest flows would still be from the fissure swarm like 1783 and would be associated with caldera collapse or subsidence at the main volcano either grimsvotn or thordarhyna, as well as probably a temporary reduction in eruption size and localisation to caldera filling, which might be the closest it gets to slow shield eruptions. It will be very interesting in 200 years as none of the active central volcanoes bordering the dead zone is subaerial today and as far as I know they never have been since settlement so we havent seen vatnajokull go effusive very often, only in 1783, 1864 and 2014 in recent times when eruptions occurred outside the glacier, but if that is anything to go off if it says these are some serious volcanoes…

          • Turtlebirdman
            Have in mind that Grimsvötn haves a very large supply
            Of around 0,4 to 0,8km3 every year from passive spreading decompression and alot from the Mantle Plume combination.
            The hotspot supply may resemble Kilauea but its likley much lower in supply. There is 5 volcanoes under vatnajökull and 2 outside that shares of the large deep supply and there is the passive rifting that steals and decompresses alot of magma. There is 7 volcanoes in Iceland that feeds directly from central plume supply. Grimsvötn, Bardarbunga, Kverkfjöll and Askja are purest plume magmas in Iceland in compostions.
            And there is the Iceland Mantle Plume directly under there.
            Iceland is pulled apart 0.9 centimeter per year at Grimsvötn. When it spreads the magma decompresses that rises the system mainly via the mantleplume, but partially alot from MORB-processes.
            Grimsvötn is the heart of the Icelandic Hotspot.
            The Hotspot and Mar spreading both decompress and makes magma.
            Each of the vatnajökull volcanoes may have 100 s of km3 of liquid magma thats not eruptible.

            This tells how a 500 cubic kilometer deep chamber can form under the Grimsvötn volcanic system. Under an area of 200 km long 60 kilometers there is around 800 cubic kilometers of basaltic magma.
            Temperatures deep down is around 1190 C and its around 1510 C in partial melting zone in the astenosphere source, even deeper down.
            This magma reovair feeds hot open pathways and Grimsvötns much smaller upper magma chamber and Thordarhyrna, Kverkfjöll and partly the Dead Zone. When these upper magma chambers drains we gets the 10 to 20km3 events and calderas in the central volcanoes.

            Most of this magma can never erupt and is doomed to become gabbro.
            When Grimsvötn dies in millions of years there will be a giant gabbro batholith like thing there.
            The very large rifting events coud partialy tap from this upper parts of this magma resovair. But most of this cannot erupt.

          • That flood shield Turtle is talking about
            Is vatnafjölls thats east of Hekla where large km3 basalt flows piled on eachother. More than two dozen eruptions have occurred at Vatnafjöll in holocene era according to geological data.
            Vatnafjöll last erupted around time of settlement.
            This is a place where holhuraun sized events and much larger have happened many times in same local arera.
            Each one of these episodes is called a fire

          • Grimsvötn easly may do another Laki
            Laki like events are not feed from the upper magma chamber
            but from deeper down
            Grimsvötns main eruptive styles are as Turtle says
            small to medium sized to large intra caldera surtseyan and plinian and subplinian eruptions thats goes from VEI 4 to VEI 6
            and very large flank lava floods that can exceed many km3 ( VEI 5 to VEI 6 )

          • Without disagreeing on the large magma amounts underneath iceland, a large fire also needs a mechanism for bringing it to the surface. The weight of the overlying mountain can do a lot, but only if it is in the right place and much higher than the location of the final eruption. I think that aspect is till missing for grimsvotn. it doesn’t matter whether the source is deep or shallow: you need the stress regime with high local stress and a weakness along the rift.

          • @farmeroz ….. “volcanoes do just what they bl**dy well want to do”.

            I believe that the great philosopher T.Swift would also agree.

            In one of her seminal works, she concludes that the players gonna play play play play play and that the haters gonna hate hate hate hate hate.

            In that vein, Volcanoes gonna volcane volcane volcane volcane.

        • Grimsvötns south caldera where most eruptions have taken place..
          Is where the smaller shallow magma chamber are that may contain a few km3
          This is shallow just 2 km below and explains the extremely high geothermal actvity in the subglacial lake floor in that nested caldera and why the caldera is always liquid water under the icesheet.

          1983, 1998, 2004, 2011 occured all in the same place
          suggesting a hot dyke pathway up there
          The caldera may have trapfloor like caracther like Sierra Negra
          Where pressure in grimsvötns chamber builds and choses same weak point in the caldera ringfaults to erupt


    Magnitude mb 5.0
    Date time 2019-04-05 10:28:04.3 UTC
    Location 12.67 S ; 45.47 E
    Depth 10 km
    Distances 29 km NE of Mamoudzou, Mayotte / pop: 54,900 / local time: 13:28:04.3 2019-04-05
    264 km SE of Moroni, Comoros / pop: 42,900 / local time: 13:28:04.3 2019-04-05
    725 km N of Antananarivo, Madagascar / pop: 1,392,000 / local time: 13:28:04.3 2019-04-05

    There are currently 67 felt reports at the link above (under Testimonies) with auto-translate to English.

  14. Something ive been working on for months now, its hard to find info on kilaueas eruptions before last century…

    This is the best information I could find about what kilaueas east rift flows probably looked like before the modern period of ERZ activity that started in 1955 and is (obviously) very well mapped. Most of it is based on the 2005 USGS big island geological age map, with relevant flows being of the age group p4y, but the area covered by pu’u o’o is based off a much older map that was made some time in the mid 1980s and is not very high resolution and at times seems to contradict the newer data, so it is a bit smoothed and in some cases I had to make up boundaries though this is minimised. The patch of lava in the upper ERZ is largely extrapolated, there are no geological maps of the lava age that predate mauna ulu that I could find anywhere.
    The notable part of this is that the majority of the activity is outside the national park (yellow line), in contrast to the recent episode where most of the activity has been inside it so far.

    The purple shapes are vent structures formed by eruptions of these flows, the upper two are pu’u kamoamoa and pu’u kahualea, which have both been buried by pu’u o’o. Pu’u kahualea apparently was about 300-400 years old, and I couldnt find a date for the age of pu’u kamoamoa but it was very likely the source of the uppermost long flow by its location, and is likely also somewhere about 300 years old. Most of the flows in that area are completely unknown and are entirely buried by pu’u o’o. The next circle is heiheiahulu, a lava shield, which likely erupted for at least a few years based on its size. It formed over the central vent of a fissure that probably erupted a lot of the lava around it. Heiheiahulu is assigned an age of around the year 1750 but it has not been accurately dated so this may change, and it still looks distinct in the vegetation so it is possibly younger. Beyond that is pu’u honuaula, a the cone PGV is built on and probably the last time an eruption comparable to last year happened, a small unnamed cone, and then pu’u kea and the much larger kapoho crater next to each other at the end.

    The total area of the combined flow field is about 370 km2. Assuming the average thickness of the lava is 5 meters, that is a volume of about 1.8 km3 – 2 km3. Heiheiahulu is much thicker, probably an average thickness of about 20 meters overall, and with an area of 30 km2 that is a volume 0.6 km3 on its own, and probably at least 1 km3 factoring in the probably substantial volume that was lost to the ocean and volume of the stage one lavas (area given is the pahoehoe stage and stage one is hard to define), which going by assumption this was similar to eruptions recently, also gives a probably duration of that eruption being about 5 years, maybe slightly longer, very similar to the kupaianaha eruption.
    That brings the volume total up to about 3 km3. pu’u honuaula has an area of effect of 25 km2, and while it doesnt have a big perched channel like last year formed, a lot of its volume sems to have dlowed into the ocean, similar to the first stage of last years eruption, and to 1960. The fissure 22 flows are about 10 meters thick and so it is likely the flows from pu’u honuaula are similar, and so the likely volume on land is 0.25 km3, and total volume maybe about 0.4 km3. The 1790 eruptions, though likely not erupted in that actual year or at the saem time, I had previously calculated at about 0.3 km3 combined. Kapoho crater is only the cone, and rather limited, the eruption volume is roughly 20 million m3, about 1/10th the size of many of the other eruptions in that area, but little is known about this eruption and with all its surroundings buried last year it will probably remain that way. Overall the erupted volume of these flows is about 4 km3, which compares to the roughly 12 km3 erupted by kilauea since 1952. Maybe the caldera was frequently refilling and collapsing during this time period much more than today, or the south flank movement was higher so more magma went into storage. Or maybe I did my maths wrong and it is actually a lot higher…

    Im not going to post it now but I have also made maps like this of every time period from the early 1700s up to today.

    • Based on the detail presented in this paper (and the associated maps) which was just published in May 2017, the best bet for today would probably be to just to move to Hawaii and try to get to know John Lockwood, Frank Trusdell and Don Swanson.

      They probably have a mountain of very detailed, but still unpublished, data.

      I would not be surprised if they are planning to publish similar detailed maps of other parts of the island.

      • Not of the ERZ, their field work has been summit focused for a while now and with the new exposures in the summit it will most likely remain that way. I understand summit explosive events seem more threatening but they are leaving an important part of the system out. When the Leilani eruption started they had to cite a publication of ~1990 to refer to prehistoric ERZ activity as there has been basically no new studies since then. They also left the ERZ out in their model of explosive-effusive cycles

        • Yes, their study and the recent study of the SWRZ is very interesting and I suspect we will get some really important results in the next year.
          One of the problems that I think gets thrown around is that because the eruptions of 1790 happened in a deep caldera that it gets assumed the caldera was new, and im not exempt from that. In reality all we know is that there was a caldera then and it existed. Probably for a large amount of time between the calderas original formation in about 1500 up to the mid 1800s it was very deep, most of the eruptions of that time are below the 600 meter elevation while most eruptions in the current ERZ cycle have been between 600 and 1000 meters elevation and the caldera floor at 1000 meters ASL until last year. Probably not coincidental. I expect that any ERZ activity in the future will be in the lower half now, exactly where the deep inflation is occuring. Obviously this will appear like there is low supply, but in reality it is just erupting somewhere else.

          In this case the so called massive collapse of the late 18th century could have been minor and just a pit crater but it sank past the water table, and the eruption forming it need not be very large, one of the 1790 flows could have done it easily. This existence of a pit crater could also be evidence of there being an active lava lake in the caldera and that long lived eruptions on the rifts may often be accompanied by long duration lava lakes in the caldera, this was very likely the case around 1800 on the southwest rift as well as mauna iki also on that rift and now puu oo on a different rift. Its a stretch but if puu honuaula had any important part in this then it didnt create a caldera that sank below the water table, no explosive deposits around that time. This most likely means there was no ongoing long duration eruption in the caldera back then.

          Of course the other problem is obvious, the east rift isnt well dated, and that is a problem because since 1992 when that study was done a lot more has been learnt about kilauea. At the moment HVO still maintains that the very slow aila’au pahoehoe shield summit eruption was responsible for the formation of the 1500s caldera, despite it predating the event by decades even in the original dates, and last year showing exactly how its really done, and not to mention holuhraun where the exact same thing happened on the same scale in an entirely different volcano…

          • I dont see why Puu Honualua wouldn’t have collapsed the caldera below the water table, there are plenty of explosive eruptions in the Keanakakoi to choose from. Neither Puu Honualua nor the explosions are very accurately dated.

          • Not past the water table, but I really find it hard to believe it wouldnt have done anything. 1840 was a summit collapse, and so was last year, both in the same area, and pu’u honuaula is a bigger eruption than 1840 by a significant degree so it very likely would have made a decent collapse. In any case, though not very well dated both definitely happened well after the initial collapse of the caldera in the 1500s, and so the situation would be broadly comparable to today.

            It is actually the fact there are no explosive deposits that are about that age that tells me it didnt go past the water table. It could have made the caldera deep though, and for whatever reason eruptions afterwards didnt fill it much, but there could have been a lava lake that formed in the deep caldera during the heiheiahulu eruption (maybe at the transition to when it began erupting pahoehoe) and when the ‘1790’ eruption or eruptions happened this both drained out heiheiahulu and the hypothetical summit vent which then filled with water and allowed the keanakakoi tephra eruption to be as violent as it was. That might have been a number of years after the eruptions in leilani and at heiheiahulu, maybe in the start of the 19th century, and multiple similar but less explosive eruptions happened with some probably post-dating 1823. That surge of magma seems to have also driven the SWRZ activity, as well as possible ERZ activity that is poorly documented, possibly some of the flows that are now under pu’u o’o, as well as maybe one of the 1790 LERZ flows, as said many times I doubt these are from the same eruption, they have very different character (south rift is voluminous, north rift seems to be mostly intrusive with limited eruption).

            Something else I just noticed regarding that too, after looking at it more closely, the main line of the southern rift lines up very well with the trend of the fissures heiheiahulu erupted from, so this line is probably better associated with that eruption. The northern rift postdates heiheiahulu, part of that flow is on top of its lava, and the uppermost vent is impounded by the smaller shield. It was probably erupted within a decade afterwards though we wont know any of this without a study.
            The main flow of the southern 1790 rift also is very thick, over 20 meters in places near the coast, so it alone might have a volume of over 0.2 km3, though the rest of the flows of that eruption seem to be much smaller. That sort of volume is probably quite enough to sink the likely already deep caldera below the water table. But that all still depends on dating that doesnt exist yet…

          • Just as a quick comparison, these are 3 pictures I took out of google earth, showing heiheiahulu, 1790 and 1955 lava, all a’a flows to try and get a fair comparison.
            These were all literally within about 4 km of each other so there should be no differences regarding the climate.

          • The 1840 (and 1832?) caldera was wider in the northern half of the caldera and probably smaller than 1924 in the southern end, from paintings of Titian Ramsay Peale. It was very different in shape to the later collapses which have been centered at Halema’uma’u.

            There are tephra layers of Keanakakoi which could coincide in time with Puu Honualua. There are also descompression explosive events like 2018 or 1924 and that we dont really know how big can get in a faster or bigger collapse nor how frequent they are or how many of the Keanakakoi eruptions were actually similar in mechanism, nor if these ocurred in 1823 and 1832, and 1790 too?

          • Im leaning more on the side of caution regarding assigning the ash layers to any age at all right now, not until there is a source that actually gives a date accurate to within a decade. The 1790 date is given based on extrapolation from 1823. This explosive summit eruption could well have happened in 1820, all we know is that it didnt happen after the 1823 expedition, and that there were at least 2 eruptions on the southwest rift after it and before the keiawa lava flow that was erupted shortly before the Ellis expedition. Ellis described fuming in these other areas in the kamakaia hills which is a sign of eruption very recently.

            One way of dating the minimum age of the lava would be to find a tree growing inside a volcanic vent, and take a core sample for tree rings. Heiheiahulu has lots of trees in its crater which is perfect, though as far as I know no one has ever actually gone there at all to test this. Some assumptions might need to be made about how long after the eruption ended that trees could actually grow there, but then you add that to the tree ring count, and that number is probably not high in that area, maybe 30 years. Ohia like wet places and craters are usually wet so the trees inside heiheiahulu are likely of the fast growing jungle variety that turn into massive trees while those outside are more scrubby and just get tall and thin. There are actually almost no trees at all on the summit of the shield, and parts of the lower flow field still have original bare lava which hints at the young age of this eruption. Part of the flow field from heiheiahulu also entered the ocean at the same place as the easternmost point of the kupaianaha flow, which allows a direct comparison, and with the 1991 flows as a standard it would be possible to estimate how long it takes for vegetation to take root and then add that to tree ring data.

            C14 isnt the only way to date the lava, and the method above actually works better in puna than anywhere else because there are so many trees.

          • You are getting close to defining a research project. With a complete plan, it is time to go to Hawai’i and carry it out! A few comments: I am not sure how good tree rings work in tropical climates (they are more a temperate feature). The height of the trees might be useable. The colour of the lava may be a clue (more weathered should not be as dark). In my recollection of Hawai’i, the fern density stands out as being highly variable and perhaps related to the amount of lava weathering.

          • These 3 pictures (heiheiahulu on top, 1955 on bottom) were literally taken within 3 km of each other, near kaimu. The ferns are variable but all of these flows are close to identical environmental conditions, so I wouldn’t take that as much of a consideration with them. Further along some of the heiheiahulu flows are even less forested, looking like the 1790 flow. The lava flows in all 3 cases are still dark, and the fact they all still have visible surfaces at all shows the young age of the flow surface. It is also interesting that there are a’a flows at the ocean from heiheiahulu, maybe it was more variable in eruption than pu’u o’o, with some faster eruptions

            It would help if the growth rate of ohia trees was a constant, but they are very variable so that is why I suggested tree rings, which should still work in hawaii, it still has seasons. The trees seem to grow fast on the actual rift, even the 1955 vents are forested, but near the ocean it is less so, except along roads which makes it a bit inconvenient. It is actually interesting that heiheiahulu is unforested at its summit, maybe the internal heat of the lava within its center is still too high to allow trees to grow there, the shield is something like 60 meters above its base, and the lava surface before the eruption might be similar again, by comparison to recent similar eruptions that have made lava fields over 40 meters thick at their vents, up to 90 meters at pu’u o’o. Lava that thick might take hundreds of years to cool to a temperature that it is low enough to not kill the deep roots (probably about 50 C), the crater itself likely collects water and would cool faster but the lava on the southern slope is thick and would hold heat more.

            I also think that lava cools slower the colder it gets, the 1959 lava cooled to solidification after about 20 years but even today is probably hot enough to be slightly incandescent at night in its center, at a temperature of about 500 C. Thick lava shields probably stay hot inside for similar lengths of time and may take much longer to become ambient temperature. It will probably take the 1959 lava lake about 100 years to cool to below the curie temperature of basalt according to a study in 2017, the curie temperature of basalt is apparently about 540 C. The lava was about twice that hot at the end of 1959 so that means it might take about 200 years to cool that much lava to ambient. Heiheiahulu is therefor using that data probably at the very most about 250 years old and likely closer to 200 years old, which would but it forming at some point between 1770 and 1820, 1795 being the median, which is rather a lot later than the 1750 date that is usually given. The eruption likely started some years before though, at least 5, maybe more than 10 years. The eruptions in leilani are evidently similar in age, generally younger looking even, so likely also took place in the time period around the turn of the 19th century. The main flow that underlies the malama ki forest reserve also shows a large area of light green, which is probably a pahoehoe surface, maybe the overflows of a lava channel like those in the eruption last year.
            It is never actually stated anywhere that there was a large flank eruption associated with the 1790 summit eruption, but actually it is inferred quite strongly that eruptions on the flank were frequent afterwards, which given the general layout of the populated areas is much more likely to be talking about lower puna than kau where the southwest rift is.

            Again this is frustratingly hard to confirm until HVO studies the area better…

          • I did feel that the trees on the top picture were more mature but was not certain enough to make the call! Water must be crucial to the tree growth. And oxygen. Trees can make it themselves but tree roots can’t.

          • This is another of the p4y flows, it is one of the flows that i marked as ‘other’ on that map, most of the others are buried under pu’u o’o but this one is further east so it survived.

            This flow is on the west edge of the heiheiahulu flow field, it is dated as being part of an older flow on the 2005 map but this flow is definitely very youthful. It is very hard to determine where this flow originated though, above the highway it is cultivated and indiscernible, and on the rift basically all flows more than 50 years old look the same and are forested.

          • There was apparently a westerner who witnessed the presumed plume of the 1790 eruption from far away, and in the month of November. I haven’t read the original report though.

            My intepretation after reading the expedition report of Ellis was that the fuming came from the 1823 Keaiwa fissure, he noted that the fumaroles and cracks extended down from the summit to the eruption site. In his description of Kamakaia Hills I only recall that the natives told him that they were craters.

            The trees could be used to date pit craters too.

          • It would be useful to date the pit craters, except the ones that are believed to have formed in the late 18th century are all buried now, mostly by mauna ulu. I suspect that with only one of the large craters actually having a crater now that it might be not too far in the future that more form.

            Trying to keep everything kilauea related in this comment thread, I also finally did some actual calculations of the likely volume of the ‘1790’ LERZ flows.
            Despite having a total length of 19 km, the northern rift is quite limited with the flow extent, the lava didnt reach the sea so it probably was not a long eruption either. The area covered by the lava is about 14 km2, the lava below highway 130 is pretty thin, making the tree molds at lava tree park, it probably is only about 2 meters thick overall, which is a volume of about 30 million m3, quite small compared to a lot of other eruptions in that area, it might have been active for only a few days and most of the lava stayed underground. My interpretation is that the main part of that eruption was brief, maybe only a few days, and the majority of the lava erupted from the vents at the highway with lesser amounts from lava tree park and the lowest vent. The vent next to heiheiahulu seems to have been impounded so it could be a lot thicker but there is no way of confirming this.

            By contrast the southern rift appears to have been a major eruption. Most of the flows are also thin, but the flow originating from the southern edge of leilani estates seems to have been a larger flow, it is big enough to have diverted the initial lava flows of 2018, which were about 10 meters thick at the ocean, and the upper part of the flow seems to have developed a pahoehoe overlay, similar to the upper parts of the channels in 2018. It also has a wide area of ocean entry, and a littoral cone, so a significant amount is likely lost to the ocean. This flow alone is probably at least 0.2 km3, maybe even 0.3 km3, which is larger than 1840 or 1960. The other parts of the eruption, the flow downrift, and the two long flows uprift, seem to have been much less voluminous, the downrift flow might have been mostly a slow pahoehoe flow like the vents in the same area last year were, while the upper vents might have been like the first flow in 1955 to reach the ocean (which was almost exactly in the same place as one of them), and hence might have only erupted for a few days. This southern rift probably overall has a volume of up to about 0.35 km3, some 10 times bigger than the northern rift. The fissures of the southern rift also align very closely with those of heiheiahulu.

          • The relevance of the water table as a factor seems way over hyped and is too often dogmatically cited as fact.

            For the most part, heat will probably push the water out of the immediate area long before it can directly interact with magma.
            Plus, magma doesn’t need external water sources. It already produces more than enough inside the plume. Water lives in magma as dissociated separate hydroxyl and hydrogen ions (OH− + H+ → H2O) that quickly bond back into water as the lava erupts.

          • Lakes of water in a caldera are responsible for explosive activity in otherwise effusive volcanoes like Fernandina (Galapagos) or Karthala (Comores).

            The way you describe it seems that the southern vent produced a perched channel, the lost volume will depend on how long the channel was throwing lava into the ocean, unfortunately there is probably no way to constrain this.

          • Actually, in the case of the eruption on kilauea in 1790, there is a necessity for water, there isnt enough in the magma. A basaltic lava fountain can turn into a plinian eruption but it still wont make a pyroclastic surge that way, to do that requires a true ddtonation that actually launches the flow in that direction. On a volcano like kilauea that really cant happen because the magma is too fluid. Adding a lake to this renders all these factors null. It also ties with hawaiian oral history, at least once since the caldera formed a lake existed at the deepest part. Just to say, but a lake that is 20 meters deep that gets flash evaporsted by a big eruption like 1959 happening underneath it, would make enough steam to escape the caldera and surge over the rim at high velocity, which is what apparently happened in 1790, that eruption happened in a lake that was bigger than that though because it was phreatomagmatic for several days at least, some explosions being powerful to excavate the ground that existed before that eruption, with eventually all the water gone it returned to typical high fountaining again.

            The malama flow is very interesting in a lot of ways, it does appear to have made a perched channel, but there isnt actually a visible channel, and there is also no vent structure, which means there was no fountaining at all, or just dome fountaining in a ponded lava body that overflowed vigorously to the south. This gas poor lava as well as the similar appearence and close association to the fissures heiheiahulu erupted from suggests the southern 1790 rift was a relatively shallow draining event from heiheiahulu, there doesnt appear to be any mentions of summit activity or ash on the upper ERZ that suggests pit crater formation, so while a number of the pits seem young they probably didnt form then. Apart from devils pit all of them have or had hawaiian names so they at least existed before the islands had significant outside contact.

          • Perhaps wetness might work better than a full lake. The Leilani flow wiped out the lake at Kapoho but without massive explosions. The possible reason is that the lava just sank to the bottom, so the water was displaced but remained on top, and water as a much higher heat capacity than lava so it is not actually that easy to make it boil instantly. For that to happen you need to limit the amount of water involved.

            It is different if you confine the water to inside the rock, for instance let it enter a subsurface lava dyke. That would be explosive.

          • There is a massive difference between the 1790 event and lava flowing into green lake last year. The lava there was probably around a bit less than 1100 C, it was 10 km from its vent, and entering the lake at maybe a few tens of m3 at most and likely less as the majority of the flow went elsewhere and the flow was not channelised at that location yet. By contrast, the 1790 event was a very large eruption anyway and it occurred within a lake. The closest comparison to that, as i have saud before, was 1959, which had eruption rates in excess of 2 million m3/hour, far higher than even fissure 8. It was also much hotter, the core of a fountain like that is about 1200 C and probably even hotter, it cools quickly to about 1100 C after flowing away from the vent whuch is why big fountains usually make a’a flows, but that is a moot point when the vent itself is in water. To make enough steam to fill the entire caldera you only need 3 million m3 of water, which is a lake 100 meters wide and 30 meters deep more or less, which i think wouldnt stand a chance against an eruption like 1959 happening underneath it. There is also evidence to suggest some of the 1790-1800s eruptions were a lot bigger than 1959. I also doubt the lake would be room temperature, more likely it would be near boiling, which lowers the amount if energy needed by almost an order of magnitude, not that I think it matter at that point.

            Kilauea also has the one of the highest heat flows of any volcano on earth, even now, it is not likely the water table would be able to get anywhere near magma let alone in volumes to effect the eruption, the water also deposits minerals in the basalt and makes it a lot less porous than originally believed. Several stydies in the past have also determined that the amount of groundwater in kilauea wouldnt be enough to invade a dike, but a lake would, groundwater could make explosive eruptions on the LERZ and it has done so twice in the last 50p years, one is on video, but it is also at sea level more or less and that changes everything, the caldera even at its deepest is still hundreds of meters above sea level and all of its potential groundwater has to come from rain falling on mauna loa and only on a small area.

          • Water from rocks is a non-factor. This is another example of old dogma masquerading as a current fact.

            Ages ago, quite understandably, the chemistry of the magma was not well understood and it was inconcievable that hot magma could be full of water.

            It was assumed that any water in the eruptions must have been externally sourced. This is not remotely true. The water table theory needs to go away.

            We now know, however, that the building blocks of water live inside the hot magma in quantities vastly larger than any surface lake or wet rocks could supply.


          • Yes there are hydronium and hydroxide ions in magma but that doesnt mean they will react, and if that was really an important factor then why does it only explode sometimes and not all the time? Chemistry is an under explored area in volcanology but it is not something that will make a difference in something like this. Most water is also a supercritical fluid dissolved inside the magma not in solid compounds, so it isnt really going to change much anyway.

            Again, it is supported both by observations and also the physical limits of water that the 1790 eruption was caused by a vigorous lava fountain (or at the very least an eruption with a massive flux rate) that went through a body of water in the caldera, basically think something like 1959 except through water, you get a massive interaction as well as direct contact with magma at its initial eruption temperature, which again is probably at least 1200 C for kilauea and could even be higher than that in an eruption that originates as a pulse of magma rising quickly out of the deep source (which is what 1959 was, and likely also 1790). The amount of pressure at the bottom of a lava fountain of any sort is enormous, but consider that in 1959 it is capable of launching a viscous and very heavy fluid 500+ meters in the air, at close to the speed of sound… The 1790 eruption, as well as the two eruptions after it, each made close to 3 times as much tephra as 1959, and apparently in less total time, so these eruptions would have been incredibly violent as far as hawaiian eruptions go, fountains on the order of a km high. Groundwater doesnt have a chance at intruding into magma like that, but a lake that is tens of meters deep would be able to interact with magma. We have in fact seen exactly this only last year –

          • Nope….99.9% of the water in eruptions is from outgassing.

            Most volcanoes erupt from elevations well above any water table and have the exact same chemistry.

            Just my opinion, but my guess is the character of the individual eruptions in Hawaii (explosive vs effusive/fountaining) is, more than any other single factor, probably mostly dictated simply by the state of the caldera summit plumbing at the time of the particular magma infusion.

            Two examples of the identical 0.5km3 infusion:

            Scenario 1: new magma + good summit plumbing + not entirely busted pipes to the ERZ = summit lava pool and/or periodic overflow draining to the ERZ with relatively stable outgassing.

            Scenario 2: new magma = busted summit plumbing = rapid magma cooling and explosive outgassing (including millions of tons of H20) at the summit vent.

          • Actually, the cause of those explosions was the collapse of the foamy upper layers of the lake, after the impact of a large mass of rock falling from a significant height (100 meters or so). And if you are using the overlook crater lava lake as an example of cooling magma then you should probably be aware that its temperature was 1250 C so it was not only hot but twice the temperature of the sort of magma the effect you talk about usually happens in (crystal rich rhyolite or dacite) usually only 600 C). At points it has been considered that the rock above the lake might be waterlogged but that is very unlikely especially in that video which was in the lakes final month of existance after over 10 years.

            Basically the idea is similar to dropping something flat onto soapy water and it spraying the bubbles everywhere except you scale it up to 300 meter wide pool of 1200 C liquid similar to glass and the thing falling is similar dimensions and falling 100 meters obviously the amount of energy involved will be huge. The gas in the magma foam is compressed and tgen explodes back out, blowing everything out of the vent. A lot of that ash will also be pulverised rock from the impacting rock slide. The one thing it is not is a large portion of the lake freezing and shattering in an instant from the dissolved gas coming out of solution in the glass, as said before the overlook lava lake was much hotter than the freezing point of basalt so if anything the rocks falling into it would melt rather than the lake freeze. Now imagine the lava is not static but being ejected out of the vent at nearly the speed of sound, add a small water lake (which again is supported by observations as well as other evidence) and you get a pretty massive explosion. Also despite the heat capacity of water it would probably be able to evaporate most of that lake within minutes, the general temperature of the ground at kilauea is very high and a lake fed out of mostly groundwater would be very hot probably almost boiling, so that removes about 90% of the energy needed to boil the water. Basically most of the lake turns to steam so fast it rushes out as a shockwave which is the surge that is mentioned. No chemistry or wet ground needed, the lake can just come from rain, all you need is a deep pulse of magma to erupt quickly and through the water. An eruption exactly like this but smaller happened at karthala volcano in the Comoros islands near Madagascar in 2007. And a much bigger eruption of this type happened at fernandina in 1968, which was an explosion of megaton proportions, so this is the real deal.

            Just as a note, the deep quakes indicate a sizable pulse of magma is on its way to kilauea, its next eruption might be a big one…

          • There have been bigger deep earthquakes in the Pahala area in the past, even above magnitude 5. And big swarms like the recent one are not unusual.

          • Agree to disagree on the water impact.

            However note that USGS now also severely discounts any meaningful impact from ground water / magma interactions. Magma can produce loads of gas including water vapor all on its own.

            Also, I think that the vast majority of “ash” is not pulverized rock. Nothing whatsoever to do with the busting up of rocks or bits of busted up lava.

            The thing we call volcanic “ash” is mostly just precipitates.

            Entirely different creature than dust or tiny lithic sand.

          • The ability of water to trigger explosive eruptions is most obvious at an ocean entry, a lava flow that is behaving effusively enters in contact with seawater and is fragmented into tephra by the expansion of steam. If vigorous enough large amounts of tephra will accumulate into a litoral cone, the most impressive example would be the 1868 eruption of Mauna Loa which built two litoral cones nearly 100 m high in 4 days. It is resonable to think that if more gassy magma comes directly into contact with groundwater or a lake during the onset of an eruption and particularly with high eruption rates then the interaction will be much more vigorous either blowing up the pre-existing terrain or the magma into ash. And this fits numerous observations, water-driven explosive eruptions do exist, now purely magmatic explosive events also exist even in usually effusive volcanoes like Kilauea, as shown in May 2018, in this case rapid descompression of magma causes amounts of gas escape upwards and remobilize material of the caldera floor into the plume.

            Lithic ash would be the remobilized material, while vitric ash is fresh magma fragemented and solidified rapidly. To my understanding these are the two main products of explosive eruptions that are referred to as ash. Sulphates will also form in the plume.

          • I am proposing a twist on that standard definition of ash….here’s how I got there.

            First….here’s just one example of the chemical analysis of the tephra in the ice cores.

            It’s not all sulfur. In fact, only about 1/2 is sulfur (SiO4). The rest is things like MgO and CaO FeO, etc.

            The most accepted way for sulfur to get from a volcano to the ice core is in the form of sulfuric acid. H2SO4

            This makes no sense…..

            Vast, dense clouds containing both sulfuric acid & some of the most volatile bases around will react…..and they will do so in a highly highly exothermic way.

            The sulfates that are formed are now very stable (and also mostly very white) and can travel merrily along their way in the upper level winds for a while.

            In some images you can practically see it happen. The plume transition from yellowish-brown muck into a lighter “ash”

          • Here’s a better picture of a plume turning from grubby yellow brown to white as it forms sulfate “ash”.


          • HVO doesnt discount all water interaction only that explosive events during caldera collapse are caused by it. 1924 and 2018 were both preceded by lava lakes so it is unlikely there was water anywhere near the vent. This is actually something I said months ago back in the middle of the eruption last year. It is not at all the same thing as saying water interactions never happen in hawaii because they do.

            However the mechanism regarding a lake forming and then being erupted into is completely different, there is no collapsing caldera, there is massive amounts of magma that is extremely hot, and there is a nearly boiling lake. Think surtsey except the ocean is boiling…

            If the caldera was dry in 1790 then that eruption would have been like 1959.

          • I would also like to point out that you used the term ‘volatile’ when refering to basic metal oxides. Those compounds are anything but volatile, CaO has a melting point so high it can glow white to our eyes as a solid – limelight. That is a notable physical difference between acids and bases, most acids are liquids and are volatile, which is actually a term refering to evaporation. Bases in water are all oxides, either as OH- or more rarely as O2- (peroxide) but all of these as solids will just form metal oxides if heated and lose water (Ca(OH)2 turns into CaO and H2O with heat, NaOH melts but doesnt evaporate). Actually there is a small overlap between the liquid phases of NaOH and H2SO4 so that would be a scary reaction…

            In magma most of those also exist as silicates, SiO2 doesnt exist on its own in magma either except in really felsic rocks and usually only in granite. SiO2 and Al2O3 are both technically weak acids (actually they are amphoteric but that doesnt change anything here) so S group metal oxides will react with them. Same thing happens with Fe2O3 which is slightly acidic, and iron exists as FeO as well and can act as a weak base in that form and react with Fe2O3 to make Fe3O4 which is magnetite.
            FeSiO3 mixed with MgSiO3 is olivine, though those compounds have different ratios that can make more minerals too that I forgot the name of.

            Most of these same reactions are how glass is made too, adding CaO to sand, basically making glass is intentionally making very felsic lava, though it flows very easily because of its temperature (1400 C)

          • Fair point….. substitute “volatile” for “reactive” in the description of this hypothesis.

            Highly reactive bases, especially ones like MgO and CaO, would not last long in a dense cloud of sulfuric acid.

            This is an actual sulfur spike layer (Laki) in the GISP2 Greenland ice sheet.

            SiO2 52.26%
            TiO2 2.88%
            Al2O3 13.05%
            FeO 14.64%
            MnO –
            MgO 4.81%
            CaO 10.41%
            Na2O 1.26%
            K2O 0.69%

            This happened.

            It is fact.

            I am just suggesting a hypothesis that, to me, provides a fairly simple mechanism for how all of these ions (not just the sulfur) were transported from the volcano to the ice.

            I am open to any other suggestions for how all of the ions (not just the sulfur) arrive simultaneously and in the correct proportions as the original basalt lava.

    • That is really interesting. I guess it was not well studied because the mature forest along much of the south coast/puna suggested that large flows were not frequent. The erupted volume since 1952 must be above average as it includes a 30+ year continuous eruption. How much above average is to be determined!

      • Well my number is only very average, the flows are quite big and may well be much more than 5 meters thick in places (despite actual images being rare topographic maps of the area exist fro before 1983 and some flows are clearly visible) in historical time eruptions of this size and scale have typically been between 0.03 and 0.1 km3 (1977, 1969, 1965, 1955). Heiheiahulu might also be bigger than shown here, just today I noticed that there appears to be a smaller shield just uprift of it, and in general the area between the upper 1955 vent and puu oo seems to be mostly covered with lava of the age group p4y with a lot of it probably being erupted by the same vent system that later formed the shield. It is also hard to determine how much lava flowed into the ocean and was lost, that was done for last year because it was accurately observed but even for puu oo those values are an unknown. However I have looked at every instance of there being a sustained ocean entry since 1983 and as much as 70% of the time there was something going on, which in the end brings pu’u o’os volume up from 4.3 km3 to about 7 km3, and that agrees very well with the 0.2 km3/year supply from the hotspot (36/5 is 7.2). For heiheiahulu, closer to the ocean, it was likely even more lost, and that is just the shield there is also the lava underneath that from the initial stage that also appears to have been voluminous.
        The main flow of the southern 1790 rift, erupted from a vent just 1 km southeast of fissure 8, also has a wide ocean entry and probably lost a lot of its volume to the sea too. That flow has a littoral cone that shows it had a sustained flow in one channel for some time, longer than a week at least and likely the better part of a month in order for the pahoehoe section of the channel to become stable near the ocean (littoral cones didnt form on the fissure 22 flow that was active for a week, the flow was always a’a at the coast). The 1840 flow also made a littoral cone, there is no data on how long it took to form though but it was likely in the latter half of the eruption.
        Initially I took the lack of a vent structure on that flow to mean it was brief but 1840 was active for over 3 weeks and has no cones so it is likely these eruptions were similar, passive and very fast but with no high fountaining unlike last year, consistent with gas poor fluid magma, such as that draining a lava shield with a lava lake or a big lake at the summit.

        The interesting thing is that both heiheiahulu and the 1790 flows are just as visible in the vegetation as the 1840 flows, both have fewer trees and look a more light green. Heiheiahulu is probably at most about 250 years old, 1770, and could be even as young as 1800, with the 1790 flows being formed some time afterwards but within a few years at most. This also gives a small possibility of both actually postdating the keanakakoi tephra eruption too, which really completely thows out a lot of assumptions of the volcanism of that time period. HVO actually said in their recent publication that since the keanakakoi tephra eruption no major explosions have happened but the first expeditions were told that numerous villages were destroyed, and I dont think all of those were under the 1823 flow and there are a lot of villages in lower puna… There is also the report of an eruption in the middle east rift during the Vancouver expedition in 1794.
        Really I cant confirm any of these ideas until HVO does better study on the east rift though, it actually really surprises me that so many of these cones that are (literally) in peoples backyards and in close proximity to last year are not well dated, or even dated at all… Maybe if HVO sets up a station in pahoa they might take more interest.

        If only I could stop finding so much new stuff on this subject, then I’d ask to write an article on it and settle it down. Kilauea will probably start erupting again before that can happen at this rate though…

        • Dating lava flows is not easy. You can’t date it by 14C which it doesn’t contain!. You have to dig through to find some carbon buried by the lava (carbonized vegetation) but that only gives a maximum age of the flow. There are some other methods but it is a problem.

          • Its more that the data of the studies on summit eruptions on their own give the idea that kilaueas supply frequently and wildly changes between eruptions overflowing the caldera and explosive eruptions in a deep caldera, while ignoring the fact we would actually still be in such a period of low supply explosive activity by the observations they base the study on, the only magma to leave kilaueas caldera after erupting from vents inside it in historical time has been tephra, nearly all in 1959. The ground around kilaueas summit is still to this day the keanakakoi tephra that was erupted back then, and in future without observation the 1959 tephra would be lumped in with it just the same.

            It is very important to look at the east rift, especially the lower part, because for the past 500 years that is where the majority of eruptive activity has actually been. Actually the only time kilauea has been quite in the sense that HVO is talking about is when mauna loa takes the hotspot for short periods, but that actually falls quite out of line with any link to explosive eruptions. It also has only done that once in the time the current caldera has existed.

            Better dating of flows on the ERZ would give a far better picture of the processes that happen within kilauea, it is also evident that the flows are datable it just seems like there is not a collective interest to study this area as much as the summit.

          • True though that the summit is easier to study, dry unvegetated area lots of outcrops in the caldera walls and Koae fault scarps.

          • Interested how lava flows rocks meteorites etc are dated. For older stuff then recent lava could not decay of other isotopes with longer half life be used?

          • There are other radiometric dating techniques, K–Ar dating is very frequently used with lava flows. Problem with Kilauea is that almost the totality of its surface is covered in lava less than 1000 years old leaving C14 and its short half life the only usefull one. Other methods can be used, in 1980 Holcomb used secular variation of the geomagnetic field to try and date some flows if the data is updated to recent models it could actually give a better idea on some of the flows ages, vegetation was also used by Holcomb though fires, anthropogenic deforestation and other factors may interfere with the results though tree rings might provide an accurate minimum age. Hawaiian oral history may also provide some rough constraints on the age of the flows (Holcomb also tried this one xd).

          • Hawaiian oral history actually supports the recent dates very well, and also indicate a lot of that period was poorly documented.

            “the native Hawaiians questioned by William Ellis in
            1823 claimed that no “great explosion” of Kīlauea had occurred
            since (1790), although these long-term residents also added
            that many coastal localities had been “overflowed” with lava
            since then (Ellis, 1825). In 1794, British Navy Captain George
            Vancouver, sailing along the coast of the Puna District, observed
            columns of “smoke” issuing from the central East Rift Zone,
            which accompanying Hawaiians related to an ongoing eruption
            motivation for them to make offerings to the goddess Pele
            (Vancouver, 1798). However vague and even contradictory these
            early accounts appear to be, they at least suggest that substantial,
            undocumented extrusions of lava from the rift zones took place in
            the decades immediately following 1790″.

            – appendix of

            This ‘ongoing’ central ERZ eruption is very likely to be referring to heiheiahulu, especially if offerings to pele were being made it suggests this was neither a new or brief event, and one that had been going on for some time already (and realistically it is very unlikely that expedition would randomly coincide with a 1 day eruption, as opposed to a 5+ year eruption). My earlier calculations regarding the age of heiheiahulu did in fact come to the conclusion that it was most likely erupted in the last decade of the 18th century, and the flows of that eruptive series, as well as the two rifts in leilani estates that are assigned an age of 1790, could account for the occurrence of ‘many coastal localities being overflowed’. It seems to be inferred in the report that eruptions on the southwest rift were the source of these accounts but given that only one of those flows even reached the ocean and it happened the same year Ellis was there I am very sceptical.

          • I have checked the report of Vancouver, he describes columns of smoke rising roughly halfway between Mauna Loa summit and Cape Kumukahi, most likely a Kilauea summit eruption.

            Archibal Menzies was in the party of Vancouver 1894 and he was the first to ascend Mauna Loa, I do remember reading that Menzies reported activity in the caldera of Kilauea during his ascent. Vancouver doesn’t mention any of this though, so I will see if I can find a publication from Menzies that gives more information

          • I have checked the account of Menzies ascent to Mauna Loa in 1794, apparently they approached from the southeast flank of Mauna Loa and had the “volcano” (Kilauea) to the right as they walked, smoke and ash is described to be rising from it, blowing towards them and being irritating to the eyes which most likely corresponds to the columns of smoke Vancouver describes sailing offshore of Puna and that King Kamehameha told him were related to eruptions. It is surprising that Menzies doesn’t seem to have asked the natives anything about Kilauea not even be aware of its name, no glow was reported during the expedition either.

          • Re-post…..

            You are probably not going to find anything more detailed and current on this time period than these recent publications by Mr. Don Swanson & friends.



            Personally, I still think that the “Mystery Unit” above the “Footprints Ash” may also be post-1790 and pre-1823.

          • Yes, it is very interesting and has the most updated data of the late Keanakakoi ash. The mystery unit is shown as underlying the deadly surge in stratigraphic columns, it seems then to have been the main phase in the 1790 eruption, the large blocks seem to correspond to the description of an initial eruption that sent a shower of rocks which hit some people, the deadly surge must be the one identified to have killed a third of Keoua’s armya and ocurring two days later according to Hitchcock (Hawaii and its volcanoes, 1909). There is no ash layer represented between the 1790 eruption and the eastern and golden pumice deposits, these were produced by high fountaining which is clearly not what was going on in 1794. So the activity seen by Vancouver and Menzies is most reasonably the thin lithic ash overlying the eastern pumice, maybe some minor phreatomagmatic activity ongoing inside the caldera? This may actually be bracketing the high fountains between 1790 and 1794.

          • The puna coast doesnt go past the caldera though, it starts at apua point and goes east around kilauea, I think it stops near hilo on original maps but it might be different today. It is a stretch but if there is some indication as to which way people were looking when making their observations it would help. From the puna coast if they were looking at the caldera maina loa would be behind it, so the caldera would not be half way between mauna loa and cape kumukahi but rather the actual middle of kilaueas east rift which would be a better fit with some of the other things the early explorers were told. Observations of smoke dont necessarily refer to ash, lava fountains also make large clouds of ‘smoke’, and while this could be talking about the caldera, the presence of very young looking a’a flows on the ERZ originating near and probably in some cases from heiheiahulu would suggest fountaining. Heiheiahulu is not exactly that similar to kupaianaha like I originally compared, it looks steeper and in some cases pretty cone shaped, the latter part of mauna ulu, a month before it ended, it had a return to high fountaining, some fountains were over 100 meters and sent long a’a flows down the flank of kilauea to the coastal plain. mauna ulu is 12 km from the sea, but heiheiahulu is only 7, so a similar phase there would probably reach the ocean. Heiheiahulu actually looks very similar to mauna ulu.

          • Problem is that both Vancouver and Menzies seem very schematic in their descriptions but the absence of any reports of glowing certainly discard the fountains hypothesis and maybe lava lakes or even lava flows too. Whatever the activity was some things are certain, the 1790 eruption ocurred in 1790 this meaning that the big explosive eruption and presumed collapse event had been 4 years ago. The source of the activity was most likely Kilauea Caldera, the ascent of Menzies took him near the SWRZ of Kilauea, there trade winds would be blowing from the summit which is from where “smoke, dust and ashes arising from it proved very bothersome to our eyes”, that is basically all detail he gives. Vancouver sailed from Puna to Kau so he got to see it from diferent angles, Menzies describing the source of activity to be the summit about a month later means he was probably looking at the same eruption.

          • There is a possibility their observations are of an ash cloud resulting from collapse of the edge of a newly formed crater. There have been a lot of those at pu’u o’o in the last year, and will probably be more over the next year too. Going with my hypothesis, this could be interpreted to be collapse dust from within heiheiahulu, or from a pit crater on the upper rift as it is likely some of those were very recent at the time of their expedition. Some of these go very high, multiple km, and look like vulcanian eruptions, but are just caused by collapses of the edges and air compression.

            Or maybe I have been asking the wrong question. I cant find the report from 1794, but it would be worth trying to find whether there is any mention of recent looking lava on the flanks of kilauea, if they sailed along the puna coast then that is basically impossible to miss even if they arent trying to look at it, In what I have read, that doesnt seem to be the case, and based on the lack of vegetation on the flows even today, and how long it takes to grow stuff on the lava, it is quite certain that if heiheiahulu was erupted even in 1750 (which is probably on the older side) it would still look quite barren and very obvious 44 years later, after all the 1955 and 1960 flows look barren today (nearly all growth on them is artificially planted) and they are more than 44 years old. Based on the amount of growth on heiheiahulu it might have even been barren within the last century on the shield itself, and same with the flows in leilani, which are similarly mostly covered with ferns and have few trees that are of significant size (except, again, near the road which makes things annoying on google earth). If there is no mention of this, that could indicate the eruptions that created heiheiahulu, and the ‘1790’ eruptions in leilani, both postdate 1794, and also postdate the eruption that made most of the ash. It also again fits very well with the descriptions of ‘many coastal areas being inundated by lava’ told to Ellis in 1823. If all of that really happened in the 29 years between 1794 and 1823 then that is a lot of activity, maybe even comparable to pu’u o’o, though less continuous.

          • Thats the thing, Vancouver must have passed right in front of Heiheiahulu and had it been active he would have observed an ocean entry or some other sign of eruption. The date of 1750 comes from the natives who told Ellis that Kaimu had been overflowed by lava in the days of King Alapai who is estimated (I don’t know how) to have died in about 1754. The flows of 1790 were also dated from hawaiians in roughly 1788 but I don’t know to who.

            I think I have metioned every detail they give Kilauea-related, as I mentioned their descriptions were brief. But this is the report of Vancouver in respect to the columns of smoke:
            The account of Menzies is included in the compilation done by Hitchcock on Kilauea and Mauna Loa (1909):

          • It is worth noting that it wasnt just heiheiahulu that produced lava flows that invaded the kaimu area, that whole area was pretty much completely overrun. In the past 60 years there were many eruptions in the middle east rift before pu’u o’o formed, so maybe that was also the case then, with many eruptions in the area where heiheiahulu is that predate the actual formation of the shield. That young looking flow that I linked a picture to above is one of them. Locations in hawaii seem to be very specific, villages in hawaii also seem to have been generally rebuilt at their original location, so the location of kaimu today is likely also where it was in earlier times, which is actually more east of heiheiahulu but a lot closer to a few of those other unknown flows.

            While it does seem to be clear that there wasnt an eruption going on down in lower puna in 1794 it is pretty likely that a flow as big as heiheiahulu would have been noted if they sailed past it, active or not, if anything just from the size. But there is nothing, which to me indicates that nothing was there at the time and the eruption likely would have occurred later. The most uprift section of the northern 1790 rift also postdates heiheiahulu, whether it is actually part of the same rift as the vents that made lava tree park is not clear at all bit if they are it does indicate that eruption is younger than the eruption that made heiheiahulu. I have already said my reasoning for thinking the southern rift is likely the final event of the heiheiahulu eruption, so I wont go into that.

          • What do you guys think about 1782/1783 as a possible approximate date for something very very large @ Kilauea?

          • Not for what you are probably going to link it to… Those dates are very particular.

        • HVO belong in Kilaueas Caldera!
          My favorite volcano is its home and nowhere else
          Its ridicilous moving it away from the Island .. US state sucks in that Idea.
          The HVO was founded and born 1912 on the Calderas rim and there it belongs.
          Thomas A jaggar woud crawl in the grave IF he knew what is happening to his beloved observatory…

          # move HVO back to the caldera areras just little further away from the caldera faults maybe

          • Well the place he decided to build it was also a sacred location that was not ever intended to be built on, and there were issues with that, and then it is in a national park. I think the agreement was that it would stay but not be allowed to be rebuilt if destroyed, and while it isnt destroyed per se it is damaged enough to be dangerous so they cant occupy it anymore. I suspect volcano house is similar though it is likely to be much better off being relatively far from likely eruptions.

          • Other than where it was, the only other place it would make sense to build it would be where volcano house is. HVNP does want HVO to have a presence in the park, so this might happen, but it is not likely to be where the HVO building is, that will probably be a lookout and otherwise tourist location but not a research building.

        • I can’t recall anything remarkable in any sense back then, it was most likely a period of small summit or ERZ eruptions. The action started several years later

          • This was an answer to BadWolf on anything large ocurring in 1782-1783

  15. FYI using an Occam’s razor approach, I think I solved the problem of UT/LS transport of sulfur from basaltic eruptions.

    If I am right, and I am 98.6921% sure that I am, then it provides a mechanism to show that the vast majority of the sulfur in the dense volcanic plumes from medium/large basaltic eruptions is likely injected at altitudes that are higher than most current models predict and that the vast majority of the sulfur from these basaltic eruptions is NOT oxidized or degraded in the UT/LS atmosphere which makes it possible to transport it much farther.

    Not ready to publish the idea just yet. It is going into the paper that I am working on.

    • Here’s a hint to stir the pot:

      It turns out, the following statement is probably mostly FALSE for the vast majority of the ejected “ash” volume for all types of volcanoes but especially for medium/large columns from basaltic eruptions.


      Ash is created when solid rock shatters and magma separates into minute particles during explosive volcanic activity. Ash particles are incorporated into eruption columns as they are ejected from the vent at high velocity.

      • For Toba, the ash wasn’t lifted by explosions, but by the heat from lava flows. Such co-ignimbrite doesn’t get as high though.

        • Toba did not erupt any lava flows
          Souch gas rich viscous dacitic magmas produces pyroclastic flows and Ashfall
          The pyroclastic flow deposits where very massive and thick and cooled slowly
          They welded togther and formed Columnar joints

          • No, there was lots of lava. It was mainly an effusive eruption, The lava even reached both oceans. It is a collapse caldera, not an explosion (although it will have started with impressive booms). The eruptions were along the ring fault.

        • There where no lava flows
          Are you soure?

          You misstakes the collumnar joints in the rehomorphic dacite pyroclastic ingmigbrites as lava flows. Souch gas rich evolved viscous dacite blown up into ashclouds

          There may been dome building or blocky flow action after the eruption when the gas poor stuff came out

          There are NO Flood Dacites
          There are Sillic lava flow LIPS but they dont flow 100 s of km in a few days

        • Thats a very intresting VC article
          But remeber Tobas extensive ashfall in India and Asia
          This was from a thousands of km wide
          Ash plinian ash cloud shield that darkend the skies for the whole region

          • Ash only to the west and northwest. East Indonesia escaped. Most did not go very far but fell in the ocean. There was some early explosive ash but most was co-ignimbrite, raised to 8-9 km by the heat. A bit reached Africa which is quite far.

        • There are No giant dacite flows preseved there

          Are you really soure your opinions are correct?

        • Turns out that most of the “ash” is created within the plume from the components in the plume. Lots of SO2 making sulfuric acid….but also lots of MgO and CaO….very basic….neutralization reactions within the plume create the “ash”

          Virtually none of the Laki or Eldga sulfur was transported to the ice cores in Greenland by sulfuric acid.

  16. Is there any chance for the 800km3 deep resovair under Grimsvötn – Thordarhyrna to make its way towards the surface? What if a very major rifting event happens in the Grimsvötn system in Glacier

    • No chance. Only the hottest parts have sufficient buoyancy to move closer to the surface. If rifting occurs, most magma will move sideways. But the deep magma is far below the locking depth so it doesn’t notice rifting. I am not sure that your wish for enormous eruptions is very helpful. VC tries to avoid sensationalism. Except of course in my posts.

      • Yup most of that magma moves sideways
        And fills the gaps
        In general only 10% of the magma thats on the move in huge Icelandic fissure eruptions erupts
        Holuhraun coud been 10km3 that was on the move but only 1,6km3 was erupted
        The rest stayed in the long and very deep dyke and inside Kistufell and Bardarbunga

          • That volume would then make kilaueas eruption last year the most vigorous effusive eruption since 1783, as in the highest average effusion rate. Holuhraun was 6 months, kilauea last year was only half that, but 80% of the volume. Holuhraun was much stronger at the start but only for maybe 2 weeks before slowing down, fissure 8 was consistently high, if anything it was higher in mid July than when it started at the end of May, the lava flow that diverted around kapoho crater in early July flowed to the ocean 5 km away in a few hours, a literal lava flood, and that was when the vent had been doing that rate for over a month already.

            That seems to be a rather curious characteristic of kilauea, most of its eruptions start small and only peak long after they start often closer to the end of the eruption if anything. Most effusive eruptions including holuhraun as well as those from mauna loa are vigorous for a few days then decline in a linear manner. It would be interesting to know why.

          • I wish I coud understand these complicated maths … my Math Dyslexia is so bad. I cannot even read a normal clock and Im 24.
            Not possible to be happy

            But this blog is one of the best things ever. VC is the best volcano page on the internet so far.
            VC is great 🙂

            Albert your musings on Toba we can debate about 🙂
            Im pretty soure there was no lava flows in Toba 😉

          • Different papers state it as lava flows or tuff flows. Older papers state it as lava and the more recent ones as tuff. Both come out of the ground as lava, but in the high fountains tuff partly solidifies before hitting the ground. At these volumes, it will still flow/ooze as lava. The difference is therefore somewhat academic.

  17. Scotland and England will become completely subtropical in 170 years if the Co2 keeps climbing like this

    reports that the Holuhraun eruption released one of the most SO2 (sulfur dioxide) of any effusive (non-explosive) eruption in the world on an annual basis since 1978, the advent of satellite monitoring of volcanic eruption clouds. The Holuhraun eruption released 16 times more SO2 (sulfur dioxide, 9.6 Mt) and almost twice as much CO2 (carbon dioxide, 5.1 Mt) as one year´s worth of anthropogenic emissions within Iceland (0.06 Mt SO2 and 3 Mt CO2 in 2015). The gases from the eruption were quite poor in the halogens HCl (hydrogen chloride, 0.1 Mt) and HF (hydrogen flouride, 0.06 Mt) which reduced the environmental impact the eruption could have had had the gases been richer in these pollutants. The collection of this eruption data has a large impact on the assessed amount of emissions from Iceland of the pollutants SO2 and CO2, as well as the other assessed gases. These pollutants are assessed by nation because of their significant impact on human health and the environment.
    A team of scientists led by Dr Melissa Anne Pfeffer at the Icelandic Meteorological Office, with partners from 16 international institutions, collated all ground-based measurements of the 2014-2015 Holuhraun eruption cloud. The study recommends that a diversity of methods used to measure eruption clouds continue to be improved upon, and that additional methods be developed to improve monitoring of eruption clouds in Iceland and in other places where winter-time, with very little sunlight, and remote and dusty conditions, have a huge impact on the collection of ground-based data.
    Gases released during the eruption were released from magma that erupted; from non-erupted magma; and from the Holuhraun lava field during and after emplacement. The emissions of gas from the lava field continued for 3 months following the end of the eruption. The gas emission rates reflected the deep magmatic system and also surface processes.
    The concentrations of gases on the surface, which were affecting people downwind from the eruption, were not always reflecting the current conditions at the eruption site. Gases were observed to accumulate in valleys, particularly overnight, and particularly when winds were weak. It sometimes took hours of stronger winds to flush the older gases out of a region.

  19. Looking at Internet: I can mostly only find the explosive theory on Toba
    Makes sense with gas rich viscous dacite magmas.
    They blow up ferociouly when gas rich as the gases cannot escape the sticky magma.
    The caldera ring fault eruption ash columns base in Toba glowed ferciously red – orange in the night

    With billions of volcanic ligthing flashes every hour
    Toba must have looked like one Big hell
    An event almost 4000 times larger than Calbuco 2015
    I can just imagine how a long exposure photograph of Toba volcanic ligthing must have looked like…
    Nuts of insanity in volcanic lighthing

    • The internet may not be the best source – it tends to go ballistic. Try VC instead.. Volcano wold very carefully calls it an ‘eruption’ and avoids ‘explosion’. Tuff indicates fountaining, i.e. effusive. The eruption came from the ring fault, not the crater. We have seen two big craters form/deepen this decade, both by collapse rather than explosion.

      And yes, it must have been an incredible sight but with 1km3 of lava every ten minutes (for 9 days!), the only safe place to watch would from the distant future.

      • Most large caldera eruptions ( Yellowstone and La Garita too ) are erupted through ring faults
        Yellowstone was explosive

        Are you really soure that a very gas rich viscous dacite can form a 100 km long lava flow to the coast in a few days?
        These magmas are extremely viscous
        They dont flood

        • There is no evidence that Toba started with a plinian eruption. Instead, it appears to have been fountaining right from the start. The fountains gave pyroclastics, and the pyroclastics gave co-ignimbrite. Oppenheimer has argued that such secondary heating could have raised ash to 30 km, so it is very significant but not as extreme as an explosive event this size would be. Oppenheimer also points out that the tuff is unwelded, so came down fairly cool (500-600 C). That is probably too cool to behave as lava, although as it covered the entire island anyway, where it also moved is not very relevant! The point, I think, is that the tuff was made from magma, not pulverized rock. This wasn’t a big mountain that blew up.

          Given the length of the ring fault, it behaved very much like a fissure eruption. Of a rather extreme size.

          Models that assume an explosive start pose a shallow magma chamber 3-4 km deep. It is hard to see how a chamber of this volume and depth can be stable. The alternative is a 10-km deep chamber.

          You mention how evolved the magma was. It had been in storage for a very long time. but it was not uniform in composition. Parts were more evolved than other, presumable because of intermittent heating inside the reservoir.

          Toba was one eruption. Other supereruptions may have been very different. They are fundamentally difference from small (VEI7-) eruptions, and I wouldn’t extrapolate from those. The eruption mechanism is different.

          Openheimer wrote

          “The eruption is thought to have ensued as the roof of a large magma chamber, located up to 10 km deep in the crust, began to founder. […] This opened up ring fractures, which fed massive outpourings of pyroclastic flows. If this interpretation is correct, it implies that the caldera formed in a piecemeal fashion during the eruption, and not catastrophically, afterwards. Supporting evidence is provided by the generally symmetrical distribution of YTT around the present day lake, and the absence of plinian tephra fall deposits from an eruption above a central vent. This latter feature of the YTT is unlike most studied examples of large caldera-forming eruptions, which begin with a plinian phase. There is some room for doubt with this conclusion, however. Although no pumice fallout has been found preserved beneath the base of the massive outflow sheets of ignimbrite, which cover >20,000 km2 of northern Sumatra, it is conceivable that the great erosive power of the pyroclastic flows could have scoured away any deposits from an initial plinian phase.”

          • Aaa in this model seee
            NO lava flows 🙂

            Instead massive oversized pyroclastic flows flowed out

            stiff Gas rich dacite magmas does this

          • Toba was ryholite
            Even more viscous when its cold
            And explosive when gas rich

            Dacites and Ryhodacites can be just as viscous as Ryholites when they are cold

          • Fountaining involves high pressure and gas can increase this. But it isn’t a big explosive event, although certainly very high energy, but a scaled up version of early Leilani. At least that is what I envisage. Maybe we should define things a bit. The eruption was mostly rhyolite and some dacite. There was proper rhyolite lava involved but it stayed inside the caldera. The fountains came down as pyroclastics, indeed not as lava. I expect this was nearly continuous. The fountain height may have been 5 km or more. The pyroclastic flows themselves lift ash up, and this reached 30 km. The magma became less silicic as the eruption progressed. Agreed? I am reading some newer stuff than the Oppenheimer article and clearly there still is controversy. There seems agreement that there was little sulfur in the eruption.

          • This is an intresting diskussion
            Are you really soure Albert?
            Huge Leilani style lava fountains in Toba??!
            No way…

            800 C gas rich ryholites do not behave like that.
            What Clive and most other says
            Gas Rich Ryholites blows up into pyroclastic clouds.

            Ryholites are so viscous that when they are very gas rich like that: the gases cannot escape from the sticky magma.
            It all blows up into pyroclastic flows and gets deposited as Ingmigbrites and tuff
            Thats what Oppenheimer says

          • It is how fountains work: The stuff shoots out, goes up and comes down. That is leilani. Make hem bigger, and air gets trapped within the rising stuff, gets very hot, and rises. This hot air carries the ash/particles with it and this becomes a plinian column if strong enough. However, if the flow rate is higher still, as in Toba, the amount of ash is too heavy for the air and the column collapses. This causes the pyroclastic flows. (This is different from ‘normal’ pyroclastic flows which form when a plinian column loses support because the eruption grows weaker). The pyroclastic flows entrain more air, heat it, and now a fraction of the flow gets lifted up again. This is over a much wider area (can be ore than 100 k) and at much lower density. This keeps rising and can reach 30 km (one recent paper says 70 km but I take that with a grain of salt. The atmosphere there is too thin to support it anyway). This is the co-ignimbrite. This is my picture of Toba.

          • That’s not exactly how the fountain plume works.

            The plume initially is loaded with H2O, SO2, MgO, FeO, CAO, and some other stuff.

            The SO2 very quickly forms H2SO4 in the early hottest part of the plume.

            Within a few km of the initial plume, the next level reactions inside the plume are what created the ash and the heat for the additional lift.

            Acid / Base reactions are very very very exothermoc…. especially MgO and CaO

          • That’s what I worked out.

            The plume actually creates the ash (precipitates from the chemical reactions). Ash has very little (or virtually nothing) to do with combustion.

            The plume can lift actual dust and other heavy debris (lapilli and bigher) for short distances but that stuff falls out pretty quickly.

            The stuff labeled as ash is mostly just fine particles
            of gypsum and other sulfates.

            Very stable so they can easily be carried long distances

            Most precipitates are very white and would be highly reflective in upper atmosphere.

            I think that the papers talking about sulfuric acid directly causing the cooling are mostly wrong….especially for basaltic / fountaining type eruptions.

          • Toba is over rated.

            It almost has to be the result of a series of eruptions pretty much just as the article describes.

            Another thing that would work against Toba as a single event vs a series of caldera-area events over an extended period is the proximity to one of the most active and destructive faults in the world.

            It’s hard to imagine a scenario where a single magma bubble, big enough for the speculated size of the theorized single large VEI 8 event, would ever have been able to accumulate and then detonate all at once. It would have taken thousands of years to create a single large magma pool of that size. The Sumatran Fault would have made that scenario virtually impossible.

            On the other hand it is pretty easy to imagine a Mag 9+ earthquake along the Sumartan Fault cracking the top of a magma chamber 10km deep and starting the process.

        • La Garita appears have erupted a strange lava-like tuff made of dacite just before the Fish Canyon Tuff was erupted. This is known as the Pagosa Peak Dacite, and appears have erupted as low-fountaining Hawaiian/Strombolian fountains on a massive scale, about 300 cu. km – something like Fissure Eight, but on a far, far larger scale with higher viscosity lava with bigger blobs? It’s possible that this was a trigger for the later Fish Canyon Tuff eruption, but how far apart these eruptions occurred is unknown.

          I wonder if something like this might also have occurred with the Toba’s ~75 kya eruption?

          • That’s the first explanation I have ever read that actually makes sense.

            A single massive 2,800 km3 lava dome forming undisturbed for centuries and less than 10 km from one of the most dangerous and active faults in the world makes no sense.

          • If gas rich ryholites usualy blows up into ashclouds

            If gas poor they forms lava domes or blocky obsidian flows depending on viscosity

  20. Toba was indeed scary stuff
    The Toba eruption likley began with one or a few intense ultraplinian collumns in a ringfault
    That shoot ash to 60 km elevation
    That likley lasted hours and deposited alot of ash in Asia. 1000km3 from that

    Later the caldera began to founder and huge vents opened up in the ringfaults
    This decompressed the remaining 2800 km3 of gas rich ryholite materials that extremely violently erupted in a boilover st helens close ground blast like sourge. The pyroclastic sourge / flow was a wall 9 km high and moving at 700 km an hour.
    Covering 20 000 km2 in a few hours

    The Toba pyroclastic flows from 60 km distance woud look like angry grey cauliflower wall strecthing from Horizon to Horizon comming towards you at 100 s of km an hour.
    Whole rainforests flattened and buried under 100 s of meters of materials

    • 2800 km3 in less than a day is very probably impossible, that would be cubic km a second eruption rates… The only things that eject that much material in those timeframes are massive impact events and those launch things at escape velocity… Even the eruption rates of Alberts model are really pushing it (1 km3 in 10 minutes) and the eruption has to last a week and a half still at that rate to get to the volume we know was erupted.

      • Some of the 6000 to 10 000 km3 Grande Ronde flows was emplaced in a week
        Competely massive sheet and homogenous compositions across their 600 km lengths ( some say roza and other basalt was emplaced in just a few days
        Other says a few weeks thats seem more likley

        It makes sense for Explosive sillicous VEI 8 s to go just as fast or faster
        But a few weeks maybe more realistic

        But as you says … alberts Toba model is really pushing it.

        • Others say this:

          “Roza lavas feature a wide range of pahoehoe surface structures, such as lava rise plateaus, tumuli, and surface breakouts, and we illustrate that the lava morphology is inconsistent with previous proposals of rapid emplacement for these lavas. An integral component of the Roza flow field is the sheet lobe with internal structures identical to those of inflated pahoehoe sheet lobes from Hawaii and Iceland”

          “The results indicate that the emplacement of individual Roza lobes lasted for months to years and that the lava flow field was constructed over a period of at least 14 years. We propose that the Roza flows achieved great areal dimensions and thicknesses by inflation and endogenous growth.”

          If this interpretation is correct the flows of the Columbia River Basalt were emplaced in a similar way to the “shield eruption” of Undara in Australia 190 ky ago, by erupting huge volumes slowly (sort of Pu’u’o’o in a much bigger scale).

        • Jesper think about it, if that lava flow was able to flow 700 km in a day it would be advancing at 30 km/hour, which is as fast as the lava in hawaii last year was usually flowing in the channel next to fissure 8. That is also over what was likely very flat ground 99% of the time, which is not favorable for the formation of really long massive flows, the flow would just spread out and pond. The eruption rate of a flow like that would also not fit in a dike the size of the ones that are exposed inside flood basalts…
          Pahoehoe flows start off small, so they will tend to flow along the direction of the slope, and will end up flowing more in one direction. Later breakouts expand the flow but the main tube will stay and extend the flow front.
          It is also important to realise the roza basalts were not one single flow, there are 4 flows that total 1300 km3, but none of them individually was that big, and they were far enough in time to be distinct in the geological record (flows werent erupted within at least a few months of each other and likely years).

          • The Discovery of tubes and tumuli and pahoehoe features then is proof that many of the CRFB flows where slow and long lived

            But some of these flows are totaly massive and uniform sheets for 100 s of km suggesting a rather more massive emplacement

          • Some of the Siberian Traps and CAMP flows where anything than slow pahoehohoe bleeds.

            These thick collumnar joint flows where enromous basalt flows 40 meters thick moving as huge glowing Aa flows at near walking speed covering large landscapes feed by large fissures upslope

            Faroe Islands and Deccan, Siberian Traps,Island of Skye haves good perserved ones
            Enomorus Aa sheets.

            They can also be ancient rootless lava ponds like Kilauea Iki near the vent but many 100 s of times larger but similar thickness.

          • What do you mean with a columnar flow? If you are referring to the hexagonal column, these form during later cooling and don’t say much about how fast a flow was emplaced, and nothing about how fast it moved. Either model, of slowly thickening flows or full frontal assault is possible.

          • In basaltic flows thick long columnar joints are signs of a massive thick flow ( despite they are very fluid ) souch flows cools very slowly
            Or rootless lava lakes ( Kilauea Iki will form columnar joints )
            Columnar joints almost always says that the flow was massive and thick

          • pahoehoes and Aa s form cooling joints
            But never hexagonal columnar joints like the really deep flood basalt flows and ponded lava lakes do.

            But the really thick 300 meters Kapoho coastal flows will form columnar hexagon joints and cool so slowly that crystals will form quite large
            The interior of the Kapoho coastal flows will form diabase instead of fine granied basalt. kapoho coastal flows will cool extremely slowly

          • The lava flow at kapoho isnt 300 meters thick solid rock, it is 50 meters on top of what is mostly rubbly material

            It will cool slowly but still be basalt.

          • Yes that makes sense most of it is rubble
            Adims may Edit my commentary

          • Yes, the columns show very slow cooling and that requires very thick layers, but it doesn’t say how those layers formed. It can be 100 meter in on go, or a meter per day, either by adding on top or below. There are some in Iceland.

      • ‘My’ eruption rates are just the total volume divided by the length of the eruption (9-14 days). Quoting Oppenheimer: “The granulometry of feldspar crystals in Indian Ocean YTT tephra horizons was investigated by Ledbetter and Sparks (1979) to estimate the eruption duration. By computing settling velocities for the crystals in water, and from an estimate of the contemporary water depth, they concluded that the eruption lasted between 9 and 14 days. ”

        If the eruption lasted one day, the rate becomes 1km3/min. not per sec. Typographical error.

        • Courtesy note for anyone wishing to contest Albert’s comment. Dr Clive Oppenheimer is one of the preeminent experts on Volcanoes, and Albert is well versed at nudging out details in scientific papers to vet them for inferred or implied use of bullshit and inaccurate methods. If Albert references a paper on the topic, you can bet your arse it is a solid bit of science he has found.

          While I’m good at ruminating, I don’t have the background and training that Albert and Carl have. If I get out of line, they are the ones that will smack me around and point me at my errors.

    • As mentioned above, there is no evidence for a plinian eruption at the start. This comes from the tephra. Your ultraplinian column is imaginary. Stay within the evidence.

      • Toba was the stuff of dreams or nightmares
        As soon as the caldera faults weakened the ryholite magma decompressed in a huge pyroclastic flow just exploded out
        The angry greay cauliflower wall is 70 km across and many many km high in places
        burning and crushing everything in its path 100 s of km an hour.
        St helens is just a little fart in comparison

        The glowing pumice and ash fountains along the fissures many kilometers tall
        A long line of plinian columns cannot be ruled out given the extremely heavy ashfall in India. Thats very heavy in places the ashfall
        Toba ash is light ashy grey due to high sillica content

      • How Toba looked like is hard to say: but awsome it was
        With many billions of volcanic ligthing flashes every half hour
        Toba must have looked like one Big hell
        An event almost 4000 times larger than Calbuco 2015….
        A long exposure photo of Toba in Action woud be an amazingh thing.
        Alot of volcanic lighting in the vents infront of red hot ashclouds

        Albert Toba was Lightning – Mania!
        Must have looked like some Lordi heavy metal cover
        I wish there was simualtor to simulate the vocanic ligthing in the Toba Event.
        Nothing is more spectacular than a red hot ash cloud plinian fountain column with bright ligthing flashes all over
        That video looks like mordor/armageddon its Calbuco 2015
        Tobas red orange hot columns woud be something to watch
        Its hard to grasp witht the 75 000 years old events eruptive rates.

        The whole region in sumatra woud have been layed to waste.
        Im soure the Tuff and igmigbrites are very thick aroud Toba.
        There was ceratinly humans around there it must have been pretty scary.
        In recent estimates Toba erupted more than 3000km3 of ash and pyroclastic materials
        Thats more than Vatnajökulls Icecap volume.

        If there was a plinian phase whos deposits that later was eroded away by the PDC s
        Then the ashclouds coud have reached almost 66 kilometers up in the fast model
        The plinian umbrella cloud covered much of south east Asia
        Impressive ash mammatus from the underside

  21. Latest update from Nyiragongo
    The lava lake is still overflowing like mad.
    This is beacuse of increased pressure in the magma system

  22. The impending Yellowstone disaster hype just a new level, and close to home! This, with an accompanying letter to Justin Trudeau (I will spare you), showed up in my mailbox from a concerned neighbour. I thought VC readers would find it entertaining.

    This was brexited to the dungeon by the resident deamon-in-charge. Released but with the link removed. As the submitter mentioned, the link showed a total lack of understanding of Yellowstone and readers interested in why Yellowstone is so safe are referred to Carl’s recent post on its staleness – admin

  23. My math dyslexia is so bad
    This pretty much makes it impossible to become volcanologist as volcanology is all about maths.

    Its strange that I manage to understand some physics despite my head is totalt kaputt.
    Cosmology I grasped instantly as a small child despite Half of my brain is dead…
    geology and volcanology I managed to grasp without complicated mathematics

    • Ah – a fellow maths dyslexic! Mine manifests itself as a number blindness. I can see the numbers, know what they basically stand for, but they kind of go into a blurry puddle of nothingness when I use them. I have absolutely zero understanding of the ‘language’ numbers represent.

      Curiously, I have a very good grasp of mechanics and some physics. I can understand a lot about stress points, motions and interactions, fluid movements and gas interactions. But I can’t represent them mathematically.

      I understand your frustration very well.

      • Amazing you managed to become a volcanologist

        IF You are Dr Clive Oppenheimer right???
        Famous Mount Erebus lover
        Is this the Clive I talks with here? 😊

        • Absolutely not! Dr Oppenheimer without doubt has far more intelligence than me! I’m just a retired librarian in England! 🙂

      • I struggled for years and thought I must be stupid because numbers go right out of my head. I can’t seem to remember them for more than a few seconds, let alone work with them above basic math. remembering a series of numbers (telephone numbers and such) is just not doable. Glad to hear there are others like myself.

        • For me its nightmare I cannot even read a clock properly.. and Im 24
          Not possible to be happy

          But its very good VC exist

  24. Earthquakes continues to be active in Grimsvötn
    This coud be indication of increased pressure in the upper magma body as it pushes on the bedrocks around it.

  25. Strong Earthquake (in the relative sense) ‘rocks’ Donegal Ireland.
    A magnitude 2.4 tremor was recorded of the coast of Donegal Ireland yesterday.
    While quakes of these (lacking) magnitude would not get a look in anywhere else it was picked up in the Irish media as a notable event (prehaps as a filler until the next Brexit News Report).
    Quakes (usually around this magnitude) occur most commonly in the northwest (Donegal) and South East (Wexford).
    The last ‘substantial’ quake in ireland was the 5.4Mag Lynn Peninsula of 1984 but sadly Ireland cannot claim this one as it was 60miles away in Wales UK.

    Perhaps there is an article here which could discuss the remarkable/unremarkable nature of ireland in relation to seismity and volcanism.

  26. Ancient history

    5th century eruption of Ilopango (around 450 AD)

    “A violent Plinian eruption occurred at Ilopango in the 5th century AD. It was the largest eruption in El Salvador during the past 10,000 years and one of the largest in Central America.
    The eruption was larger than the Krakatau 1883 or Pinatubo in 1991, and probably more comparable to Tambora 1815. It ranks VEI 6-7. It erupted more than 70 cu km of tephra, produced major ash fall and large pyroclastic flows that covered 10,000 sq km under 50 cm or more of pumice and ash. Pyroclastic flows and ash fall devastated an area of up to 100 km radius around the volcano.
    The eruption severely impacted the cultural history of the region: at the time, a sophisticated Mayan culture ruled and lived in the highlands of El Salvador. The eruption killed thousands, destroyed settlements and sent many to flee to the lowland areas in Guatemala and Belize. It ended the Mayan presence in the highlands, and it took probably several centuries for the region to recover.
    A major trade route controlled by the Mayas was abandoned and power shifted to Tikal. Modern excavations of some of the buried settlements have been bringing light into Mayan culture.”

    Meanwhile in Britain, the Romans packed up and left. The Emperor Constantine who gave us Christianity died and left 3 philosopher king sons, Constans, Marcus Aurelius and Uther, father of Arthur. Vortigern appointed himself regent and then ruled all Britain

    Germanic forces invaded from Europe

    “Vortigern fled to the extremities of Wales where he decided to build himself an impregnable stronghold on the southern slopes of Yr Aran, above Beddgelert (Gwynedd). However, after construction began, it was found that, each night, the previous day’s work was destroyed by unknown forces. Vortigern consulted the druids, who suggested he look for a young fatherless boy, born of the fairies, whose sacrifice would placate the gods and allow for the fortress to be finished. Vortigern’s men searched throughout Britain until just such a boy was discovered, most commonly it is said in Caer-Fyrddin (Carmarthen). The boy, called Myrddin Emrys, shortened to ‘Merlin’, laughed at the druids and explained that the building works collapsed because they were shaken by the battling of two fierce dragons buried beneath the mountain. The white dragon, representing the Anglo-Saxons, was defeating the red Welsh dragon, thus prophesying their eventual conquest of the majority of Britain.

    With the dragons removed, the fortress was completed, but Ambrosius Aurelianus seems to have have become the centre of resistance against both the Anglo-Saxons and Vortigern at this time, and he ousted Vortigern from his new home. He fled once more, but the site in which he took refuge is much disputed. It may have been to the Roman station at Old Carlisle (in Cumberland) or to an unknown fort in Gwrtheyrnion. The most popular traditions, however, indicate either the old hillfort of Little Doward, above Ganerew in Ergyng, or the hillfort of Tre’r Ceiri in Yr Eifl (the Rivals) on the Lleyn Peninsula. Wherever it was, Vortigern’s stronghold was miraculously struck by lightning and burnt to the ground, with Vortigern inside. He is sometimes said to have been buried in Nant Gwrtheyrn, also on Lleyn. This was around AD 459.”

    • “Meanwhile in Britain, the Romans packed up and left. The Emperor Constantine who gave us Christianity died and left 3 philosopher king sons, Constans, Marcus Aurelius and Uther, father of Arthur. Vortigern appointed himself regent and then ruled all Britain”

      When did Constantine live? 300s. When did Ilopango erupt. 500s. That’s 200 years in between. It’s like saying George III died and gave us three sons Prince Charles, Prince of Wales; Prince Andrew, Duke of York and Prince Edward, Earl of Wessex.

      Edited out text that failed to comply with our first rule (‘be nice’). -admin

      • Looking further the “emperor” referred to in this post was a self-declared usurper Constantine III. The Constantine who made Christianity the state religion of the Roman Empire reigned in the years up to 337. Nothing to do with Constantine III. Constantine III’s “connection” to the three named is an invention of Geoffrey of Monmouth, a very unreliable source.

        Ilopango’s most reliable dating would place it as the cause of the disasters of 535 and 536.

        Like I said 200 years in between the events. Even the earliest dating of Ilopango places it a century after Constantine the Great’s reign.

    • Volcanodiscovery still puts the Ilopango eruption in the 5th century, while more recent results favour a 6th century date. The size of the eruption they give remains plausible.

    • Perhaps a bit poorly phrased (the ability to clearly express oneself appears somewhat missing) but the actual question is a good one. Breathing on your hand with your mouth wide open feels warm. Doing the same with the mouth almost completely closed feels cold. It is basic physics but also affects volcanoes, as some erupt with mouth wide open and others explode through narrow openings. The latter loose a lot of heat in the process.

  27. How high can VEI 8 ultraplinian columns get? Bishop Tuff was a famous ultraplinian event
    Tambora reached 50 km I think. It produced a giant plinian eruption column, which is estimated to have reached more 40-50 km

    La garita and Bishop Tuff and Toba woud have reached higher if there was plinian phases
    Maybe these VEI 8 plinian columns ( Bishop Tuff ) very much reached the edge of space 70 km up.
    Its weird the ash particles can float on air thats below 1/1000 th of earths surface air density

    • Since there was no plinian phase, it does not really apply to Toba. And it appears that the plume itself reaches highest for not quite the largest eruptions. Make the eruption too big and the plume can’t carry its weight and collapses. The secondary plume (co-ignimbrite) gets higher in those cases, between 25 and 70km depending on whose model you trust.

  28. Where Felix Baumgartner Jumped in 2012 at 38 km attitude
    Air density is already down to below 1/1000 th of earths surface air density
    Ultraplinian columns goes way beyond this attitude

Leave a Reply