The piggyback volcano

The expression became famous because of Isaac Newton. Never one to ignore praise, in typical English fashion he expressed his self-assured superiority in a self-effacing way: “If I have seen further it is by standing on the shoulders of Giants“. This was written in a letter to Robert Hooke, himself both an intellectual giant and a person with power who could be morose, jealous and vengeful. Newton’s confession should perhaps be seen as a way to keep favour with someone who could (and would) become a powerful enemy. And of course, it doesn’t really reduce the dwarf. The point is that the dwarf does see further than anyone else. Rather than putting oneself down, it puts the writer on a pedestal.

But the expression was not original: Newton quoted older sources. The oldest written source containing it is from 1159, the Metalogicon (meaning ‘on behalf of logic’) by John of Salisbury (names were so easy in those days), where he writes “Bernard of Chartres (another easy name. Imagine being called ‘Bob from New York’) used to compare us to dwarfs perched on the shoulders of giants. He pointed out that we see more and farther than our predecessors, not because we have keener vision or greater height, but because we are lifted up and borne aloft on their gigantic stature.”

But make sure you stand on the shoulder of the giant, and not on its knee. Otherwise you still remain the lower of the two. This mistake is a very common one in real life. To benefit from prior knowledge, you need to learn all of it, not ‘just enough’ to get by.

Kilauea also made this mistake. In comparison to other prime Hawai’ian volcanoes, it is not large. It got a head start by growing on top of the lava of Mauna Loa, the giant next door. But it started out too low down the Mauna Loa slope to become the dominant seer.

In this blog, Mauna Loa is often portrayed as being in decline and no longer important. Here I will argue that far from this, it is still the mover and shaker of Hawai’i. Even of the dwarf on its knee, noisy Kilauea.

Mauna Loa

Mauna Loa is the world’s second tallest volcano on Earth, after Mauna Kea. It started 5 km below the sea, and took around a million years to reach its current height of 4.17 km above sea level. Although it is slightly lower, Mauna Loa is a lot bigger than Mauna Kea. Mauna Loa is a shield volcano: its fast flowing lava has spread far, building a large mountain with a shallow slope. It is much wider at its base than Mauna Kea, has much more volume and therefore much more mass. This is immediately obvious from satellite images

Mauna Loa is a very elongated mountain. It is a lot longer in the southwest-northeast direction. In fact, the name Mauna Loa means ‘long mountain’. The long ridge follows the rift zone. There is a caldera at the top (which shows the mountain used to be a little higher before the collapse); the rift extents from the caldera towards the south west and towards the east, running from the south eastern point of Hawaii to (almost) Hilo, a distance of almost 100 km. The southwest rift bends by 40 degrees some 30 km from the summit, at an altitude of 2400 m. At the location of this bend a number of eruptions have build a satellite shield. The northeast rift also bends, to the east, but this bend happens much closer to the summit.

Landsat mosaic image. Source: USGS

On the satellite image (which predates Leilani), the extent of the lava flows give a good indication of which volcano rules which area. The edge of Kilauea’s dominion is clearly visible. The lava flows from Mauna Loa reach very close to the summit of Kilauea. Kilauea owns the land to the south and east from its summit, but very little to the northwest.


Let’s look at those dominions. On the map, the two circles show the area of dominance of Mauna Loa. Of course it is not exactly a circle, but it gives an idea. The inner circle extends to the lava field of the neighbouring volcanoes, a distance of 27 km. The outer circle is the distance to Kilauea’s conduit. (The distance from Mauna Loa to Kilauea is 34 km, according to giggle maps – which also claims you can walk it in 14 hours, something I have doubts about.) Distances between neighbouring volcanoes on the island of Hawai’i are quite similar:

Mauna Loa to Mauna Kea: 41 km
Mauna Loa to Hualalai: 37 km
Hulalai to Mauna Kea: 44 km
Mauna Kea to Kohala: 40 km
Hualalai to Kohala: 46 km
Mauna Loa to Kilauea: 34 km

The separation is about the same as the depth of deepest magma, which is not accidental. A new volcano can only form at a certain distance from an existing one. Magma will always prefer to use an existing conduit within reach, rather than spending effort in forming a new one. But if the distance to the conduit becomes larger than the depth, then it becomes easier to form a new conduit. Both Mauna Loa and Kilauea are fed by 30 to 40-km deep conduits. (There are of course many intermediate magma reservoirs on the way, where the magma can age before use.) Pahala’s earthquake activity also has this depth of 30 to 40 km and it is at this same distance from both Mauna Loa and Kilauea. This may be the reason why the magma is congregating underneath Pahala, unsure what to do next.

Kilauea is a little closer to the Mauna Loa than would have been predicted from the other volcanoes. This might be why it hasn’t grown as large yet: there is competition for its magma supply.


All Hawai’ian volcanoes have rift zones: it is part of their culture. These rift zones are indicated by the dashed lines on the map below. (The rift zones extend further than shown on the map, and can reach as fas as the deep ocean.) Those rifts deserve a closer look.

A curious aspect is that the rift zones tend to run parallel to the nearest neighbour volcano.

Consider Kohala, the oldest volcano on the island. It’s rift zone runs parallel to Hualalai, the next oldest, but not to Mauna Kea which is a bit younger. Mauna Kea has strongly bend rift zones: it was clearly affected by its older neighbours. On the eastern side the Mauna Kea rift seems to bend and merge with the rift of Kohala. Off-shore it becomes the Hilo ridge which seems to have been owned both by Kohala and Mauna Kea. Hulalai’s rift runs parallel to Kohala and Mauna Kea, but is unaffected by Mauna Loa which did not yet exist in Hualalai’s glory days.

The southwestern rift zone of Mauna Loa runs parallel to Hualalai, and the northeastern one parallel to Mauna Kea. In the far southwest, the 40 degree bends brings the rift into line with that of Hualalai, similar to the case of Mauna Kea at the other side of the island. The bend in the northeast rift is due to Mauna Kea. Clearly, the rift zones show the interaction where younger volcanoes had the follow the guidance of the older ones. They are perpendicular to the flank of the older, larger (at the time) volcano, and in fact follow the same direction as the pali of the older volcano would. A rift perhaps begins as a slump in the other volcano. Once the rift is established, it will remain in place while its owner grows, although it can (and does) grow in extent, away from the mountain.

Kilauea’s rift zone runs parallel to Mauna Loa and follows the elongated shape of Mauna Loa. It is clearly very strongly affected by Mauna Loa.

An interesting effect is that Kilauea’s rift zone of offset from its summit by about 7 km, while for Mauna Loa the rift zone goes exactly through the summit. All other Hawai’ian volcanoes also show perfect alignment of the rift zone with the summit. Kilauea is the odd one out.

The rift zones are the locations of most of the earthquakes. Rifts are a weak zone where magma can force its way in; the earthquakes can happen during these intrusions or during tectonic adjustments. Magma intrusions typically happen as a dike, 1-3 km in height, 10 km or more in length, but as little as 1 meter in width. The magma forces the rift apart. A dike can slowly melt its roof. Sometimes, when a dike runs close to the surface, on one or more places along the dike the roof can collapse: this forms one or more pit craters. Kilauea’s pre-2018 lava lake first formed in such a collapse crater where the magma below melted an opening, providing a window into the flow. That magma had no intention to erupt there: it was on the way to Pu’u’O’O’o, and the pit lava lake was just a distraction on the road. Summit calderas can become enlarged by such pit craters: eventually they can become elongated as a result. Halemaumau itself started as such an off-centre pit crater, perhaps around 1825.


In addition to the rift zones, there are fault zones. These are distinct from the rifts. As mountains grow and summit rifts expand, the slopes of the volcano are pushed outward. This movement occurs on faults. It will not be a surprise that these faults tend to run parallel to the rift zones. In the case of Kilauea, the main faults run south of the rift zone but not on the north. This is because the massive bulk of Mauna Loa blocks any expansion to the north.

There are two obvious fault zones south of Kilauea. One is the Hilina fault, which follows the steep escarpment at the edge of the island. The second is the Koa’e fault which is visible on the surface south of the caldera, and which joins the east and west rift zones. The Koa’e fault can erupt lava, but the dominant signature is ground cracks, grabens and low cliffs (scarps). Just south of the Koa’e fault, the land is moving towards the sea at 8 cm per year. There are frequent earthquakes here, which in 1965 uplifted part of the area by more than 2 meters. The whole region may be unstable, and in a distant future it is conceivable that the entire block south of the Koa’e fault zone will slump into the ocean.

Measuring the drop along the Koa’e fault zone caused by the 2018 earthquakes. The people doing these measurements are known as the crack team. Photo by photo by S. Lundblad. Source: USGS

April 2, 2012 – March 19, 2014 deformation of Kīlauea’s summit area and upper/middle East Rift Zone. Areas of subsidence are caused by cooling of old lava. The Koa’e fault system shows the effect of two M 3.5 earthquakes in 2012.

The Hilina fault also is very visible on the surface, with a cliff several hundred meters tall. It traces a large slump, where material moved towards the ocean and left the cliff behind. Perhaps one day the Koa’e fault will look like this. But the real Hilina fault lies several km underground. How deep it goes is not really known. The lava pile here is 9 km deep, and underneath this is the old ocean floor. The contact plane (decollement) between the lava and ocean floor is nearly horizontal: it is tilted by 5 degrees. (The downward direction is north, where the ocean floor is depressed by the weight of the mountains.)

The best description is obtained of course from USGS. This drawing is from them, with as only VC addition a rough indication of the Hilina fault. (VC is a dwarf standing on the shoulders of USGS!)

As indicated in the drawing, the dike pushed out the land in the upper few kilometers. But the contact plane with the ocean floor (‘Pacific plate’ in the picture) is locked in place. The top moves, but the bottom doesn’t. (After 10 months of lock-down, I feel the same way!) This difference in movement causes shear between the top and the bottom. The pull on the bottom grows every year, until finally it gives. Now the bottom moves seaward (5 degrees uphill!) by potentially several meters. At the point where it begins, the top will sink, forming the cliff (or pali). This is the cause of many of the largest earthquakes in the area, including M7+ events such as in 2018. Here the stress comes from the expanding rift, but the inflating caldera can do the same thing. And even without inflation, just the height of the summit puts an outward force on the slopes and try to make it move.

In practice, the surface is connected to the deeper regions by a number of steep faults, which accommodate the shear. One of these is the Hilina fault. It appears that the Hilina fault does not connect all the way to the bottom but remains fairly shallow. Some earthquakes cause it to move, with potential tsunami consequences, but not all do. The 2018 event did not.

When push comes to shove

Now let’s look again at Mauna Loa versus Kilauea. What does a giant do when a dwarf climbs on its shoulder?

Mauna Loa is of course a far older volcano. It started out a million years ago, first on the sea bed and afterwards on the land it itself created. The growth has not been uninterrupted. On the west side, about 100,000 years ago a large block broke off and collapsed into the sea, with debris flowing out as far as 80 km. The scar and its debris are still visible. And while the Hilina slump happened slowly (mostly), this one happened in one collapse.

The south east side of Mauna Loa has suffered large, slow slumps similar to Hilina, leaving scars or palis. Some still exist (the Ninole hills above Pahala), others have been buried under new lava flows. The slumps left us with old faults, now largely inactive. One of these is the Kaoiki fault zone, which follows the buried palis between Kilauea and the Mauna Loa summit. The Kaoiki zone is still active; in recent years it has caused damaging earthquakes in excess of M6. The Kaoki zone extends close to Kilauea.

Growing Kilauea

Kilauea began to grow just below the old pali. Was this accidental? It is difficult to know. A major removal of material can allow magma an easier pathway to the surface. Once that pathway existed, the location was far enough from Mauna Loa (just) to retain its own magma supply and declare independence.

The oldest Kilauea lavas are up to 300,000 years old. They were in part submarine: the pali may have been at the coast. At first eruptions happened at different places along the flank of Mauna Loa in shallow water, with some above sea level. Some of the lava formed sand, as it does when flowing into the sea. Kilauea may have formed an island. But there was also eruption at deeper levels on the southern flank, far from the rift zone. It is not known whether there was already a central vent and burgeoning rift zone, or that (like the current activity of Mauna Kea) the eruptions came from scattered locations. This pre-shield phase lasted until 130,000 years ago.

While Kilauea grew – slowly -, Mauna Loa was still spreading. The massive weight of its edifice and the magma insertion in the rift zone pushes the sides out. One of the remarkable things about Mauna Loa is that it maintains it height in spite of this spreading. A lot of its magma is not used for eruptions, but fills in the gaps, keeping up appearances and increasing the bulk. Volcanoes should not just be judged by the surface activity but also by their internal magma digestion. Kilauea, build on the slope of Mauna Loa, was carried outwards by the spreading of Mauna Loa. But once the conduit was established, that conduit would not move. So the eruptive summit stayed in place, while everything else was moving south, away from Mauna Loa. This is perhaps how the rift zone ended up 7 kilometers (estimated) from Kilauea, and why no other Hawai’ian volcano shows such an offset. It is the effect of standing on a giant. When the giant moves, so do you. You have to go with the flow.

From 130,000 years ago, the Puna ridge quickly grew, and Kilauea entered the shield-building phase. The timeline shows that Mauna Loa was already half a million year or more old before Kilauea started. Mauna Loa must already have been a very large volcano, and Kilauea always was a small bump on a much larger giant. Over time, Kilauea was pushed up both by its own growth and by the further expansion of Mauna Loa. But even now, Kilauea isn’t particularly high (1247 m, to be precise). To put Kilauea’s origin near sea level on an already large Mauna Loa means that its own growth has been modest. It indicates that Kilauea is relatively thin, and that its lava pile is little more than 3 km thick. Below that lies the pre-existing Mauna Loa lava.

Coombs et al 2006, Journal of Volcanology and Geothermal Research
Volume 151, 2006, Pages 19-49. Red colours indicate Mauna Loa lavas, blue and green is Kilauea.

This makes Kilauea a still very young volcano. It is a teenager, not so much a dwarf as a child carried by its giant parent. The lava flows are no more than 3 km thick, underlain by 6 km of Mauna Loa flows. The volume of Kilauea becomes between 5000 and 10,000 km3. Over 100,000 years of growth, the average lava production is 0.05 km3/yr. The current rate is twice as large; allowing for a slower start these numbers appear reasonable. And it explains why, in spite of all its activity, Kilauea remains little more than a bump on the landscape, dwarfed by Mauna Loa. Its adolescent growth spurt is still to come

On the figure above, the Hilina fault is (speculatively) shown as extending down 3 km, from the pali to the contact between the Kilauea and Mauna Loa lavas. The contact between the Mauna Loa lava and ocean floor is much deeper.

Kilauea has changed Mauna Loa. It acts as a buttress, and has fixed the southeast slope of the mountain. It is no longer able to shift downslope in the way it used to do. Slumping is stopped by the new bulk of Kilauea. But movement along the bottom still remains possible.


Of the ancient faults on the southeast flank of Mauna Loa, one remains active: Kaoiki northeast of Kilauea. The fault system predates Kilauea. There are steep faults which follow the old pali; deeper down is a near horizontal fault, the decollement between the lava pile and the ocean floor.

There are two different types of earthquakes here. The larger ones (1962, 1983, both M6+) occur higher up the slope of Mauna Loa, have a depth of 8-11 km and are nearly vertical strike-slip. They cause surface ruptures. The second type consists of swarms, with individual events up to M4. They occur close to the Kilauea summit. This activity here does not reach the surface – the movement stays well below ground. A 9000-year old lava flow has been left unbroken.

One of these swarms started on Oct 22, 2020, located a few kilometers west of Kilauea’s summit, near Namakanipaio Campground. This is just below the ancient Mauna Loa pali. The largest single event was a M3.5 earthquake. The quakes were shallow, 2-5 km. It was the 6th such swarm since 1983. The most recent one before 2020 was in 2012.

The swarm had no effect on Kilauea, in spite of being in its vicinity. That is also true for the earlier swarms. But the summit of Mauna Loa, much more distant, showed a pronounced effect. Over a few days it deflated by a centimeter (a lot, for a mountain this size), and the caldera contracted by a similar amount. The change started at the same time as the swarm. It is not clear which one was the cause and which one the effect! The Mauna Loa summit remained deflated until late December when it quickly reversed. As of this week, Mauna Loa is back to its pre-swarm size, and is rapidly expanding.

Hidden motion

What happened? USGS has not left us with clear guidance on this, so we have to guess (speculation alert), based on what we do know. The change in Mauna Loa was unusually strong and fast. There has been no other comparable event over the past 5 years – not even during Leilani.

My guess is that this was a downward shift of Mauna Loa in the direction of Kilauea. There was no notable effect on Kilauea, nor was there any ground movement. Neither did the GPS measurements catch horizontal shifts. Therefore, the movement happened underground, below the bulk of Kilauea. The swarm was nowhere deep enough to happen on the boundary between the Mauna Loa lava and the old ocean floor, but it does fit the upper boundary. The lava pile of Kilauea is 3 km thick, and this is also the approximate depth of the swarm. This potentially puts the earthquakes at the decollement between Kilauea and the underlying Mauna Loa lava. The Mauna Loa lava was on the move, but it left Kilauea in place. The underground downward movement reduced the pressure on the shallow Mauna Loa magma chamber; over a few days the magma chamber adjusted and this caused the deflation at the summit.

It has been suggested that Kilauea grew on top of a layer of sediment. Such a soft layer could help accommodate the slide, and could cause the movement to occur in swarms of small events, rather than a big quake. The deeper decollement with the ocean floor shows large quakes (such as the 1868 M7.9 Ka’u quake), aseismic creep, and regular slow-slip events. Underneath the Kilauea summit, it makes sense that the upper contact plane also shows movement, leaving the summit itself in place.

The unexpected eruption

We can take this model one step further (speculation alert). Forty days after the swarm happened, a dike nearly broke the surface at Kilauea. Twenty days later, a second dike did break through and started the Kilauea summit eruption, still on-going and forming a rootless lava lake already more than 210 meters thick. (As I write this, it is the first day the new lava lake has gone down rather than up (by 1 meter), and perhaps the eruption is nearing its end.) Perhaps the eruption was not a coincidence. It did not follow instantly after the swarm and Mauna Loa’s deflation, but the process may have been started by it.

The Kilauea summit has a number of magma chambers, at various depths. A very shallow chamber is suggested to be at a depth of some 2 km, and this is feeding the current eruption. A second chamber is suggested at 3-4 km depth and a third chamber at 8-11 km. Earthquakes tend to happen at these depths. It is a very complex region and very difficult to map. The deeper chamber seems to lie at the depth of the decollement between the lava and the old ocean floor, extending into oceanic crust. The intermediate chamber may lie at the contact plane between the Mauna Loa and Kilauea lavas. Magma finds it easier to form sills at such planes, which form a structural weakness.

From Lin et al., Geology (2014) 42 (3): 187–190. The white box shows the proposed deep magma chamber. The black dots show the locations of seismic activity, wit a focus at 3 km depth. The colours indicate velocity of the seismic waves.

A shift of the top of the Mauna Loa pile may have increased pressure on the intermediate magma chamber which lies in this region. The pressure caused magma to begin moving upward, and after a month enough pressure had transferred to begin breaking open the route to the surface. And so the eruption began, and this new pressure is why it is lasting unexpectedly long.

The idea behind this model is that the Mauna Loa pile is moving, but that the summit region of Kilauea remains in place, fixed by its magma chambers. The rift zone is not so lucky. It follows a structural weakness in the rock, and as the pile moves down, it takes the rift zone with it. The southern slope moves as much as 8 cm per year. Over 100,000 years, that amounts to 8 km, and this is indeed about how far Kilauea’s rift has moved away from the summit.

This is speculation. Is it right? It is a model that fits the data we have. There may be data we do not have and cannot see. But isn’t it nice to think that Mauna Loa could squeeze an eruption out of its little neighbour?

Albert, February 2021

(The title of the post is plagiarised from the paper: ‘Piggyback tectonics: Long-term growth of Kilauea on the south flank of Mauna Loa’, Journal of Volcanology and Geothermal Research, by Peter W. Lipman et al., 2006)

73 thoughts on “The piggyback volcano

  1. Kilauea is not small by any means… she is already almost 200 kilometers long and 55 000 km3
    Most of Kilauea is underwater and alot been built undersea in the puna ridge… Kilaueas undersea foot.
    Kilauea also extends far south east out in the ocean. This is already a giant in the making… insanely huge.

    Hawaiian shields can grow to over 130 000 km3 ( PunaHonu ), Most land volcanoes rarely exceeds 200 km3

    • Anyway great article Albert! I think that Mauna Loa will erupt soon… and the next eruption coud be close to 1km3 after accumulating magma so so long. The next eruption may come from the swelling south rift. It will be a spectacular sight with curtains of lava fountains for many kilometers and 1000 s of cubic a second. Perhaps a larger version of 1950…next time

      • The volume I think will be higher than that… A dike fed by 1 km3 of magma can reach basically any altitude, and that would potentially induce a feedback loop that will only result in a very big eruption, im talking a nearly laki scale event.
        Theres no 15 km3 a’a flows in Hawaii because the magma chambers that directly erupt are not that big but Mauna Loa has by a huge margin the biggest gravitational potential of any volcano on Earth, it positively dwarfs anything in Iceland by nearly a factor of 4. It might lack the volume but it is surely capable of erupting at a comparable rate, if not higher. Most of 1950 was erupted in 2 days, even though the eruption lasted 23.

    • I find the volume of Kilauea 10 times less than your number, because much of it is actually Mauna Loa.

      • The Puna Ridge is by itself enormous. Its No Mauna Loa under there.
        Thats a feature of Kilaūea. Kilauea is very long as an edifice.. just as ancient Haleakalā is thats another giant. Most of Kilaueas volumes erupts through its rift zones

        • It is enormous. But I did include the Puna ridge in the calculated volume. The volume of Mauna Loa is estimated at 75,000 km3 (minimum volume). Kilauea is a lot smaller than Mauna Loa.

          • Mauna loa is growing on Hualalai the same way as Kilauea is growing on Mauna Loa though. Theres possibility that Mauna Loas southwest rift is a pirated structure, south point has a submerged beach that was last at the surface when Mauna Loa was too small to explain it, at least supposedly.
            The Alika landslides could have been from Hualalai and Mauna Loa being more active invaded the space . Hualalai was by this point in its early postshield stage, so not active enough to reclaim the space, it also had become mostly silicic by that point too.

  2. The 1868 earthquake was located on the decollement between Mauna Loa and the ocean floor near Pahala. I find the estimated magnitude (7.9) to be quite staggering, this is an incredible amount of energy for a non-subduction quake. It illustrates the sheer amount of weight and pressure placed on everything adjacent – Kileaua is like the men of Minas Tirith, with Mauna Loa as Grond battering at the city gate.

  3. This is interesting, I have though read a paper that the linear formation of the volcanoes is actually not so linear, Kilauea and Loihi are both a lot younger, and Mauna Loa did already exist when Kohala was still in its late shield stage and Hualalai was the giant of the island. Mauna Kea also apparently never actually grew rift zones, it has always been radial, or perhaps had rifts only at close proximity to its summit like Etna.

    Basically what seems to have happened is all of the volcanoes on the north side of the island are syn-eruptive, Kohala and Hualalai were both once colossal twin volcanoes and have since been subsided by their own weight and the weight of the island as a whole. Mauna Loa and Mauna Kea are not dissimilar in age too, surprisingly, and Mauna Kea was just trapped in the interior of the island, probably it has been the height it is now for a long time and Mauna Loa has only recently reached anywhere comparable, in the Holocene. Kohala began to form 1.4 million years ago, Mauna Loa began to form 600,00 years ago, so that is 200,000 years between them on average, or even less if they pair up.

    In effect this means Mauna Loa is big because it could grow unobstructed in the last few hundred thousand years, it was once itself the small volcano getting shoved around by Hualalai. Kilauea only began its tholeiitic shield stage by some estimates as recently as 80,000 years ago, though a bit over 100,000 is more typical, and data from the above pater did determine the Hilo ridge of Kohala was formed largely in the earlier stage of growth, that is it built along the rift instead of up at first. In a the next few hundred millennia Kilauea will be allowed to really grow, but at the moment the only part of that we see is its exaggerated supply compared to what it would need to create itself, that might be a thing it aqquired only in the late Pleistocene or even in the Holocene. Loihi is also not actually a lot younger than Kilauea, and is at a similar point as Kilauea was about 100,000 years ago, so it might actually be a legitimately small volcano (for Hawaii anyway), it also got unlucky by having to grow right on the deep sea floor, Kilauea was nearly subaerial as you say.

    This is though a very thorough article and I do like the topic of discussion being Mauna Loa, it has been a bit boring in recent decades but its time to impress is on the horizon 🙂

    • The oldest data lava on Mauna Loa (a deep subsea sample) is dated to 650,000 years, with an uncertainty of 150,000 years. Of course the volcano must be older than that. I would quote the age as at least 600,000 years. It is very hard to date the origin of each volcano, because this lava is very deeply buried and inaccessible. The rift zone of Mauna Kea is not obvious and is deeply buried. It was studied in the 1980’s, and one paper (paywalled so unquotable) states ‘ Shield-building volcanism appears to have been concentrated along two principal rift zones (east and west). A third minor rift zone trends south’. Not much work has been done on it since.

      • I think the ages come from extrapolation of where the sample is and putting it on a simple cutout of the thickness of the volcano, then going back. The point that was being put across is that the age gap of Kilauea and Mauna Loa is bigger than Mauna Loa and Mauna Kea, or Kilauea and Loihi, it is a bit more complicated than most models that show a linear progression.

        • The initiation of shield stage at several hawaiian volcanoes has been inferred from dating of the submarine rift zones which often have much older surfaces than the subaerial lavas. The Hilo Ridge of Kohala, the Ka Lae Ridge of Mauna Loa and the Hana Ridge of Haleakala have all been studied.

  4. Yea! You drove the thought of winter from my mind….. Thanks, Albert! Enjoyed the journey. Your driving is smooth and i trust You to explain the scenery… A+ Best!mots

  5. If I remember right, it’s been suggested that kohala was a large volcano and mauna kea was a late-arriving neighbor that grew high on its flank. Among the reasons for thinking this was the lack of long rift zones, with the Hilo ridge being assigned to Kohala. Would anybody have more insight on this? I’m forgetting where I read the articles on this and I don’t have the leisure time at the moment to go dig (I’m in the middle of moving)

    • This is certainly possible. There is still uncertainty on whether to assign the Hilo ridge to Mauna Kea or to Kohala as it lines up with both. But an origin from Kohala seems likely. There is a bit of a tendency on the eastern and western side for volcanoes to borrow the rift of its predecessor. This is because rifts do not cross: if a new rift approaches an older one, it will bend into the same direction and run parallel. Mauna Kea may have been late to the party. Perhaps if Pahala would develop into something, it would become something like Mauna Kea.

    • Most of Mauna Keas thoelitic shield stage is buried and sunken. Mauna Kea looks like an overgrown version of Etna with its alkaline postshield cap.. Mauna Kea does not have a significant magma supply these days ( But its still probaly much higher than typical subduction zone volcanoes )

      Even postshield Hawaiian volcanoes haves a huge supply compared to the west coast US cascade cones

    • The oldest parts of the Hilo Ridge have an age of 1.15 Ma, which is too old for Mauna Kea, and thus shows the rift belongs to Kohala. The oldest lavas are transitional between alkalic and tholeeitic which shows the Hilo Ridge started to grow when Kohala was entering its shield stage, this is also about the time Haleakala left its shield stage. The dominance jumped Mahukona and went from Haleakala directly to Kohala. Kohala grew to be the second (or even first, since the estimate is not as good) largest volcano of Hawaii Island until the most of the supply went to Mauna Loa, which was around 500,000, can’t remember now the exact date.

      • You are forgetting Hualalai was also once a massive volcano, its south rift is what makes up a lot of the volume of the structure we call Mauna Loas southwest rift, it used to be longer and was cut off by the Alika landslides. The summit of Hualalai in its shield stage is at the current location of its saddle with Mauna Loa, deeply buried. The modern volcano is at a location which was once its northwest rift, because this spot is now above the deep source. Kohala is the same its shield stage summit is under Waimea not the summit it has now.
        Kilauea as it grows I suspect will heavily distort Mauna Loa, but I doubt it will bury Mokuaweoweo, it will be interesting.

        Basically Hualalai was as big as Mauna Loa is now though maybe not as tall.

        Please limit yourself to one identity. You are giving the spam demon kittens. -Admin

        • Hualalai was big and had a well-developed rift system (unlike Mahukona and Mauna Kea), but I don’t think it would have been comparable to Mauna Loa. The rift system of Hualalai is shorter than that of Mauna Loa, and only about half that of Kohala.

          Hualalai had a lot of competence, for a time it may have been in shield stage together with Kohala, Mauna Kea and Mauna Loa, I don’t think it was well positioned for spreading either. The best direction for spreading is towards the east, northeast or south, I think, which is where the Hawaiian Through is located, Kohala and Mauna Loa would have buttressed this direction and limited Hualalai to using the weak Kona decollement to the west. Currently, Mauna Loa could choose to spread SE or NW which are similarly partly buttressed by Kilauea and Hualalai respectively, but clearly goes for the SE direction, I think because it is easier to move the flank over the subsiding and sediment-filled Hawaiian Through, there may be even a pull force from the through, although this is speculative.

          Mauna Loa on the other hand had no competence because it was the only volcano in shield stage, erupting tholeiitic magma from the hotspot centre, for more than 200,000 years, until Kilauea started to erupt tholeiite less than 100,000 years ago. All the supply from the hotspot centre would have gone to Mauna Loa for a long time, and the eastward spreading was also easy throughout most of Mauna Loa’s life, only recently did Kilauea become an obstacle and deflected the flank motion slightly.

          • Hualalai shield stage ended about 150,000 years ago though, of every bit of information I could find. It sounds like you have a different source on that. I did think it was sort of strange though that it erupted trachyte for a long time, if it was evolving its upper magma system from its shield stage it should have turned to rhyolite, but it was already alkaline then.

            I also get the impression now that Loihi is actually very small for being where it is, its in the early shield stage but still way underwater, Kilauea is only 50,000 years ahead but a 200 km long rift with often times the whole hotspot to play with and is 1 km above the waves without even growing upwards that much, Loihi will probably join the island in time but not be a giant like Kilauea.

          • A note on language: I think the English word you’re looking for is “competition” there: Mauna Loa had none for a long time, etc.


  6. Thank you, Albert. Very accessible and informative! Tallis and you are keeping us well entertained thanks!
    And Hi! to Motsfo! Good to hear from you. Hope you are surviving the Alaskan snows!

        • Yes its a hornito, there is a video of it somewhere and it has lava erupting from parts of it. It is interesting how it is only erupting out of this one place now, before 2007 it was erupting out of a lot of scattered vents.

          Apparently the occurence of carbonatites here is very recent, within the Holocene, and before that it erupted lava like Nyiragongo, maybe not as hot, which it even still does on occasion like in 2007. Carbonatites are also geologically associated with kimberlite volcanoes and there is a very young kimberlite volcano on the Tanzania craton not far away from this area. It would be very interesting and very scary to see a kimberlite volcano erupt, a VEI 6 ultraplinian eruption in the middle of a continent after 3 hours warning…

          • It would be interesting to witness such an eruption (from a distance!). Any idea on what/where is the most recent such eruption?

          • Late Pleistocene at Igwisi Hills in Tanzania. This eruption was not very big though just a normal monogenetic cone at first glance. Most kimberlite fields are old, so chances are there will be more eruptions in Tanzania in the next million years.


            This is a much more impressive set of kimberlite volcanoes, and where I got the above number from.
            Related to kimberlites in that they also erupt diamonds, there are lamproite volcanoes that are probably about as common but are eroded away so only Cenozoic eruptions are easily visible. Lamproite volcanoes are effusive and the lava is probably very similar to the lava in Nyiragongo, it is an ultramafic alkaline rock much the same, just with diamonds. Maybe that Nyiragongo is on the opposite side of the Tanzania craton from Ol Doinyo Lengai… 🙂

      • Yes This is the most fluid lava on Earth! Liquid limestone soda.

        Lengai is even more fluid than Kilaūea and Nyiragongo. Some Lengai lavas almost approach the viscosity of water, but many are a bit higher. Its very fluid this lava..

        How does souch carbonate magma form?

        • Melt sodium carbonate, basically. Its not hard to do that, alkali metals tend to be strange in that it is very hard to decompose their salts, most carbonates turn into the metal oxide and CO2 if you heat, but Na2CO2 melts. Same with NaOH, most hydroxides turn to the oxide and lose water. Probably normal magma melts carbonate rocks deep underground.

          Pure Na2CO3 has a melting point of 850 C though so probably carbonatite is a very mixed rock, which is prett obvious because it is also black. If you mix calcium chloride and sodium chloride, which both melt at over 750 C, the mix stays liquid down to about 500 C, I have done this myself once and probably a very similar thing happens in carbonatite.

          I wonder if carbonatite lava conducts electricity, if it actually is basically a melted salt it should. I wonder if anyone has tried that before.

        • Hawaiian basalt magma is 1600 C in the astenosphere uncooled more than 100 km down. Is that magma at similar viscosities to Lengai at souch high temperatures?

          • It would have to be very low to be able to flow in the mantle, which is still pretty much solid. The asthenosphere is like wet sand, except a lot more densely packed because the crystals grow into each other. The magma is the liquid that is left between the crystals. In Hawaii there might not be much crystallization but it probably isnt a total meltdown of the mantle, though that might be possible and if that actually is what happens under Kilauea I wouldnt be surprised. Picrite basalt and komatiite are basically the same, just in picrite basalt the olivine is able to crystallize out before erupting, and Hawaii does erupt picrite basalt which is usually hotter than the normal stuff.

  7. Kilauea lava lake back to pre deflation levels after just a day of stronger inflation. Rim is already overflowing again.

    • It looks like there could be a dome fountain again. It is something to wonder if all of those older tubes and channels within the cone still exist and feed lava still, the lava lake now is right about at the level the west vent was when it first formed, just that now there is a big 30+ meter tall cone on top. The dome fountain that existed in early January was active when the lake was 24 meters lower than it is now, shows how much the cone has grown too.

      If i was to guess now if lava does eventually flood the vent it will either stop completely or continue passively, not unlike Pu’u O’o. If it stops though it wont be long before another eruption happens, especially as the volcano is overall in weak inflation since the north vent shut off. It is crazy that this entire eruption only started not even 2 months ago, how fast things are changing. Who knows what the caldera will look like after a year, but I wouldnt be surprised if a lot of the pit is gone or even if a new vent opens.

    • The response to the inflation is so fast that I wonder whether either the inflation is squeezing the bottom of the lake, or that the rootless lake has developed a root.

      • There are the drowned vents on the submerged ledge, one of them was still very active when the lake flooded it and it didnt turn into a drain.

    • Three areas of small lava outbreaks along the edge of the floor have also appeared over the past several hours. It would seem that sills have intruded from the lake into the solid ledge surrounding it, and are pushing some lava upwards through the contact with the crater walls. The rise in several meters of the lava lake must have pushed the lava into the ledge and up.

      This last deflation-inflation event was very large, with 9 microradians of amplitude it may have been the largest since the 2018 eruption at least. Since this last eruption started there have been multiple DI events strong enough to show in the UWEV-CRIM caldera length, something which had seldom happened before.

  8. Why is there so much small eq in iceland.? It has been this way a long time now. Something has to give soon

    • Wasn’t the system upgraded recently so it now records tiny quakes?

      Grimsvotn will erupt this year, I suspect it could be on a short notice because unlike in 2011 the lake is very high and will probably drain out in a big flood like it did in 1996, with that drop in pressure setting off the eruption. If anything the eruption might be somewhat secondary compared to that, unless a subaerial stage happens and there is visible red lava. I dont expect a big eruption, but it could be very intense if it is set off so fast.

      • While it is indeed common for Grímsvötn to have an eruption following a jökulhlaup, that is not what happened in 1996. The lake level was high and a draining event was probably close, but the eruption started before the hlaup and the melt water from the eruption emptied into the already overfull lake. The result was a record breaking jökulhlaup.

        • I mean the level is as high as 1996, not that there is any common factor to how that happened. Now is just the result of the glacier overall melting anyway as well as higher heat flow after 2011, and probably an overall rise on a longer scale since 1996, all combining. I read about the high lake level on, because I cant read Icelandic to navigate the IMO site.

          I dont know if there is any seasonality to jokulhlaups but it might be a bit less likely now than in the middle of the year, in summer. That is maybe the most accurate we can be on an eruption and it is a big if. July 2021 eruption of Grimsvotn.

          How deep is the lake? It must be pretty deep if it draining can induce an eruption, a deep caldera created in 1783 maybe.

        • Grimsvötn have never been this geothermaly active before since
          photography of the caldera began… soo its getting an increased magma supply compared before 2004. The Grimsvatn lake is quite deep probaly almost 250 to 320 meters enough to put pressure on a very shallow magma chamber 1,7 km down. The lake is of course keept liquid by souch shallow magma.

        • There exist chemosyntetic bacteria in the lake.. perhaps there are something like hot smokers on the lake floor? The magma chamber is as shallow as it can be.

          • Technically… Most “Hot Smokers” as we know them are below the depth where the pressure of the critical point of water is exceeded. ROUGHLY 2.5 km depth. Below that there is only a liquid phase of water. (No boiling or flashing to steam). Any shallower than that and you get at the very least, bubbles on the surface. Now, that doesn’t mean that you can’t have extremophiles busily munching away at greater depths down in cracks in the rock. Can fumeroles exist there? Sure, not a problem. But if it is shallower than the critical point pressure, you potentially have boiling/steam production. (with the right temperature)

    • It would be interrsting if Mauna Loa erupts while it has snow, at the same time as Kilauea is erupting. Kilaueas eruption now is potentially going to last a long time and it is going to stay very active for a long time regardless, chances are the next eruption of Mauna Loa will be a simultaneous eruption again 🙂

      It would be even more of a stark contrast if Kilauea makes a new satellite shield in the jungle east of Pu’u O’o and Mauna Loa erupts during that, same island but a world apart.

  9. Translation from Icelandic – must be perfect since it is from google translate

    A continuous GPS meter on Grímsfjall indicates that the volcano is expanding, presumably due to magma accumulation, as has been the case for the last quarter of a century between eruptions. There are indications that the rice plant is now further east than it has been for a long time. It is unclear what this change means. However, it could mean that magma is now accumulating in a new place, north of Eystri Svíahnjúkur. Measurements in the coming weeks and months will show whether this interpretation is correct. This spring, 10 years will have passed since the last Grímsvatn eruption and the volcano has expanded almost continuously during that time. Grímsfjall and land movements there need to be closely monitored in the coming months.

    (rice plant on Grimsfjall?? Extreme global warming or volcanic heat?)

    • Grimsvötn haves today an elevated/ extra magma supply compared to before 2011.
      Very large sulfur gas emissions was seen by last spring 2020 measurements.. suggesting that magma is very close to the surface

      • Grimsvotn is even more charged then it was back in 2011 if I am not mistaken, since I’m a felsic guy. I ask you Jesper Maficberg, which is more likely, a violent explosive eruption similar to 2011, a softer explosive eruption like 2004 or mostly effusive eruption.

      • The next eruption coud be quite large ( possible longer lived than just 5 days ) since alot of fresh magma have accumulated. If we are really lucky we may get a caldera surtsey Island. Huge ammounts of sulfur emissions been measured there last spring 2020.

        But Grimsvötn is not known for any shields or long lived activity. Its generaly not that type of volcano.

        I woud go for chads opinion too another brutaly fast pheratoplinian VEI 3 or VEI 4. Perhaps we gets a smaller version of 2011 again… or perhaps a large version of 1996 But in the caldera.

        • 1996 eruption was volumetrically larger than 2011 by DRE, both would have been similar subaerial eruptions, probably similar to fountains on Etna, a bit more viscous than Holuhraun or Hawaii. 2011 was under thin ice so it had no obstruction to the atmosphere and could blast high, in 1996 the main stage was over before it melted through and it was also a fissure, limiting the output of a single location.

          Is there any reasonably exact number for how much ice has melted since 2011? Lava can melt about 4x its volume of ice, so it might give an idea of how much new magma has been added.

        • Did you see this video Jesper? This guy makes paintings on top of 3D-printed landscapes. In this video he shows his work on Grímsfjall. At around 6:00 in the video he tells a story about some rescue workers going on a training mission shortly after the 2011 eruption. They went to the lake, found the water still warm and took a swim out to a small island that had formed in the middle of the open water.

        • Beautyful video Tomas Andersson!

          Sometimes I imagines sleeping 💤 up at svianukur a

          And Gollum is slowly creeping around in the arera .. slowly climbing up the caldera wall with his icey cold humid hands .. long fingers

          Rest of time he likley dwells in the Ice caves of Grimsvötn living of waste and dead animals that washes in from the coast.

          Insanely silmey caracther

          The Island in Grimsvötn 2011 is the infant stage of a table mountain

      • The very large sulfur emissions and open lakes ( just thin ice on them in winter ) are signs of magma accumulating at very shallow depths. Some parts of these lakes where steamimg hot in recent videos summer 2020…

        I dont know How much Ice that have melted But it coud be alot and alot dissapears in water drainage. The subglacial lake itself contains astonishing ammounts of water. The meltwater pools lakes are quite large. Alot drains into the rivers too.

        Grimsvötn haves a quite robust magma supply.. only Icelandic volcano that constantly swells after eruptions, suggesting the system is very active and lack of quakes suggesing and open magma system. But Kilauea and probaly even Etna haves larger direct surface supplys…

        Grimsvötn certainly haves the biggest influx among Icelands volcanoes.

        Eyfjallajökull.. is so tired 💤 not going to do anything in our lifetime

        • Well you are older than 11 as am I so Eyjafjallajokull has done something in our lifetimes 🙂

          Im not sure Grimsvotn is the only volcano with a constant supply. Trapdoor caldera inflation tends to be really obvious by nature. Bardarbunga erupts more magma long term just it erupts in big eruptions infrequently, it also rifts and that could mask reinflation. Hekla also outperforms Grimsvotn at least it did in the 20th century, it just has no shallow magma system to show deformation well and there are two magma sources that merge into one system there too, its complicated.

          I remember that a while ago now Carl made an article on Grimsvotn and the melting ice, and there was 3 km3 of ice melted since 2011. I dont know how that number was determined but that would be about 0.8 km3 of magma, maybe 1 km3 at high end. Its not likely a long term trend but a big eruption might be possible after all. But it also could just be small still, its all a waiting game until it happens. Best be watching for a flood because that will likely set it off.

  10. HVO update today says the SO2 on Feb 10 was 1600 t/day, a lot lower than other numbers and also outside of a DI event too. The lake has also not really risen in the last day or two, and since the north vent shut down the GPS cross caldera has been going up strongly, as though the eruption isnt enough to relieve pressure. I wouldnt be surprised if we see a new vent open soon.

    • The magma supply to Kilaūea is competely crazy these days
      The largest single supply on the planet.. the deformation is just going up and up

      Its normal for kilauea to erupt non stop constantly everyday in human timespanns

    • The shallow magma chamber is increasing in pressure, but I would say it has a long way to go before it can open a new path. Earthquakes at the summit and Upper ERZ, which act as a pressure gauge, are minimal, this shows recovery is still incomplete. Magma also continues to flow back from the ERZ conduit to the summit.

      • I more meant a new vent within the caldera, probably near an existing one. This probably would look similar to the side vent in January but more persistent and maybe with more vigorous fountaining and high effusion rates.

        It also could involve initial eruption under the lake which then turns to draining, which might rapidly pressurise the system. There is also the option of a vent east of Pu’u O’o now, as that is lower than the summit again.

    • The GPS stations do not show much inflation. OUTL is down a tad, CRIM and BYRL are almost flat, UWEV is up a tad. UWEV is moving west and a little north, while OUTL and CRIM are moving south. You’d need more stations to pinpoint the change. I expect that Halemaumau is sinking a bit under the extra weight of the lava, and the area around it is being pushed up in response. If that ‘moving up’ is focusing on the southwest-northeast rift bisecting the caldera, and esp on the northeast side of Halemaumau, that could fit the pattern, but other distributions may fit equally well. The northern part of the caldera has not been active for more than a century, and still showed a bit of contraction due to cooling of the rock. Perhaps one day this area will erupt again.

      • The lava lake wouldnt be sinking the ground, it was just in the ground under that very same place, overall there is less pressure from it anyway as the magma has pressure and higher volatile content, and the lava lake is going to be dense but not likely as dense as solid basalt until it is solid itself, though this lava is much denser than the overlook lake ever was.

        The deep inflation isnt common, there was probably a bigger batch of magma that got stuck on the way up and has since moved to the shallow system, but usually this source (source 2) is open and only the shallow halemaumau chamber (source 1) responds in post summit eruptions. Rift eruptions mostly come from source 2 but some distal eruptions in Puna also interact with source 3, which is the deep rift. The fact this eruption interacted with source 2 is unusual for summit eruptions, it shows its trigger was deep even though the lava that has actually erupted isnt fresh deep magma.
        Again a new vent isnt referencing an entire new massive eruption somewhere it is probably a new vent close to the already existing one, or an increase in eruption rate with higher fountains and building up the cone a bit

        • With my physics hat on, a few comments. It doesn’t matter that the lava was already under ground. That isn’t the point. The point is that it moved. Two options. 1, it came from sideways, in which case the weight moved and depressed a new area. The area it come from would have risen after this. 2, it came from below. That is a mogi model. A deeper reservoir affects a larger area on the surface, depressing a larger area a bit less. Move it up. and the depression becomes smaller and deeper, i.e. the central area will depress and the area around it will come up. Basic physics.

          • Is there enough magma to actually give this effect though? I dont disagree this effect takes place but is it enough that it would be visible over everything else. HVO has instruments that are in the caldera, the one they use the most is actually not public, and they dont talk about any sort of deflation of any kind except for DI events which are visible anyway. When the eruption began the intrusion pushed up the ground within Halemaumau and the ground sank outside of it within the greater caldera, this rise was very similar to the intrusion in early December which didnt erupt so we can see the effect it had without being drowned by the greater noise of the magma chamber deflating.

            The strange thing is the gas emissions are the lowest for the whole eruption but the effusion rate isnt low, theres even a dome fountain at the inlet again and the lake is overflowing the edges and erupting in the moat.

          • What I described is after the inflation phase. You assume that the magma is already in place and is moving up. There are other effects. Degassing changes the density. Cooling does too. DI events are funny things and they have been very large recently, perhaps because the magma is so shallow. I was trying to make sense out of the GPS data. HVO will do a much better job, and of course they see all the data. we don’t.

Leave a Reply