Mauna Loa from 1852 to 1868 – part 2 – The 1868 eruption and disaster

We continue where we left in the previous article. Last time we were looking at the eruptions of Mauna Loa in 1852, 1855-56, and 1859. These three large, closely spaced eruptions took place in the Northeast Rift Zone (NERZ) and north flank of the volcano. The next major flank eruption would happen in 1868.


The 1868 eruption

After the eruption of 1859 came a more prolonged dormancy of 5.1 years. An eruption occurred in 1865-66 at the summit, which raged on for 4 months. Following the 1965-66 eruption came a dormancy of 1.9 years until the eruption of 1868, which was in the SWRZ. This follows the “pattern” of Mauna Loa. This “pattern” is that summit eruptions will happen in between rift eruptions, and that rift eruptions will alternate between NERZ and SWRZ. The dormancy between summit and rift is shorter, and the dormancy between rift and summit is longer. This pattern would be cyclic. However, it is a behaviour that has been broken more often than not. Sometimes summit eruptions in a row have happened. Twice there has been a summit-NERZ-summit-NERZ series. There have also been two series of three unbroken rift eruptions, 1852, 1855 and 1859 in the NERZ, and 1916, 1919 and 1926 in the SWRZ. The plumbing of Mauna Loa is in constant change, which is probably the reason no stable, predictable pattern gets established.

The 1868 eruption started on March 27 at 6 a.m. It commenced from the summit as usual. One report mentions the location of this initial outbreak to be near Pohaku Hanalei, a lava cone in the uppermost SWRZ. The initial effusion was over in a matter of hours. Frederick S. Lyman witnessed this eruption from Kau:

“You may have already heard that an eruption broke out on Mauna Loa, a little to the south west of the summit, about 6 o’clock last Friday morning, 27th inst. It gave no forewarning; the fire burst up out of the ground, throwing a spray of red lava high in the air, then a great column of smoke rose straight up thousands of feet and arched over to the east. In a few minutes a new jet was thrown up a little S. E. of the first, with its column of smoke ; soon followed by another jet; and then by a fourth ; soon the red lava began running down the sides of the mountain in four streams in a southerly and easterly direction. About seven o clock we began to hear a roaring sound, which grew louder and louder until the air seemed to tremble with the incessant roar of the volcano; but it finally subsided, and ceased entirely about eight o’clock.”

That very same day, in the evening, a large series of earthquakes started rocking the Kau district, which makes up the south and east slopes of the massive mountain. The next day, March 28, two enormous earthquakes occurred in quick succession, estimated at M 6.1 and M 7.0 in the Catalogue of Hawaiian Earthquakes. A letter from Kau in the Pacific Commercial Advertiser details the earthquakes:

“The earthquakes commenced on Friday night (Mar. 27); some say we had thirty or more shocks. Saturday morning I awoke about daylight, and up to 1 P. M. I counted ninety-seven earthquakes; none were very severe, but their frequency frightened us. While at dinner we had a heavy shock crockery jingled and the house creaked like a ship in a storm. I left the table and went over to _ and while talking with them we had a fearful shock. I caught up one child, and another, and little _ and _ followed, running out of the house, while vases, books, boxes, lamps and dishes were falling about us. When we got out, walls were falling down with a thundering noise, and the air was filled with dust, the earth still quivering. In less than twenty-four hours we had over two hundred shocks!”

According to Abraham Fornander, who was at the Volcano House that day:

“The first (heavy) shock was felt throughout, Kau, Puna and Hilo. At Keauhou, the ground shook continually all the afternoon on Saturday. In Kau, the shock seemed to have been stronger the more southward you went. At Waiohinu, it shook down walls, and cracked the stone church from top to bottom. At Kahuku, Capt. Brown’s place, the dwelling house, cattle pens, stone walls, etc., were thrown flat on the ground fortunately no one was hurt”

Mrs. Sarah Joiner Lyman, who lived in Hilo, kept a diary of felt earthquakes. This diary, which was later continued by her son and his wife, records a list of earthquakes felt in Hilo for eight decades. She writes:

“Saturday (3/28) is a day ever to be remembered here: in less than half a day, there were 51 decisive shocks, one of which was the strongest shock that the oldest inhabitant has experienced . . . This occurred at 1:28 pm. This lasted for some time, causing everything to rock backward and forward”

What exactly was going on? The smaller frequent earthquakes being felt on the night of March 27, and the morning of March 28, would have probably been caused by a magma intrusion propagating into the Southwest Rift Zone. The intrusion had broken out near the summit of Mauna but was now making its way underground. Frederick S. Lyman mentions a line of smoke propagating down the rift zone. This line of smoke may have already reached a location just above Kahuku by March 29. But because he may have had later knowledge of the lava breaking out near Kahuku I’m not sure how reliable this is.

The flank of Mauna Loa is slowly sliding away into the sea. The way it does so is more complicated than a simple slump. I will explain more about this later. The intrusion had triggered the faults under the mobile flank into the major M 7.0 rupture of March 28. Earthquakes continued, now aftershocks from the M7.0 were likely being mixed up with quakes from the continuing intrusion.

“On the 30th and 31st, and April 1st, a number of shocks were felt at Hilo, some of them sufficiently powerful to awaken persons from sleep, causing furniture, etc. to rattle”

In Kau people were being awakened multiple times each night by vigorous shakes. Some considered that lava was trying to break out, and until it erupted, there would not be much rest. Unfortunately, the worst earthquake was still to come:

“The heaviest shock occurred Thursday, April 2nd, being the same that was felt so sensibly at Honolulu. This destroyed every church and nearly every dwelling in the whole (Kau) district. From 10 to 12 o’clock of that day there had been
service in the large church in Waiohinu, and it was crowded with people. Only four hours after they left, the heavy shock came, the walls tumbled in, and the roof fell flat all the work of twenty seconds. At the same instant, every man, woman and child was thrown off from their feet. Horses and cattle dropped down, as if dead. A man riding on horse back had his horse tumble under him so suddenly that he found himself and horse lying flat on the ground before the thought of an earthquake entered his mind.”

The earthquake of April 2nd would be known as the Great Kau Earthquake, is estimated as M 7.9, and is the largest earthquake in the history of Hawaii. Professor Frederick S. Lyman found himself near the epicenter of the earthquake in Keaiwa, close to the present day location of Pahala:

“First the earth swayed to and from north and south, then east and west, round and round, then up and down and in every imaginable direction, for several minutes; everything crushing around us; the trees thrashing about as if torn by a mighty rushing wind. It was impossible to stand, we had to sit on the ground, bracing with hands and feet to keep from rolling over. In the midst of it we saw burst out from the pali about a mile and a half to the north of us, what we supposed to be an immense river of molten lava, which afterwards proved to be red earth, which rushed down its headlong course and across the plain below, apparently bursting up from the ground, throwing rocks high in the air, and swallowing up everything in its way, trees, houses, cattle, horses, goats and men, all in an instant, as it were. It went three miles in not more than three minutes time, and then ceased. Someone pointed to the shore, and we ran to where we could see it. After the hard shaking had ceased, and all along the sea-shore, from directly below us to Punaluu, about three or four miles, the sea was boiling and foaming furiously, all red, for about an eighth of a mile from the shore, and the shore was covered by the sea.”

The rupture of the earthquake propagated across the entire south flanks of Mauna Loa and Kilauea and relieved compressive stress than had been built along the shore of the island from the upper flank, creeping downward and pushing against the locked portion of the flank. The coast subsided 1-2 meters for a distance of 110 km, from the east point of the island, Kapoho, to the south point, Ka Lae. Near Kapoho coconut groves now stood underwater at high tide. The subsidence as far as Kapoho can only be explained by the flank of Kilauea shifting away together with that of Mauna Loa. Likely this same fault later ruptured in 1975 and 2018.

At the same time as the subsidence, the sea rose up, pushed away by the motion of the submarine slopes. A tsunami claimed 46 lives. Villages were entirely destroyed by the tsunami and coastal subsidence. Some villages weren’t resettled, which was a good choice, for a 1975 earthquake again produced a tsunami in those same areas. Old volcanic ash from Kilauea, the Pahala Ash, was triggered into many mudslides from Kapapala to Waiohinu. One of these mudslides, described by Lyman above, caused 31 fatalities.

At the time of the earthquake, Kilauea had a lava lake that covered much of the caldera floor. It had filled and drained multiple times. After the previous draining in 1840, it became concealed by a large thickness of lava flows and had pushed up the centre of the caldera like a piston about 100 meters by 1846. This activity had continued up until 1868, with decreased strength. The M 7.9 April 2nd earthquake created a sudden extension across the rifts and summit of Kilauea, and it was seemingly strong enough to start a sudden dike intrusion. An hour after the earthquake, lava erupted in the Kilauea Iki crater. A dike formed from the lava lake and followed the typical trajectory of Kilauea summit dikes. East into Kilauea Iki, and also west into the Southwest Rift Zone of Kilauea. In a few days the whole central region of the caldera had fallen by almost as much as it had risen since 1840, leaving an empty bowl as the lava lake drained away, which would be the final destruction of the caldera-wide lava lake that had grown since 1790. Some small amount of lava erupted later, on April 9, in the Southwest Rift Zone. The physician William Hillebrand described Kilauea after the earthquake:

“The great south lake Halemaumau is transformed into a vast pit, more than five hundred feet deep, the solid eastern wall projecting far over the hollow below . . . More than two-thirds of the old floor of Kilauea has caved in and sunk from one hundred to three hundred feet below the level of the remaining floor.”

Map of Kilauea caldera after the 1868 earthquake, the whole central portion has caved in and is shown in a darker colour. Map is from Hillebrand.

This rather unique series of events was still far from over. On April 6, at 4 PM, earthquakes in the Southwest Rift Zone of Mauna Loa suddenly reinvigorated, as recorded in the earthquake catalogue, a series of numerous felt earthquakes had started, with 8 events within just an hour. I think at this time the summit of Mauna Loa may have started collapsing. That very night, an explosive eruption scattered ash across the Kau district:

“During Monday night [April 6-7], prior to the eruption, the ground throughout the district was covered with a coating of fine sand and light pumice stone, of a light yellowish color. Where this shower of sand and pumice stone came is as yet unknown”

“April 7th.–The deck [of a ship anchored at Kaalualu] covered this morning with very fine ashes”

The only eyewitness report of the eruption seems to have been from Abraham Fornander, who saw the eruption from Hilo:

“April 7. Last night, between 12 o’clock and daylight, several right smart shocks were felt here. I was awaked four times by the shocks and the rattling of things in and about the house. At 6 o’clock, A. M., as I and others were standing near the Wailuku bridge, looking up at Mauna Loa, a heavy, dense, dark column of smoke, deeply tinged with red at its base, rose from near the top of the mountain, apparently from the southern side of Mokuaweoweo. Though afterwards the smoke grew thinner, yet it remained visible for nearly an hour, with every now and then that lurid glow of red at its base.”

I will explain below why I think the summit of Mauna Loa may have collapsed during this explosive eruption. Magma was on the move again on April 7. That very morning a crater opened up some miles above Kahuku, and seemingly it did so near some farms. When I was reading this report, I recalled a comment from Chad in part 1 of this article. He had found a pit crater that lined up with the 1868 fissures, that more or less could fit with the reported location of the new crater of April 7 that I was reading about. There is a description of the crater:

“The new crater, when visited by Mr. Swain, was at least one and a half miles in extent, nearly circular, but constantly enlarging its area, by engulfing the sides. While the above gentleman was looking at it, a tract of at least five acres In extent tumbled in and was swallowed up like food for the devouring element. The enlargement is going on mainly the lower side, towards the farm houses.”

The crater of chad is smaller than the one described here, but it is circular and located more or less where it should be according to the few descriptions. Horizontal distances are often exaggerated in these old historic reports. The crater in reality is only 150-180 meters wide. Reports shows it grew like a sinkhole. Ground collapsed into it. These pits are sometimes found along dikes. They are very common in Mars too, called catenae. It was likely formed due to the rock above the dike collapsing into the magma and being carried downstream..

The formation of this crater shows magma was rapidly moving down rift. Somewhere at 6 PM lava broke out 4 km downslope from the earlier crater. It did so at elevations of 920-670 meters. No eruption had taken place this low in the Southwest Rift Zone for more than 2000 years:

“On Tuesday afternoon, at 5 o’clock, a new crater, several miles lower down, and about two miles directly back of Captain Brown’s residence, burst out with a heavy roar and frightful crash. The lava stream commenced flowing rapidly down the beautiful plateau, towards and around the farmhouse, and the inmates had barely time to escape with what clothes they had on, before the houses were all surrounded and enclosed with streams of fiery aa lava, varying from five to fifty feet in depth. Fortunately all the inmates escaped safely to Waiohinu, but how narrow the escape was, and how rapid the stream flowed, may be inferred from the fact that the path by which they escaped was covered with lava within ten minutes after they passed over it.”

“At six p.m., when the point was about ten miles astern, bearing E. by S., a volume of flame shot up from the mountain Loa, in what appeared to be the neighbourhood of Kahuku. The heavens were lighted up at once, and the reflection extended rapidly in the direction of Waiohinu and Kaalualu. . . It reached the sea somewhere in that direction at nine and a half p.m., when an immense body of steam at once arose, through which flashes resembling lightning were constantly darting as long as we were in sight. The top of the mountain was concealed by the dense clouds of smoke (maybe dust or ash from ongoing collapse?).”

The lava advanced rapidly near the fissures as thin flows of aa lava, Lava made small streams, creating numerous islands that were not flooded, called kipuka. One family was apparently trapped inside their house in a kipuka for ten days. Several cattle were reportedly stranded in these islands for days until they were rescued.

Where lava rushed into the sea, it exploded and ejected pyroclastic material that built cinder cones around the two principal entries. The largest with a crater diameter of 400 meters. A perpetual thunderstorm played above the eruption, fed by steam from the ocean entry. This plume could be clearly seen all the way from the neighbouring island of Maui, as shown in the following excerpt:

“During the night of April 7th a bright but varying crimson light over the volcano was visible from the Seminary at the distance of one hundred and twenty statute miles as measured on Wilkes’ chart. This light was a reflection from a mass of cumulus cloud through which vivid lightning was constantly darting. After daylight and through the morning of the 8th, this stupendous column of cloud was visible pouring rapidly up to the ether, with ever varying shape. It was usually well defined on the westward side, where it, at times, presented a perpendicular wall of miles in height. On the east it was ill-defined. Above, it often spread out, especially toward the east, as if borne off by the southeast wind of the upper air. The base, so far as visible, appeared to be commingled with murky brown strata. The apparent altitude of this cumulus above the horizon, when at its highest was 3°30′ which reduced for a base of 120 miles with 500 feet altitude of the point of observation, gives a height of 7.8 miles.”

Mr. H. M. Whitney visited the eruption on the 10th of April and gives a great description:

“On ascending the ridge, we found the eruption in full blast. Four enormous fountains, on a line a mile long, north and south, were continually spouting up from the opening. These jets were blood-red, and yet as fluid as water, ever varying in size, bulk and height. Sometimes two would join together, and again the whole four would be united, making one continuous fountain a mile in length. From the lower end of the crater, a stream of very liquid, boiling lava flowed out and down the plateau, a distance of two or three miles, then following the road ran down the precipice at an angle of about 30º, then along the foot of the pali or precipice, five miles to the sea, the stream being about eight or ten miles in length, and in some places half a mile wide. One peculiarity of the spouting was that the lava was ejected with a rotary motion, and as it ascended both lava and stones rotated always in one direction towards the south. . . It lasted only five days, the eruption ceasing entirely on the night of the 11th, or morning of the 12th. During its continuance, the atmosphere was filled with smoke so dense that the sun appeared like a ball of fire, and the whole island was shrouded in darkness. . . As the lava entered the sea, clouds of steam and smoke rose up, and flames of bluish fire were emitted, rising from the water to a height of from ten to twenty feet.  During the night we were at the volcano, the air was highly charged with sulphurous gas and electricity, and frequent flashes of lightning were seen directly over the lava stream, accompanied with short claps of thunder.”

The volume of the 1868 eruption has never been properly estimated, given that most lava was dumped into the sea. With the help of bathymetric data, I measured the submarine lava deltas of the 1868 eruption the best I could. There are two lava deltas, the largest, the northwest delta, has about 0.14 km3, while the southeast delta has 0.08 km3. The subaerial volume is about 0.065 km3. Taken altogether, the eruption totals roughly 0.25 km3 or about 250 million cubic meters of lava, which does put it as one of the largest eruptions of Mauna Loa, although still behind 1859, and probably behind 1950 and 1880-81 too.

Approximate extent of 1868 submarine lavas.

The flank eruption of 1868 was, however, of remarkable intensity, lasting only a little over 4 days, the averaged eruption rate was 700 cubic meters of lava per second, and the ocean entry may have been the most intense Hawaii has seen in historical times.

Did the 1868 eruption and intrusion make a caldera?

There are several lines of evidence that 1868 was a caldera collapse of Mauna Loa. The caldera of Mauna Loa, as we see it today, formed probably in 1710, in the Hapaimamu eruption. In this caldera, there were several pits, presumably from different collapsed magma chambers. The Inner Pit was the deepest, placed in the centre of the caldera. A nested crater. It is to Mauna Loa what Halema’uma’u is to Kilauea. The Inner Pit is known to have collapsed again at some point between 1841 and 1874. A detailed survey of the caldera in 1874 found the Inner Pit to be 80 meters deeper than the previous survey in 1841, even though the pit had been rapidly filling with lava since 1871, and that summit eruptions in 1849, 1851, and 1865, had also added lava to the crater. No one has attempted to link this collapse to any particular eruption.

Of the eruptions that took place between 1841 and 1874, only 2 of them had very high eruption rates and occurred from low elevations, 1868 and 1852. Other flank eruptions in 1843, 1855, and 1859 were slow and long-lived. The 1852 eruption doesn’t seem to have been particularly big, nor is its intrusion too exceptional. Instead, 1868, had a massive intrusion, as I will explain, possibly the largest historically. The eruption itself was remarkably intense. Its lava carries as much as 30% olivine, a higher content than other historical eruptions, likely showing how deeply it drained the storage of the volcano. 1868 erupted from lower elevations than major prehistoric caldera-forming events, like Hapaimamu or Panaewa. This alone makes 1868 the preferred candidate for collapsing the summit.

Mokuaweoweo in 1885. The nested crater in the centre would be the one that collapsed in 1868.

Caldera collapses are often linked to explosive eruptions. This was seen at Kilauea when it collapsed in 1924 and in 2018. The shield Piton de la Fournaise in Reunion has also usually exploded when its summit collapsed, as well as Karthala, Fernandina and possibly others. Mauna Loa has also sourced explosive eruptions. Explosive eruptions of unknown timing have scattered blocks of lava and gabbro, as large as 2 meters across, around the Inner Pit of Mokuaweoweo. The April 6-7 explosion of Mauna Loa described from Hilo came from the summit. This explosion, given that there is no other known mechanism that could have triggered it, would have likely been caused by a collapse of the summit of Mauna Loa, and may or may not have contributed to the ejected blocks around the Inner Pit.

Lastly, lava erupted in the caldera in 1871 and ponded into a lava lake that gradually rose up within the Inner Pit until 1877. When you open a hole the volcano will try to fill it up. We’ve seen this with Kilauea after its 2018 collapse, with a huge lava lake that seems to be intent on filling up the caldera. Other past collapses of Kilauea have filled rapidly too. No historical summit eruption of Mauna Loa has lasted beyond 5 months, except for 1871-1877. This very unusual activity must have been the response of the volcano to the Inner Pit collapsing into a crater.


Why so many earthquakes?

The 1868 intrusion stands out for the dramatic earthquake swarm it triggered. Tens of earthquakes were being felt in Kau within a day of the initial summit outbreak, and before any of the major flank ruptures happened. Reports claim more than 3000 earthquakes were felt in 12 days. Highly exceptional. As I mentioned in the previous part about the intrusions of 1852, 1855, and 1859, none of them had any felt earthquake. At the time of the 1984 eruption in the Northeast Rift Zone, there was already a dense seismic network in place. As recorded by the USGS catalogue, the 1984 intrusion produced just five M 3 earthquakes up at the summit, no located earthquakes along the rift, and just a few M 2 earthquakes down the flank, that are too small to be felt.

Another oddity of the 1868 intrusion is that the crater and fissure of April 7 are offset from the “main” rift, about 7 kilometres off to the east. The main rift here refers to the line followed by dikes that intrude directly from the summit, following the direction dictated by stresses, like in 1950. The direction of the dikes seems roughly common to all fissures and fractures of the rift zone, including those of 1868. The problem is, a dike following the direction of the rift cannot possibly end in the location the 1868 intrusion did. A more complex intrusion geometry is needed. While there might be more than one possible option, there is only one I can see that solves both problems at the same time.

It is important to understand the structure of Mauna Loa. The flanks of the volcano are sliding away from the summit and rifts. The southeast flank of the volcano seems to be ever so slowly sliding away.

Arrows show deformation of the summit, note how in 1997-2002, when there was only slight deflation of the summit, the volcano was slowly moving to the southeast. From the USGS (Michael Poland).

Mauna Loa’s SWRZ is surrounded by a ring of faults, normal, reverse and strike-slip, and their main function seems to be accommodating outward slip of the volcano. The majority of the faults seem to deep inwards towards Mauna Loa, and are, as such, reverse faults. Many large Hawaiian earthquakes, like 1975 or 2018, have thrust focal mechanisms. So yes, the flank is sliding up slope. Is this even possible? I think we are not dealing with normal slumps, but that instead magma aids in this motion. Below is a map of the faults. I mapped these faults with the catalogue of Robin Matoza, which relocates the earthquakes, and dramatically sharpens their coordinates.

Red lines show fissures and fractures which develop where magma intrusions intersect the surface. Coloured lines are strike-slip faults which reach from the surface (red) to a depth of 10 km bgs, below the ground surface (dark blue). Polygons are subhorizontal faults, mostly reverse faults, with frequent seismic activity. White is shallower (up to 400-600 meters bgs), and black is deeper (13.6 km bgs or more). Pink arrows show direction of slip: 260 º in the west flank of Mauna Loa, 140º in the southeast flank of Mauna Loa, and 155º in the south flank of Kilauea.

There are many superimposed faults. Kilauea’s south flank has two or three major overlapping thrusts. There are at least 5 major faults running under Pahala, and this is not even considering the mayhem happening deeper down. These faults are probably inter-connected, the areas in between them probably move only during major ruptures. The faults shown are only those being affected by slow continuous slip, which generates strain and many small frequent earthquakes. The south flank of Kilauea for example has had ~40 very small quakes this week. At Kilauea the slip of the upper flank pushes against the lower flank and raises it up. We can see this in GPS data or InSAR. Apua Point, at the coast of the island, next to Kilauea’s most actively seismic part of the flank, has been uplifted by about 10 cm since the 30 cm subsidence during the 2018 earthquake.

But what is most remarkable about Mauna Loa is the enormous area of aseismic slip east of the SWRZ. And I think this is very important. For a distance of 12-16 km from the rift, the southeast flank slips in an almost completely aseismic manner, that then encounters an arc of strong, continuous seismic activity which encloses the area. A further clue as to what this area is comes from gravity. Studies have discovered that the area south and east of the rift has a high gravity anomaly, meaning there is dense heavy material under the flank. Presumably gabbro and dunite, which are dense crystalline cumulates deposited within intrusions.

Map which shows earthquakes 0-13 km deep from Robin Matoza’s catalogue. Red is shallower, purple deeper. The red polygons show areas with particular gravity anomalies, adapted from Zurek, 2015, which is linked below. The dense intrusive area is surrounded by an arc of earthquakes which in places makes a very well defined boundary at 8-11 km bgs.

This is what I believe happened. The structure is like a magma lubricated slide. The upper flank slips over the magma and/or crystal slurries and exerts pressure against the lower flank, causing frequent earthquakes where the two portions clash. I think in 1868 an inclined sheet intrusion started at the summit and then propagated downward through the slump detachment surface. It may have reached several kilometres downward, approaching the seismic arc. The upper flank would have started sliding faster over the intrusion, creating strain across the seismic arc, and numerous earthquakes. Eventually, the force would have been large enough to shift the entire flank away, probably all the way to the submarine edges of the volcanic pile. This was the M 7.9 earthquake. When the leading edge of the intrusion reached the end of the slump, it switched to growing a a vertical dike, erupting offset to the east from other fissures. The following is the speculative geometry of the intrusion:

Intrusions in 1887 and 1907 would have also intruded the slump structure and erupted offset to the east. 1887 also generated a massive earthquake swarm, with hundreds of felt earthquakes in a matter of days, although none as large as the March 28 and April 2 earthquakes of 1868. 1887 and 1907 intrusions were increasingly shallow. From 1916 onwards, intrusions in the SRWZ would have been mostly shallow vertical dikes. An intrusion like 1868 is probably a rare event, and in fact last time lava erupted so far down and offset from the rift was more than 2000 years ago.

I think that slip over outward dipping sheets of magmas and crystal slurries is the force driving the reverse faults around Mauna Loa. It would be an improvement over the deep rift model, where these forces seem to be somewhat unclear. I took inspiration from an article on Piton de La Fournaise by Quentin Dumont and his team. This article shows how sill intrusions of the past 20 years at Piton de La Fournaise have intruded downwards into a detachment of the flank and created slip. I found this article very insightful. Piton des Neiges, the ancient volcano of Reunion that came before Piton de la Fournaise, has known exposures of sills intruded into a detachment/slump of the edifice, on top of a large gabbro intrusion that is perhaps akin to the high density material under Mauna Loa’s SE flank. Related processes may be at work in Hawaii.

The 1868 eruption and intrusion was a disastrous event for Hawaii. It is probably something that will not repeat again for a long time. There is much to learn from it, and there is much we will never know. In these past 2 articles, I have looked at Mauna Loa and shown what a fascinating volcano is. And also terrifying when it wants to. While people wait for the next eruption of Mauna Loa I look at the past and see many eruptions, and am amazed at the scale and complexity of the volcanic processes shaping the island of Hawaii.

Links and references:

Various releases of the Hawaiian Gazette and Pacific Commercial Advertiser that cover the 1868 eruption:

Catalogue of Hawaiian earthquakes (Klein and Wright):

Holocene eruptive history of Mauna Loa volcano, Hawaii:

A very personal account of the 1868 events:

Relocated catalogue of Hawaiian Earthquakes (Matoza):

Zurek et al, 2015, article with analysis of bouguer gravity data:

22 years of satellite imagery reveal a major destabilization structure at Piton de la Fournaise:







118 thoughts on “Mauna Loa from 1852 to 1868 – part 2 – The 1868 eruption and disaster

  1. This is one of the first ideas I thought of for a VC article. Now the 1868 eruption has a deep analysis with a modern understanding of volcanology, something that was thoroughly needed, I think.

    • Your idea for how the earthquake originated makes it almost like a subduction quake.

  2. Great stuff good read yes magma sourely must act as lubricant during flank intrusions and slips of Hawaiian edifices .. specialy knowing how low in viscosity this lava is

  3. Chad’s crater 🙂

    I was not aware it had been observed forming though, my assumption is it was maybe a tiny early vent that caved in, or was an erosional structure from the lava as you say but formed during the eruption.

    Fascinating how there was no cone created at the vent, and yet it had some massive fountaining. Even in 2018 where the eruption rate was similar a cone had started to form around fissure 8 in less than 5 days. Maybe the lava of Mauna Loa has less volatiles than st Kilauea, so the high fountains in this case are a result of the eruption rate and not degassing. Nyiragongo eruptions are the same, always described as having strong fountaining but with no or very rare spatter cones. And I had put this idea up for 1823 and 1840 at Kilauea too, that there was strong fountaining but not as gas rich jets but more like violent dome fountaining, or a jet nozzle on a hose. The Great Crack must have been eroded by the lava erupting this way. There are a lot if Mauna Loa fissures that look very similar to this too.

    • The best way to eradicate a cone is by collapsing it into a crater

      • but the crater in question is not on the actual fissure that erupted, it is uphill a few km. Now that I think about it there is probably no vent there if people were investigating it on the day… My comment on fountaining was a reference to the main fissure, active for 5 days and with fountains tall enough to see beyond Kahuku pali (in the 100+ meter range or more) but no evidence at any of the fissures that a cone began to form, or even that any part of the fissure had begun to dominate over the other segments. Even in 1950 the fissures there had begun to make cones within 5 days. It is like the 1868 fissure was going at full force sustained for its entire duration and then stopped very suddenly with no warning.

        • As far as we know 1868 never focused into a single vent. When the fissure was described in April 10, a day before it ended, there was still a curtain of fire. It may be easier to build cones around isolated fountains than a continuous curtain of fire. The eruption rate probably hundreds of cubic meters per second when the eruption ended, such intensity of effusion along a large fire curtain, all the way to the end, was probably why no cones were formed around the 1868 vents.

          Its like trying to accumulate snow on a river, it is impossible. A similar thing probably happened during the 1868 eruption as spatter from the fountains fell into a mass of flowing lava.

  4. Right.. If the fountains are powerful and dense enough .. they dont form cones as everything flows away in clastogenic flows. Hawaiian lava are also extremely fluid. These fast fissures also generaly dont last long enough to form cones either. Ionian lava fountains like Pillans recent eruptions may reach many many tens of thousands of cubic meters a second perhaps way above that as well, really powerful lava fountains from narrow vents only form pyroclastic fall sheets .. but and Mauna Loa fissures it flows away .. as it woud do on IO as well in many cases

    • The Pillanian eruptions are probably plinian eruptions in a vacuum. Without the air to cool the lava and carry it away the fine particles still behave as a fountain, so can flow away. But an eruption this powerful on Earth would have probably been explosive unless it was a long fissure. Tarawera in 1886 was a similar eruption rate and fluid magma but was explosive.

      Looking at Io actually there arent really many huge fast effusive eruptions, the eruptions are relatively passive. Those plumes that go hundreds if km, might nit be actual eruption plumes rather are the SO2 rising and then freezing out, so a bit more of an atmospheric effect than anything to do with an explosive eruption. I guess really, apart from overflowing lava lakes or domes (of which the latter is not found on Io) any other form of volcanism is dependant on being erupted into a nedium thst isnt a vacuum. So Ionuan eruptions really will all only be Hawaiian to varying degrees of intensity no matter the scale.

      It would be interesting if ignimbrites can happen on Io. The ignimbrite would erupt fastlike it does here (although only 1/6 as fast) but once erupted there would be the same problem as above, no air to cool down the lava, so no pyroclastics. So the ignimbrite would still be a liquid, so would be like a lava flow just one erupted far faster than possible in any other situation. The 2015 thermal signatures on Pillan Patera trace an arc around the volcano, much larger than the flow observed in 1997 or anything else since then up to that year, maybe this is exactly what happened. Looking at a map the 1997 eruption there was the first of many huge eruptions from that year to 2015, and possibly the smallest… will be very interesting to see the damage done, I would nit be entirely surprised if there is a new caldera, and the old Patera is gone.

    • Pyroclastic flows are hot fluid dynamics basicaly hot fluid currents in a fluid, the hot ash particles heat the air around them and you get these dangerous convective currents. you can do that in a tank of freezing cold water and then inject hot water at near boiling point .. from a pipe at the side thats full of clay particles.. and you have your PDC. Atmospheric fluid dynamics. On IO there is No atmosphere to heat up and you may not get these currents at all..

      Pillanian eruptions are probaly long Fissure fire fountain eruptions as seen with Twasthar, rather than single vent eruptions

      • I mean, a ring dike or cone sheet will erupt as basically a long fissure, in the case of the former it will become a complete ring. This eruption was maybe not an ignimbrite than, but it might have been an eruption that was an order of magnitude bigger than the 1997 event, or something like 500-800 km3 of lava. Given that the heat was detected for 3 months, this might be how long the eruption lasted for. In that case the eruption rate was about 10 km3 a day… So this was basically a VEI 6 intensity effusive eruption. It is also possible that the eruptions were a lot faster too, there is no signal from April, it might have been two eruptions in February-March and then again in May, much shorter than 2 months…

        Pillan Patera might then be an example of a real LIP volcano in action, with fast eruptions not just lava lakes overflowing. I guess we will find out for real when we get new images.

      • Pillan Patera filling filling the Patera as lava flowed into it. The Patera is 70 km wide and haves 3 km high walls.
        That means 3 km high lava falls .. must have been a magnificent sight, with many more souch eruptions the Patera Maybe more or less destroyed today

  5. Fascinating story. I’ve never before heard of an earthquake as large as 7.9 associated with a volcano. But the pressures must’ve been immense, off any scale from those observations.

    The observations and stories from the time are excellent (and scary!) They remind me a lot of the anecdotes and descriptions collected after Krakatoa erupted in 1883 only fifteen years later. The British Royal Society pulled together all these reports and observations into a publication, which is a very good read: The eruption of Krakatoa and subsequent phenomena (1888)

    • Thanks! I too am fascinated by old descriptions of volcanic eruptions. I have read the two letters of Pliny the Younger about Vesuvius many times. The eruption of Krakatoa and subsequent phenomena is a read I would greatly enjoy.

  6. It looks like more lava lakes might start appearing in Halemaumau. One of the areas os persistent degassing in the center of the crater has collapsed to reveal a glowing skylight, and it has been there for a logn time now, suggesting the lava within it is actually circulating. The lake is hardly rising anymore so now the crust has a chance to collapse into the lake more as the lava circulates under it and perhaps erodes the crust from below. It is already known the crust is rather porous and fragile as it is persistently floating, presumably this material is easy to erode as it was in the case of Pu’u O’o.

    The underground filling of magma storage at Kilauea continues too. Pu’u O’o and the middle ERZ are now showing extension across the rift, and every station west of there to the summit is showing an eastward movement reflecting the inflation of the deep storage under Kilauea. Not a lot of magma is being pushed up under Halemaumau but the connection seems to be open and free given the main lake still exists and has not changed in well over a month now. But there is pressure building, some point in the next year (likely much sooner) there will either be a brief but powerful intensification of the summit eruption, or there will be an intrusion somewhere on the flanks of the volcano, possibly with an eruption.

    All of this while Mauna Loa is also inflating, not quite so much but far from negligible and it has been inflating persistently now for nearly 20 years. It will be interesting to see what happens, especially to see the reaction from the volcano that doesnt erupt first.

    • There is also a cracked overhang close to the lava lake vent, I Will make many screenshots of this over a month and see If its moving like a flipbook. Coud it collapse into the lake ?

    • There are three small lava pools now around the larger pool. Getting more and more like 19th century Kilauea.

  7. Right …. I see now holes too there
    Woud be fun If the 10 m thick crust overturned, it coud do as it gets denser and older, althrough is probaly extremely pourus

  8. Scanning Inside Volcanoes with Synthetic Aperture Radar Echography Tomographic Doppler Imaging:
    This method could have a big impact in many research fields. In fact they are using this method for many other things, for example they scanned cheops pyramid discovering unknown internal and underground structures (
    IMO it’s an example of military knowledge becoming public.

  9. I noticed that the majority of cross sections of Hawaii have all the slump faults going down with an arc, being near vertical close to the surface but curving away from the island going down to eventually be near horizontal, either merging with the decollement fault or above abd parallel to it.

    In the case of the 1868 intrusion it looks like magma basically got into one of those faults, and when it reached the end it formed a more normal dike. Possibly, the eruption up near the summit was on the fissure swarm of the 1949 eruption, being part of Mokuaweoweo but not actually in the pit, so sort of a separate summit intrusion but at the same time as the deeper rift intrusion. These sorts of models make alot more sense than the idea I was going for with veryical dikes going at weird angles to the rift, here the vertical dikes are at expected angles just not directly in line as they would be starting at the summit.

    • Yes, it would be similar to what has been proposed for Piton. A slump structure which curves from vertical to horizontal as it goes deeper. As the flank slips on top of the magma extension is created in the slump. The vertical part builds extension more rapidly and is often intruded by near-vertical shallow dikes, in typical SWRZ eruptions. The horizontal part also builds extension, but much more slowly, and is very rarely intruded by sills/inclined sheets of magma, like the 1868 and 1887 intrusions. I also think dense olivine/magnesium rich material from the magma chamber may flow downward into the slump as crystal slurries or very dense melts. Deformed olivine crystals which are common in Hawaii might form this way.

      It is also likely that Kilauea has a similar structure given that it has the same flank slip dynamics. A puzzling thing about Kilauea is that since 1970 all intrusions have taken place at depths less than 3 km deep below the surface. The only exception being the 1982 deep intrusion, but that was an unusual slow-intrusion, and it is unclear whether it started at the summit or not, probably it didn’t. And yet the flank moves away from Kilauea at a depth of 6-8 km, becoming deeper towards Mauna Loa, so there is a lot of space being opened below 3 km, when does it fill? A possibility of course is that this area fills slowly by dense crystal slurries which slide down from the summit. However it is also possible that intrusions like 1868 can happen at Kilauea and fill up the slump structure down to the start of the reverse fault at 6-8 km depth.

      No certain example of this happening exists. But I do have considered 1832, which could well be the largest draining of Kilauea in historical times, possibly as much as 1 km3, or the second-largest after 2018. But no one really knows where it went. We know it started at the summit and intruded eastwards, erupting in Kilauea Iki, and opening fractures east of Kilauea Iki and in Puna, it also produced powerful M 6 earthquakes in the south flank. Another time a very large volume of magma/lava disappeared into the ground without erupting much, was in 1868. So as a result of the M 7.9 earthquake of Mauna Loa, lava may have also intruded deeply under Kilauea in 1868. Other possible cases are 1823 and 1840, 1823 was an enormous 0.5 km3 draining with a M 7 earthquake in the south flank. 1840 was able to erupt in the East Rift Zone even though the East Rift Zone Conduit was probably obliterated from 1790, and also had a very large intrusive volume of 0.3 km3? and erupted very rare picrites. It would be interesting to speculate 1790 may have been like this. And that 1823, 1832 and 1840 may have been follow ups to the 1790 event, same as 1887 or 1906 may have been follow ups of 1868. One very brief report mentions that 80 earthquakes were felt the same day as the 1790 explosive eruption, some strong enough to throw down houses.

      Intrusions starting from the Halema’uma’u-Kilauea Iki area and intruding into the slump structure down to 7 km below the surface, maybe even reaching into the East Rift Zone, or the Puna Ridge, and erupting there, is a possibility I have only started to consider. But unfortunately we know almost nothing of 1790, 1823, and 1832.

      • Hector would you consider an article on 1790 Kilauea? For me that’s one of the most fascinating Hawaiian eruptions and I would love to read a deep dive into it with your knowledge base.

        Just a thought! Thank you for this incredible article sequence and as always for your extremely informative comments!

        • Thanks Ryan! I was thinking of writing an article on the explosive eruptions of Kilauea at some point, featuring 1790. There is little known about 1790, it is before historical records started, only a few reports from the Hawaiian people are known, which were written down a few decades later, so it would be hard to look deeply into it.

      • It would be interesting to consider what yiur other comment could mean then for all of this. If magma right now is possibly going into the deep rift, then it could infer an era of these massive intrusions is on the horizon.

        On the subject, this could have been a factor in 2018. That did begin as a normal dike on the ERZ, probably two actually (one from Pu’u O’o, mostly non eruptive, going maybe 10 km east, and the other from the JOKA body). But after the big quake the eruption was very different. Especially as fissure 8 got really active. Some of the lavas from 2018 were highly magnesian, not as much as picrite but they didnt come from Pu’u O’o. The summit also didnt do anything as the initial intrusions happened, but drained out rapidly immediately following the quake.

        Another case could be 1960, given it began in Kilauea Iki with very hot fresh lava and skipped the entire ERZ to erupt right at the far end only a few weeks later, and only a few years after another eruption in the same area. Like 2018 the eruption was intense and erupted more msgnesian lava than the stuff common to the normal ERZ eruptions. There was no south flank activity that year though.

        • I don’t know if what we are seeing now is really something unusual or not.

          After looking back at earthquake and deformation data, I think it is likely that the 2018 intrusion started 2-3 km under Pu’u’o’o as a sill of the Middle East Rift Zone sill complex, then propagated eastward and turned into an inclined dike as it reached Highway 130.

          At a depth of 6 km almost directly under Leilani there is a line of earthquakes that is aligned with the direction of dikes in the area. It resembles the seismogenic base of a dike swarm, similar to other seismic structures of Kilauea that are much shallower 2-3 km. I wonder if that is where Puna Ridge dikes start. It is possibly where the 1924 intrusion became a dike, given how deep that intrusion was from the width of graben opening. It also seems to be the tip of the deep rift. Maybe intrusions from the summit can reach here.

      • There is a very small but definitive spatter vent in the Koae fault zone that was formed during one of these intrusions between 1823 and 1868. So it seems these intrusions were very extensive. It is near impossible to actually locate where this vent is though, it erupted only for a few minutes to make a spatter field around a crack but no lava flows.

        Best guess is it erupted in 1840, given that eruption had many fissures in the upper ERZ.

  10. Interesting chart from Hekla…

    A quakes on the drum plot too….

    • A lot of adverts! Tenuous for northern winter weather effects but definitely something to keep an eye on.

      • I read that article a while back and it at the very least is interesting indeed. It’s observational and based on correlating other instances of strong SH stratospheric cooling with NH stratospheric warming (which splits the polar vortex and can send arctic air into lower latitudes) as well as suggesting it can flip the North Atlantic Oscillation into a negative phase, which is usually one of the things that drives major coastal blizzards in the northeast US.

        I didn’t refer back to the article or see if there are any updates to the thinking here, but I definitely am onboard with the idea that there’s going to be some sort of “chaos” variable this winter with an unknowable impact on the teleconnections / patterns.

        This upcoming week we’re about to witness a record setting -EPO with an absurdly powerful Alaskan ridge and very cold air plunging into the eastern half of the US. AFAIK the strength of this -EPO has not been seen before.

        I’m sure Craig can give further insights here.

        • It would be interesting to know of a proposed mechanism for the SH strat cooling leading directly to NH strat warming.
          Correlation is certainly not causation as the author points out.

  11. What Tongas plume woud look like from low orbit reached almost 60 km high, the huge Hatepe eruption ( 30 km3 per minute ) must have reached the base of the thermosphere perhaps ?

    • Not necessarily. The reason that plumes don’t go higher than 20-30km is that the dust in the column is heavy. An eruption strong enough to go that high would create a dust column so heavy that the convecting air can’t keep it up: it collapses. Hunga Tonga had less dust than normal (it was ‘only’ a VEI5.8) and instead created lots of water vapour. Water is much lighter and therefore was able to rise much higher.

    • Tonga vent up 60 km almost
      But all volcanic plumes do contain alot of water condensation or some of it. A nuclear explosion will form some amazing water condensation as well.

      Whats the air pressure at 58 km altitude?

    • The volcano made a snow – storm in the skies very much, the cloud was full of hails and Ice and graupel
      Something very contrasty when thinking of its fiery origin

      Injection in the Mesosphere.. but most of the water vapour seems to be in the stratosphere hovering at 30 to 45 km elevation

    • Mesosphere lower part is around
      100 times less dense than Mars surface density?

    • The previous google chief Alan Eustace jumped from 41 kilometers in 2014 If I remebers correct beating 2012 s Felix Baumgartners 38 km. Still quite some time until a person jumps from the karman line. It depends how high helium balloons can take a person. During the first minute you dont notice the air because its very thin up there

      Joseph Kittinger jumped from 31 km in 1960

      ”There’s no way you can visualize the speed. There’s nothing you can see to see how fast you’re going. If you’re in a car driving down the road and you close your eyes, you have no idea what your speed is. It’s the same thing if you’re free falling from space. There are no signposts. You know you are going very fast, but you don’t feel it. You don’t have a 614-mph wind blowing on you. I could only hear myself breathing in the helmet”

      But for a bacteria sized magma particle the air density will be much greater for small objects than for large ones for same pressure

      • There was one aryicle I saw that got a maximum plume height of 87 km based on shadows. The numbers otherwise agreed with the general data but gave a higher absolute peak.

        Not sure if this is really true, that number doesnt show up anywhere else, but if there is anything to it at all this is really something, an eruption that almost reached the Karman line…

        It might be impossible to get an eruption that powerful as to actually get into space though. The eruption at Hunga Tonga was a caldera formation that happened in shallow water, so not only was it a near instantaneous evacuation of a magma chamber but also a rapid decompression of all of the water in the rock above the chamber. It is about as explosive as it is possible for a volcanic eruption to be. Making the volume bigger wouldnt have changed this too much I expect. We probably saw an eruption just as powerful as a Yellowstone or Toba scale event, those are just where this lasts for a week instead of a few minutes… 🙂

      • Yes coud be so that yellowstone lasted at that intensity for a week

        Althrough Hatepe was much faster 🙂

        • Hatepe was also ash, not mostly water. Probably there was very little ash in the part of the plume that reached the mesosphere for Hunga Tonga Hunga Ha’apai. A higher eruption rate would not necessarily make the plume higher either, because that is based on air being sucked into the fountain (all plinian eruptions are basically lava fountains) and that really is volume limited. So you need an actual explosion that can ignore the above factor, which in most cases will probably be hydrothermal fluid decompressing. So despite the eruption rate for Hatepe there might not have been an explosion like that although I dont really know for sure. High eruption rate is also only explosive if the vent is small, effusive eruptions can have the same eruption rate as ignimbrite eruptions without turning explosive.

          What a high eruption rate will do is dump a lot of degassing magma at the surface, so it races off sideways under gravity. That is what did happen at Taupo during the Hatepe eruption. But there is quite some evidence these ignimbrites dont move as fast as smaller pyroclastic flows, and way slower than directed blasts. If you have seen the Elephants Toothpaste experiment I think it is like that although obviously way bigger scale. There is an eruption of foam that also continues to degas as it moves, so it is in some ways more like a lava flow with no friction than it is like a small pyroclastic flow seen at a typical eruption that occur almost daily. It is really just a different scale of volcanism.

          • We have a number of temp anomalies in the geologic record which are suspected to have been of volcanic origin but without any plausible candidate. I suspect most Tonga-like eruptions are unknown, unlocated and undated.
            It will be interesting to see if any significant temperature changes happen due to Tonga and its not impossible that the H2O may persist in the atmosphere at altitude for a significant time. Not least its much lighter than O2 and N2 and it may be that removal comes from photodissociation and loss of the hydrogen rather than nucleation and falling to earth.
            Its also implausible that Tonga is the smallest such megaphreatic eruption, indeed there are likely to have been significantly larger ones in the past.

          • Nearly certain that bigger explosions have happened in this style, morre what I am getting at is probably nothing that is significantly bigger, not more than an order of magnitude bigger even though eruptions 3 orders of magnitude have heppened when talking about volume. .

        • Air density is also very low up there
          1/10 000 th of surface pressure?

  12. Dear all!

    Due to the assassination of Bluebird I have terminally left that platform.
    I have moved to greener pastures to munch and galumph with the large herbivores.

    Bye silly bird, hello thoughtful Mastodon!
    If you wish you can add me at

    Edit: Volcanocafé will also integrate with Mastodon shortly, so do not worry, you will not miss out by transitioning.

    • Aaa a reaction because Elon Musk buying up the doomed canary bird?

      • For the time being we will both Toot and Tweet our articles.
        I do not think Twitter will be around in a couple of weeks.

    • Thanks Albert, haven’t seen anything on it yet.

      Any known details yet (anyone)?

    • Gesundheit!

      Is it just me or does that volcano sound like a big sneeze?
      Would be cool if we got another island poking up in the Mariana back-arc.

      I really love new volcanic islands popping up. 🙂

      • It is an interesting area with several active volcanoes nearby that have had historical eruption.
        I have hopes for this one poking up in a not to distant future.

    • Any risk of a repeat of HTHH from this thing? Seems that (and Krakatau) make clear that we, as a civilization, need to treat every undersea caldera within a certain depth range as if it were a ticking nuclear bomb that could go off with essentially no warning.

      There should probably be no habitation of islands within a certain radius of any one of those things, especially any that are its own caldera rim peaks (as with pre-1885 Rakatan and Perboewatan, and more recently the HTHH archipelago). What should that radius be? Well, larger than the distance from Krakatau to the towns on the mainland that got wiped out by tsunamis in 1885 and damaged again a few years ago, and comparable to or larger than the distance from HTHH to the main inhabited Tonga island, where they had quite a rough time of it last year and undoubtedly suffered some casualties.

      Shipping should steer clear of those circles as well. Close enough and it could simply get blown sky-high and for some larger distance there’d be danger from tsunami waves, especially where there are shallow underwater ridges or other topography to amplify the waves above them. There are/were a number of those near HTHH, particularly on the old preexisting caldera’s rim, with no visible sign on the water surface, but they’d likely interact with such an eruption to produce extreme water waves in their vicinities.

      • And yes, this likely means evacuating Santorini and half of Crete. At 3600 years since the last big boom there, it is likely quite capable of another one already, if it doesn’t take a particularly large magmatic eruption to serve as the “fission trigger” and if it doesn’t take tens of millennia to recharge the supply of supercritical water, and if the latter really is the main driver of this type of event.

        One must also wonder whether Campi Flegrei can generate this type of event too, or if it’s too “restless” to build up a supply of supercritical water, which likely requires some centuries of calm stability. Not that Naples is particularly safe from it even if this specific eruption type can be ruled out there …

        • Nea Kameni is pretty small as far as a resurgent volcanic edifice in such a large caldera goes. The pre Minoan Eruption island of Thera was significantly more filled in and “complete” than what we have today.

          That’s not to say there can’t be enough material to make a moderately large boom, but I don’t think it’d be anything like what the Minoans unfortunately saw.

          Still, I totally agree HTHH is a huge wake-up call for volcanologists, geologists, and really anyone with a vested interest in the power of a “natural nuke” residing in an otherwise unremarkable seamount, of which there are many.

        • That is still an average of 325 years between eruptions of this power. Krakatau was only 139 years ago, so there are probably at least twice as many as the above in the relevant time period. If the eruptions are phreatomagmatic with mostly water at high altitude then there would be very little trace later, except for a caldera somewhere that would probably get ignored under assumption it is too small…

          • The 20 were just in the kermadec arc. I would imagine these eruptions can take place elsewhere. What I wonder is how much water do you need to do an eruption like HTHH?

          • Probably better for Carl to answer that. But it is probably not as much as might be expected. It isnt the icean flooding the magma chamber but supercritical water in the rock above the magma chamber or in contact with it. When the caldera forms it is necessarily a loss of oressure. That opens the ring fault abd allows the full evacuation if the magma chamber, leading to an ignimbrite eruption. But in a volcano with a saturated crust that supercritical water will explode at this time too. This probably has nothing to do with the magma composition either. There are probable cases if this in Iceland, possibly as recently as the 1875 Askja eruption, and I would expect this has also happened at Kilauea, there are pyroclastic surge deposits going up to 20 km from the caldera there after all…

            The Big Basalt Blasts series by Hector a while ago is part of my inspiration. A caldera collapse is the best way I can think of to decompress a supercritical aquifer. Ring dikes can allow an enormous eruption rate but I cant see any purely magmatic process being able to generate such a powerful explosion especially if calderas actually usually erupt crystal poor (and thus more fluid) magma the whole viscosity factor sort of falls apart. There was no pressure wave from Pinatubo, despite the similar magnitude.

          • I’ve kind of moved away from the ideas I gave in Big Basalts Blasts though. In the articles “Ring dyke formation on Taal?” and “Rome’s world’s weirdest caldera” I showed my current understanding of how caldera-forming explosive eruptions happen.

            I later realized there is strong evidence that ring dikes were involved in past explosive eruptions of Kilauea, and the 2000 eruption of Miyakejima, and was the main driving mechanism of extraordinarily violent eruptions in otherwise relatively harmless magma. I also side with the idea, which is not new and has been around for some time, that ring dikes drive ignimbrite caldera forming eruptions and that a ring dike may have been responsible for the the intensity of Hunga Tonga Hunga Haapai too.

          • Ring dikes do make a lot more sense, in a general scenario. And it would explain most of HHTH from the volume and the intensity to the 30 km umbrella plume. But the high water content of the plume and the extreme elevation of the high point in the center are sign there was water involved anyway, and the mechanism of action from that ends up being very similar to the BBB idea. There have been a few caldera collapses in historical time but only one other that was so powerful as HHTH and it was also just below sea level. That shock wave in the atmosphere isnt really a thing seen at the eruptions without water.

          • I see Hector mentioned Taals caldera forming process. I recently read a study that basically goes in-depth on all of the Taal caldera forming eruptions. They remeasure many figures. one interesting thing they found is that water didn’t act as a triggering mechanism or played a significant role in the eruptions. I think it may give the ring dyke idea more credit. Study is linked below. It changed how I thought of Taal.

  13. Here’s something you don’t see every day: a very deep M2 quake directly under Hekla.

    14.11.2022 07:18:37 64.010 -19.566 25.6 km 2.0 99.0 5.4 km ENE of Hekla

    • Probably a ghost regardless of the 99% marker.
      Nothing visible on the seismos.

      • I couldn’t tell. It was more than 24 hours ago, so that part of the drumplot doesn’t show anymore. There is however a blip in the tremor plot around that time for FED and HES.

        • I checked before the 24 hour cutoff had passed. I was a tad surprised to not find nothing.

    • 25.6km is not “very deep” when I’ve been locating earthquakes at ~600km depth recently 😀

      • Those are at subduction zones.. where cold brittle oceanic litopshere takes time to warm up.
        The deepest is almost 800 km down if I remebers correct

      • Location, location. location…
        In Iceland this is well below the brittle part of the crust. Any quake at this depth is rare bird and is likely to have a magmatic origin. I have never seen something located at this depth under Hekla before (assuming this is real and not a ghost). Being an Icelandic quake – very deep indeed. Compared with subduction quakes – barely scratching the surface.

        • But quakes also means that something is brittle as well

        • Isn’t there supposed to be some continental crust underlying that area? From memory, so could be wrong.

    • Two more, also deeper than usual, but a lot smaller.

      Tuesday 15.11.2022 22:42:07 63.982 -19.775 11.3 km 0.6 99.0 5.3 km WSW of Hekla
      Monday 14.11.2022 15:15:19 64.011 -19.696 11.5 km 0.5 99.0 2.5 km NNW of Hekla

      And while speaking of Icelandic quakes at this depth, check out this article about deep seismicity in relation to the first Fagradalsfjall eruption:

      • One more. A bit to the east this time.

        Friday 18.11.2022 14:37:41 64.011 -19.415 9.5 km 0.8 99.0 12.5 km E of Hekla

  14. Did anyone see the swarm of 33 earthquakes at the Takawangha volcano in the Aleutian Islands yesterday? They were concentrated 2-3 miles below the eastern flank of the volcano.

    Another volcano I’m not familiar with.

    • Another swarm of earthquakes at Takawangha today, this time with around 49 earthquakes including a M4.2.

      The aviation colour has been riased to yellow due to the earthquakes likely being magma intrusions and migration under the surface.

      • It looks like a complex of several contemporarily active basaltic stratovolcanoes, the plumbing under this group must be so intricate. I do not expect it can erupt too big, although it would be interesting nonetheless if it does erupt. It had at least one other seismic crisis in the past that did not lead to an eruption I think.

  15. The M6.1 near Toba, Japan was a rare quake indeed.
    Depth was 357km suggesting a thrust fault was involved, yet moment tensor shows an almost pure lateral strike slip motion? Not sure if I’ve ever seen a large strike slip so deep…may somebody can offer some other examples?
    The quake’s epicenter appears to be in an area west of the Tonankai segment of the Nankai trough, close to the major M8+ epicenters of 1944 and 1946. These subduction earthquakes ruptured segments A, B (1946) , C and D (1944) of the Nankai trough…leaving only segment E in a fully locked condition. Segment E is where experts say will cause a major/devastating quake along the Tokai segment (E), and the Toba quake was right on the southern tip of the expected rupture zone.
    So far, also interesting to note there is zero aftershocks being reported…for an M6.1 strike slip motion one would expect some aftershocks (deep thrust faults though often produce zero aftershocks)…and may be related to the odd focal mechanism?
    Only time will tell if this latest minor shock is part of a longer term reactivation of the long-overdue E segment…which I think will be punctuated by some other precursor quakes, such as what happened prior to the great Tohoku M9.1 earthquake which saw several large foreshocks in the time period right before the “big event”.

    • Most powerful machine we have ever created and used. The Saturn V lost that title today, after 50 years on the throne.

      Question is, how long will SLS keep it… Starship is expected to launch in December, and even factoring Elon Time and his usual optimistic timelines if the rocket is actually flight ready it wont be more than a couple months out. Starship Superheavy is supposed to be twice as powerful as SLS, and did a full power static fire yesterday.

      I can imagine some time in the future when Starships are abundant that one gets launched with a triple core booster setup, like a Falcon Heavy but with Starship boosters 🙂

    • How hot is the rocket engine?
      As hot as the surface of the sun?

      • LOX/hydrocarbon fuel rockets are between 3500 and 3900 C. In practice though engines in real life are not close to this, around 3000 C to prevent destroying themselves. Raptor might be the hottest because the propellant is a hot gas already by the time it gets into the combustion chamber, but just my guess.

        LOX/H2 rockets are up to 3300 C in operation for the Space Shuttle engines, presumably so for Artemis 1. But H2 has lower energy density per mass so the heating is not as intense.

        I cant find any data on the temperature of hypergolic rockets but presumably they are extremely hot, hydrazine and N2O4 are liquids at room temperature and also can both decompose on their own to release energy. Test videos of these engines usually show the flame diverter heated to glowing within a second or two, these rockets are more for power and reliability than efficiency.

        Solid rockets apparently are not quite as hot but have much more energy and thrust as the exhaust is in large part Al2O3, which is heavy compared to H2O or CO2

        Generally though numbers for all rockets are about 3000 C.


    ”Earthquakes on Mars – or marsquakes – with epicentres originating around a region called the Cerberus Fossae, consisting of surface fractures, have been studied by an international team led by seismologists and geophysicists in Switzerland.

    Their analysis of deep, low frequency marsquakes suggests a structurally weak, potentially warm region, around 30-50 kilometres beneath the surface. This, they suggest, is consistent with recent magmatic activity at these depths. ”

    Looks like there is active magmatic systems under Cerebus Fossae. Makes sense that Mars should still be alive althrough not as much as Earth. Mars is between Earth and the Moon in mass and the cooling rate should be too. Not as active as the Earth .. But not as dead as our moon.

  17. Battle between giants. Which will go first?

    I guess, technically, the battle has already been won. But you dont get a deformation signal like that from a volcano that isnt preparing something… A signal like this cant help but fuel speculation the next eruption will be another dual event.

    • Right magma is filling Kilaueas storage when the lake is not rising as much as it did in 2020

      Will either be a rift eruption or summit eruption will get alot more vigorous soon

      Mauna Loa looks exciting as well

    • It seems a little odd that Kilauea is inflating as fast as in times of no eruption. This past month the lava lake hasn’t experienced any rise either. The Middle East Rift Zone has also stopped supplying the summit eruption for the past ~2 months. Maybe the conduit is clogged. Lava might still be convecting in the lake but possibly no new magma is coming in, that seems like the most reasonable option to me right now. It would be very difficult for Kilauea to inflate so fast if it had an open conduit into a lake that is so large in area as the one at the summit right now. But one never knows for sure.

      If that’s the case then surface volcanic activity will probably slowly die down while the summit keeps inflating. At some point we will start to see earthquakes along the East Rift Zone Conduit indicating pressurization, and then another dike or sill intrusion will happen at the summit-SWRZ-Kilauea Iki area, which may or may not restart intense eruptive activity. I don’t see signs of over-pressure at Kilauea right now, so we are probably multiple weeks to multiple months away from next intrusion. There hasn’t been an intrusion of Kilauea that has not been preceded East Rift Zone Conduit seismicity, with the only exception being the rootless lava lake-fed sill of past September.

      • The lake still responds to DI events though, which it probably wouldnt if the connection was blocked. Maybe it took until 1000 m elevation to erupt before but after 2018 that number might not hold true anymore. The caldera was also filled with solid rock until then, which can support itself a bit, where now the weight comes from 300 meters of heavy liquid. Maybe there is a limit to the depth of the lake that is determined by the density of the lava and how much force it takes to erupt fresh lava into it. If the conduit is wider then it will be harder to stop the dense lava draining and to erupt fresh lava. The fact there was a sill intruded recently probably means this limit is very near, and now instead of filling the lake it has capped itself. So magma is flowing in but cant push up the lake anymore.

        The lake in the 19th century was a lot bigger, but was maybe not any deeper than todays lake. It isnt really actually clear how deep the 1790 caldera was, it is assumed as deep as the water table but if the eruption that year was a magmatic eruption from a ring fault then that is not a reliable indicator anymore. It is also not really known how deep the water table actually is either, the water lake was still rising when it was destroyed, it may have gotten twice as deep for all we know, the 2018 collapse seems to have been very deep for how big it was.
        The 1790 collapse could have also been much less vertical height over a wider area, the caldera could have collapsed by only 200 meters to give more than 2 km3 volume collapse, and the lava lake would therefore be a lot shallower than todays lake despite the much larger volume and area it covered. And if it is the depth of the lake that matters then todays lake would be putting a lot more pressure on the system than the one in the 19th century. The bottom of the lake is like being 1.5 km underwater.

        It is unfortunate this critical observation window was in a gap of the historcal record 🙁

      • Maybe a better way to put it, if the conduit isnt obstructed (unknown) then the lake now could be a bit like a geyser, and the pressure is building because there is a cap of dense degassed lake lava above it. So instead of an intrusion another possible outcome is where alarge volume of lake lava drains down and us replaced by fresh magma. In a small lake or an actual water geyser this would become a tall fountain (Fagradalsfjall) but here I suspect it will be more subdued given the depth the known vent sits at, and plausible presence of other deeper vents too.

        In terms of a signal it might look like a massive DI event, so a sudden drop in the whole lake folliwed by strong eruption out of all the cracks, possibly with fountaining up to 100 m in the beginning like the eruption began last September. It is possible that the D part could induce an intrusion somewhere anyway though, it could get complicated…

      • Yes, that is true. The drop in the lava lake during the past 2-3 days clearly matches the shape of the last DI event. So the conduit is not clogged. And yet the eruption doesn’t seem to be sending magma to the surface.

        I have checked the GPS data though, and I think something deeper is at play. Every station from the summit of Kilauea down to the south coast of the island is being affected by a rapid extension centred on Halema’uma’u that has been gradually accelerating. CRIM, OUTL, GOPM, APNT, AHUP, DEVP, MANE and KAEP, all these stations which are found in the area from the south rim of the caldera to the coast show a rapidly accelerating south displacement. This southward displacement is very little at the coast but increases towards the summit of Kilauea peaking an CRIM, on the south rim of the caldera. Meanwhile UWEV in the north rim of the caldera is accelerating northward. I think that magma is possibly flowing into the deep rift/slump structure deep under Halema’uma’u, perhaps to refill draining from the 2018 eruption or fill space created by the 2018 earthquake, and creates this deep spreading. This has possibly diverted supply away from the eruption. Also the Middle East Rift Zone seems to be deflating faster than ever. The signal only covers several days, but it does look like JCUZ and NPOC are very rapidly moving down and southeast, towards the centre of the Middle East Rift Zone storage. So this flow of magma into the deep storage of Kilauea may have both accelerated deflation of the Middle East Rift Zone, and stopped the lava lake rising, or even reversed it to draining.

        • Looks like the future will see another episode of voluminous eruptions on the ERZ again then. This same activity after 1975 is what probably allowed Pu’u O’o to form. It looks like the ERZ is active down further east of Pu’u O’o too, as far as the highway based on where the south flank quakes go to. So potentially magma could flow this far east. It is very fuzzy but the JOKA station could be showing some uplift, as well as more definitive southward movement. There is also a bit of eastward movement. So it looks like magma could be moving down the rift but not as a real intrusion. It might be possible these distal magma bodies get some feed from these deep magma flows though, not just a direct connection to the summit. The 2018 dike began after years of slow magma supply to the JOKA magma body, but there was no open connection the area to the summit until the big quake on the 5th, which was after the dike already started. Not that an event like this is likely now ir near term but it could be more complicated to predict eruptiins here than it is for the upper rifts and summit.

          This is actually very similar to the first deformation signals after 2018 eruption ended, there was immediate inflation at JOKA, but still deflation elsewhere for some time after. I guess the south flank has moved enough to give a similar effect. It would be interesting to know if the southwest end of this structure merges directly into the Hilea area once it goes past the Great Crack. It might well be a trigger for a quake there, and set off Mauna Loa. In 1868 Mauna Loa was in control, doing the pushing, maybe now it could be the same but with a roll reversal, Kilauea not so much pushing but removibg the support, letting Mauna Loa slip. Perhaps the 1983 Kaoiki quake was induced by the deflation of Kilaueas summit from the eruptions and rifting at Pu’u O’o, or perhaps it was a coincidence.
          One could spend a lifetime trying to work out how the south flank of the Big Island works and what it means for volcanism, it is probably the key part to understanding the interaction of Mauna Loa and Kilauea properly, maybe even more so than the magma supply rates.

        • There are still a lot of quakes at Kilauea, now centered mostly at Kilauea Iki this week, and also with some that are deeper down than before. Not all of the quakes are shown on the live map but the bigger yellow one within the caldera is 13 km deep.

      • Probably would have looked the same except the scale of the deformation sugnal reversed so Mauna Loa is more prominent.

        Hopefully HVO will release another map of Mauna Loa, of the summit and the upper half of the rift zones. So far only distal ends of the rifts have been released and eruptions there are rare and dont give much of the full picture. The last map was in 2020.

      • That small lava pond that was in the center of the lake has now collapsed into a big pond, maybe even big enough to count as a second lava lake. I guess there has probably always been a thin area there, it is about where the first lava of this eruption appeared in September past year, and there has been a persistent and often incandescent degassing area here basically ever since then.

  18. I can see the New lava pond appearing .. right of the main lava lake

    • If we are lucky the whole ponded crusted lake will become an active lava lake surface, like it was back to early 1900 s when there was episodes of huge lava lakes

      • Thec massive lake seen by Ellis in 1823 was probably because the eruption had just resumed filling, same as in the first few weeks of this eruption and back in 2020. The 1823 fissure at Keaiwa was still very new, smoking in places, and so ws another undetermined vent on the SWRZ, most likely the Kamakaia Hills or Black Cone, which erupted between 1815 and 1823. The lakes in 1840 and 1868 also surged, but before the lake was drained. 1868 might have been pressure from Mauna Loa but in 1840 it was maybe just on its own. It could well have been a surge of magma to Kilauea in 1868 too though, the short pulses of magma do seem to go to both volcanoes even if one is dominant long term.

        Probably the lake wont become active over its whole surface unless it either drains and breaks up, after which it is gone anyway. Or if it stops erupting completely and restarts, to erupt a few months of lava in a couple days like happened in September. I think that the crust will generally survive small changes, at least as long as it has the tephra island in it to hold it up. But it is also possible it slowly gets consumed from the bottom up and collapses into that pit we saw form yesterday, maybe in a few months it will be largely new surface. This is not really the same as the whole lake being active though, the long term supply is probably not high enough to keep an over 1 km wide lake continuously circulating, unless maybe the conduit got that wide but that seems very unlikely before it would all drain out somewhere.

  19. This might also be of interest to you Hector. What might possibly be the largest recent eruption of mauna Loa that I can identify from the available maps, that of Pu’u O Keokeo north of Ocean view. The summit overflows of Kilauea and Mauna Loa might be bigger, but those are not single eruptions.

    It appears this was a closely spaced line of at least 3 massive eruptions, lasting probably several years to maybe a few decades, unlike a typical shield eruption the lava is dominantly a’a with massive flows although a lot of large lava tubes and pahoehoe are found in the eastern flow field. There was possibly two different rounds of shield building and paroxysmal activity, so the activity was rather more complicated than what we have seen from Pu’u O’o where the eruption became more passive with time. The Pu’u O Keokeo flows are C14 dated to about 1370 years old, so likely erupted in the first half of the 2nd century AD. The slightly older Papa and Miloli’i a’a flows are not dated, but on the newest map appear to be interlayered with the Keokeo flows, and are impossible to distinguish visually, so I tentatively included them in this map as well as uncertain vent locations. Both these two eruptions and the first stage of the Pu’u O Keokeo eruption were fast, and flooded a massive area of the southwest flank of the volcano. The volume of all of these fast eruptions together is likely over 1 km3, and was probably emplaced in a few weeks by analogy to observed eruptiosn in the area. The effusive stage seems to have begun immediately from this and been relatively minor in comparison, but pahoehoe still covered about 50 km2, which is an area that would have taken years to cover at the same rates seen from Pu’u O’o. A second surge formed a massive a’a field going directly south, as well as a new cone and lava lake.
    Recently many massive lava tubes have been found in this flow, some with erosion of the lava below, showing the eruption was also long lived in this stage. Strangely the tube seems to have formed directly in the a’a flow field, as pahoehoe doesnt really dominate until down on the coastal plain. The total volume of the Papa/Miloli’i/Keokeo eruptions is likely at least comparable in volume to that erupted by Pu’u O’o, and may be rather more as the area is much larger and no depth of the flow field is known, plus the caveat the large volume of a’a that may have advanced offshore as is often typical of such flows – and a thing that was not a factor for Pu’u O’o.
    The eruption sequence also ended with a caldera formation of Mauna Loa, the nearby and offset vent of Pu’u Ohohia fed a massive a’a flow down to the southwest to form the coastline between Kapu’a and Manuka bays, followed by pahoehoe flows burying much of the north side of the flow field in turn. This flow has a very high concentration of olivine like that of the other massive caldera formation eruptions, and of 1868, and quite unlike the earlier flows above. The volume of the Ohohia flows is by itself probably between 1 and 3 km3, by analogy to the nearby Hapaimanu flows, and the 2018 flows. The flow formed a large lava delta, indicative of a substantial submarine volume.

    This eruption might represent the most recent time a conduit actually formed within the SWRZ of Mauna Loa. There have been other large eruptions in the area since but all were all fast eruptions. Large long lived eruptions like this at Mauna Loa have all been at a much higher elevation, and most of them have been shallow dikes, like Mauna Iki or Kupaianaha, rather than proper conduit satellite volcanoes like Pu’u O’o. It is also remarkable for the large area the a’a stage covered, even if you only go by the flows identified as 2i6 on the newest map, and exclude the other flows, the area covered is still larger than the later pahoehoe stage. It might represent the biggest SWRZ eruption of Mauna Loa in the past few millennia.

    • Yes, those eruptions are interesting. Long-lived, large-volume, high-effusion rate SWRZ eruptions of Mauna Loa is a kind of eruption I don’t think we have seen historically in Hawaii.

      The eruption that seems the most recent and most dramatic example of this kind of activity is the Kipahoehoe eruption, which happened around 1550. Kipahoehoe is massive, its two giant lava deltas reach more than 2 km underwater, as seen in bathymetry data, and although I haven’t measured them yet, I wouldn’t be surprised if this eruption had a comparable or superior volume to Hapaimamu, which was 2.6 km3.

      In the Kipahoehoe eruption lava was supplied from a vent buried under 1950 lava, lava flowed continuously through a lava tube and broke out from rootless shields near the coast of the island. The flow is dominated by aa type lava, typical of high eruption rates. As such it seems to me that the Kipahoehoe eruption lasted for a long time (months?), it was not episodic but continuous, as shown by the sustained flow into the lava tube, and eruption rates were very high to make mostly aa lava.

      Kipahoehoe was probably closely spaced with two similar eruptions, although not as large. The Kolo eruption, and the Kalahiki eruption. The Kalahiki flow is very impressive too, but it cannot be followed adequately once it enters the forested area. It also reaches the coast and makes a mid-sized underwater delta, made of solid lava that flowed underwater. These eruptions seem to have taken place under a particular evolving stress field which caused fissures to open on the west side of the rift with a gradual migration towards the axis of the rift, with eruptions happening at mid-elevations in the rift zone. That is why I think they were very closely spaced.

      The Keapohina episode of activity in 1180 is also of this kind, two or three eruptions happened from almost the same spot in the rift, although migrating westwards. The first two Keapohina eruptions were both very large and long-lived, although they are impossible to follow in the forested flank of the volcano, and reached the coast in multiple places, without making large deltas. They had lower eruption rates and a more significant proportion of pahoehoe compared to Kipahoehoe, but still do have a lot of aa.

      There were earlier episodes, like the one you mention, although I haven’t looked too much into them. Hualalai and Mauna Kea often seem to erupt this way, building shields of aa lava. And the Pimoe eruption of Haleakala might also be similar. They seem to be long but sustain rates that are above the long term magma production of Hawaii. I find them somewhat enigmatic.

      • Thank you and Chad so much for this discussion, I was asking about the larger Hawaiian events from the not too recent past and for info on what we do know or can surmise, and you guys have both delved into exactly what my curiosity was looking for!

        Awesome stuff!

        • There is not likely to be a resolve to this discussion, there is still a lot we dont know about volcanism in Hawaii.

          I find Hawaii to be potentially much more interesting than a lot of other places. Iceland has a lot more varied volcanism, but to me there is just something that really sets Hawaii apart. Hawaii is unique on Earth today in that it is a very powerful plume that feeds only 2 volcanoes. Galapagos has 5, and Iceland has over 20. The volcanoes of Iceland are also huge, but they are huge because they grow in a rift zone and have a very long reach.

          But Hawaii just really is on a whole different level. The volcanoes are so big they create their own rift zones…

          It is also just a general fact of life that Hawaii will see a hell of a lot of lava erupt. There is 0.2 km3 of it every decade and half of it ends up on the surface. If Kilauea is active then a century can go by where you count the years it didnt erupt on one hand. If Mauna Loa is active, Kilauea still erupts all the time anyway, but there are massive lava floods every other year too… Iceland just isnt like this. Volcanism there is fascinating in other ways but to me it just isnt the same. And there is a lot less mystery, we know pretty well how divergent plates work. But a volcano that is so big it pretends to be its own divergent boundary anyway, and is literally 1000x as big as a normal volcano, I dont think we have any idea of the full potential.


          • Thats the reason Im too are fascinated by Hawaii as well

    • In fact most prehistoric SWRZ eruptions of Mauna Loa were clearly long-lived. And 1950 is the only eruption of its kind visible. The only eruption that produced fluid, thin (less than a meter thick at the edges), extensive aa lava. And in general lava flows erupted since 1800 (including Manuka and Pele Iki) have characteristics of having been erupted at high rates and short duration, with open channels of lava, and no tubes. While almost every single prehistoric eruption in the SWRZ of Mauna Loa, and all the big prehistoric ones as far as I know, are tube-fed lava flows.

      • Maybe the mechanism of a dike intruding into the deep rift and erupting at a very distal location for 1868 is to blame. There are still some other old fast fissures, so manuka and Pele Iki might represent a bit unusual but not that weird eruption.

        It might also be that the caldera formation before 1710 was from Pana’ewa’ which was far down the NERZ, and so the SWRZ magma system was left intact, allowing transport of large volumes of magma. These high rate long lived eruptions that create massive tubes might be fed by a deeper magma body than the shallow ones that collapsed to make Mokuaweoweo, a magma body that is deep enough it cant collapse all the way to the surface. So these a’a shields may be a bit of a variant on a caldera collapse. They would also share some similarity with the eruption on La Palma last year, which was also sustained at a very high effusion rate (for a shield tper eruption) for a long time, at least 30 m3/s which is much higher than the pahoehoe transition. The LaPalma eruption began very deep, 40 km, at Mauna Loa I imagine still under 10 km, but deeper than where the Mokuaweoweo magma chamber is.

        The bigger hills on the Puna Ridge might be the same as this too, direct deep rift eruptions. Kilauea might not be tall enough yet to erupt this way above sea level, as a comparison the Keapohina and Kalahiki vents are still more than 1 km below Mauna Loas summit. But Kilauea seems to have a generally more voluminous and extensive storage than Mauna Loa, and it is a lot more leaky, so not enough time to get the volume saved for these a’a shields.

      • Looking at this a bit further, these eruptions seem to be the bridge between the caldera formation eruptions and lava shields. Lava shields like Pu’u O’o erupt at the base rate of supply from the source, which is in Hawaii between 1 and 5 m3/s. At this rate really the eruption can last forever, the end likely being due to a longer term tectonic movement like seen in 2018, Pu’u O’o had previously survived many other intrusions and rifting events.
        Caldera formation eruptions are hundreds if not thousands of m3/s eruption rate. 2018 had a DRE eruption rate of 250 m3/s, and had a peak of over 500 m3/s. In bulk terms it was well over 1000 m3/s. Hapaimanu looks like it was very similar to this in a lot of ways, the channel is about as big as the fissure 8 channel, although it began with a much larger curtain of fire stage as evident from the story and the preserved fissures above the main vent.

        The a’a shields, like Kipahoehoe, or Pu’u O Keokeo, form lava tubes but also mostly a’a. Eruption rates here are probably in the range of 10-100 m3/s, so not as fast as a caldera formation but well above the base supply rate. As I proposed above, these eruptions might drain deeper magma bodies in the volcano, too deep to collapse into a caldera and set off the piston collapses that cause the huge eruption rates.
        There are some examples of this sort of activity in Hawaii historically though. The 2007 Episode 58 TEB vent on the east side of Pu’u Oo was able to sustain eruption rates that were much higher than the long term average for almost 6 months, somewhere between 11 and 15 m3/s compared to the typical ~4 m3/s, making a massive system of lava ponds along a lava channel, and for many months feeding sequential a’a flows to the east, and then again to the south forming massive tubes and rootless shields when the flow diverted, a’a flows going as far as the coastal plain. When the flow tubed over to the coastal plain pahoehoe flows crossed the plain in only a few days, not the weeks to months that were typical. This was associated with strong deflation of the summit, following a similar episode of strong inflation before. The Overlook vent formed in Halemaumau as the ERZ slowed back to normal rates. This would be at the low end of the a’a shield range I guess, but nonetheless a historical example from Hawaii. The 1955 eruption was also like this, at least in the final stage 3, where two moderate fountains made a big lava lake that overflowed a’a flows over a wide area up to several km away. 1955 had one summit collapse quake but lacked the huge eruption rate and olivine rich magma of 1960 and 2018.

        A much better example is probably the recent eruption from Fagradalsfjall although that was volumetrically very small, it was still erupting well over 10 m3/s for most of the eruption and as such only really made a’a, unlike the 2021 eruption that was much more variable and mostly pahoehoe. Lanzarote is also like this, 3 km3 of lava in 6 years is an average of 0.5 km3 a year, about 3x the eruption rate average of Hawaii or about the same, maybe a bit more, as what Pu’u O’o did in 2007-2008. And of course the eruption last year on La Palma. And now the comment from Zach below has me thinking the Matavanu eruption on Sava’i in the early 20th century was also like this too.

        I guess these eruptions really need a name, just calling them a’a shields isnt really very descriptive, as in action they would be rather different from something like Pu’u O’o. And you could also get a shield shaped effusive basaltic volcano made of a’a that was made by repeated very fast lava flood eruptions from the same spot, like early Pu’u O’o or the enormous lava field of Alayta in Ethiopia, which is an entirely different thing to what we are talking about.

        In all these cases except for Kilauea the key thing is a magma body of some sort within the crust, but too deep to collapse into a caldera which limits the eruption rate to how fast the volcano can contract, which is evidently a lot less than can happen if a free moving block is what is compressing the magma.

  20. Quite an article there, but there is a bit of an interesting eruption on Savai’i, where there was a bit of a long eruption from 1905-1911, where it looked like, according to Google maps, looks like a shield that formed there and destroyed coral reefs and a few villages.

    What is quite interesting is that, before this, there was another eruption less than 3 years ago, of which it had woken up after more than 240 years of dormancy (according to It would be quite interesting for someone to write an article on it.

    • It is quite interesting that (based on Wikipedia, which is not much of a good source but it is what is found) it was explosive with lava fountains (?) at first and there was a 100-foot high lava flow that travelled 2.5 miles. Over that course, it had filled a valley in 1500 feet of lava and destroyed a few villages and, by September of 1906, it reached the sea but died down and that is when it formed a lava lake there (a true lava lake or a pseudo-lava lake?). It was only shortly after that on October an explosion occurred (actual explosion (phreatomagmatic?) or a lava fountain?) and sent out more lava and was really active until its decline began in 1908 and died down in 1911.

      What I find interesting is that it sounded like Pu’u O’o but quicker and also probably had a combo of effusive and explosive activity. It quite ambiguous as there were few photos of the even other than a crater in vog, a lava field and the ocean entries and that is it. Would be quite incredible to see the lava lake there and it is quite odd that it had a maybe approx. a century’s worth of dormancy until it erupted again but yet there was a smaller eruption, like I said, in 1902. Maybe that was a “fore-eruption” before the big one as it was accumulating in a small pocket of magma before erupting? Like I said, this is quite interesting.

  21. Hawaii is the most powerful hotspot on the Planet at current, it dwarf anything and Dwarfs Piton

    Can anyone here find a paper from USGS comparing Hawaiis supply to the much smaller supply under Reuinion?

    • Kilauea over over 0,1 km3 per year and Reuinion sit at little over 0,01 per year

      Any data om this to prove my point in a FB debate

    • Still Reuinion have a gigantic supply compared to example st helens .. But Hawaii is far far larger supply still

    • Albert perhaps any good sources on supply rates for both ?

      Trying To winn a FB debate : ) with proving my point that Hawaii have the largest supply

Comments are closed.