Kilauea II: Roots of the Hawaiian Islands

In my previous article, here, I discussed how Kilauea and Mauna Loa volcanoes are connected to each other through the Pahala Swarm. Now I have to deal with a confrontation of theories that is inevitable. There is a classical model of how Hawaii works. It is all about the mantle plume. The classical view establishes that the different volcanoes have various magma chemistries which come from different parts of the plume, a mantle upwelling under Hawaii, and they follow vertical paths towards the individual volcanoes. Where is the problem? Mauna Loa and Kilauea have different chemistries. If they share the same magma source, which I’m quite convinced they do, then the classical model is wrong.

Nowadays no one dares go against the traditional view of how Hawaii works. But I will. I will explore the amazing possibilities that could come from a whole different way of thinking about Hawaiian volcanism. When all the magma comes from one.


The many paths of magma differentiation

As I pointed out in the preceding part of the this series, Kilauea has a highly variable magma supply. From 1840 to 1960 the supply was low. During this time certain elements, known as incompatible elements, increased their abundance in the magma. At the same time the Sr87/Sr86 isotopic ratio decreased, while the 206Pb/204Pb ratio increased. In other words the chemistry of the magma was shifting. During this time Mauna Loa was more active.

Something interesting happened in 1960. A large eruption took place in the East Rift Zone of Kilauea which used up so much magma that the summit nearly caldera-collapsed. This event produced an enormous amount of deflation. The next year a surge of magma came in which produced the highest levels of inflation ever recorded at Kilauea. The seismic pathways deep under the summit also reorganized. More east rift eruptions and intrusions followed in the next years. Ever since then, when not erupting, Kilauea inflates rapidly with magma. More so than in 1840-1950.

The summit of Kilauea starting to collapse in 1960. The not yet solid lava lake of 1952 is leaking into the collapsing crater. From USGS.

This seems to be a top-down process triggered by the 1960 eruption. What is interesting is that it affected the geochemistry enormously. All the trends drastically reversed. Since 1960 the incompatible element abundance of Kilauea has been decreasing while the Sr87/Sr86 ratio grows and the 206Pb/204Pb ratio falls. Or in other words, the volcanic processes affect the magma chemistry. Different chemistry thus doesn’t necessarily mean a different source but can also mean that the processes acting over the magma are not the same.

The incompatible elements are those that are left out from the crystallization process. Barium, Niobium or Lanthanum, for example, are some of the most incompatible. A magma that has undergone more crystallization is expected to be more enriched in them. They become concentrated in the remaining liquid. Here mushes might be important. Magma sponge? The rift zones are thought to be filled with these mushes. Because the rifts are continuously spreading enormous amounts of material go into them that never reach the surface. Excess incompatible elements might be expelled into the conduits or magma chambers during these processes. If magma has to reach Kilauea from Pahala through the southwest rift, then what happens to this magma as it moves underground through the rift zone? This is poorly understood. If anything is clear though it is that the faster the magma moves towards the surface the less enriched it would be. Slower would be more enriched in incompatible elements. What happened 1960 was a change in how fast magma rose into Kilauea, a transition from low to high supply, so that was possibly the reason for all the chemical changes it triggered.

Instead I don’t know much about the properties of the Sr87/Sr86 and 206Pb/204Pb ratios. How or why, or when they would change, I don’t know. I will have to explore this issue in the future. The events of 1960 do suggest however that they can also be altered by volcanic processes. Each Hawaiian volcano usually has particular isotopic ratios that distinguish them from each other. Mauna Loa and Kilauea have their own distinct ratios. This is usually assumed to mean that the magma source is different. I however would consider that like in 1960 there must be volcanic phenomena that alter these characteristics, and that it would rather be a difference in the way magma is transported towards the surface.


The one magma

Another type of geochemical variability that exists in Hawaii is the alkalinity. This refers to the amount of NaO2+K2O  in the magma. Volcanic rocks are classified into different types depending on the alkalinity. The kinds of rock that Hawaiian central volcanoes commonly erupt are, in increasing order of alkalinity, increasing order of NaO2+K2O: tholeiite basalts, transitional basalts, alkali basalts and basanites.

Types of volcanic rocks depending on silica and alkali content. From Wikimedia by Woudloper.

As usual this is explained by these types originating from different parts of the mantle plume, with the central part generating the tholeiites, and the edges making the more alkaline types. Problem is that Mauna Loa and Kilauea have slightly different levels of alkalinity. And also I will be soon saying that Hualalai is connected to them, and Hualalai erupts transitional basalt. So I prefer alternate models.

All the types of magma erupted in the Hawaii Island could be obtained from an homogeneous melt similar to Mid-Ocean Ridge Basalts. It just depends on the minerals.

Hawaiian magma first crystallizes the mineral chromite, but it does so in insignificant amounts, second comes olivine, which is more important, then clinopyroxene, and then plagioclase, also important. The order has to do with the temperature. Olivine has about 40% silica, while clinopyroxene and plagioclase have around 52-53%. The parental magma is somewhere in between, around 47 % silica. This is a crossroad. If the magma crystallizes mostly olivine then it will increase in silica rapidly. If it crystallizes mostly clinopyroxene or plagioclase it will actually decrease in silica and move towards more alkaline compositions. Temperature could have a lot to do in shifting the process to one side or the other. Higher temperature would favour olivine. Depending on the amount of minerals of one type or the other that it forms, then it would result in a very wide range of alkalinity.

From this point of view the tholeiitic lavas erupted by Kilauea and Mauna Loa are tholeiite basalts that are almost unchanged from the original melt and have underwent only small amounts of olivine crystallization. The alkaline lavas of the older volcanoes would be more evolved and generated through a more clinopyroxene and plagioclase dominated fractionation. This also seems to be supported by how the amount of incompatible elements seems to be directly proportional to the level of alkalinity. The most ultramafic and alkaline magmas always are the most enriched. Presumably meaning they have underwent more crystallization.

The high concentration of incompatible elements in the basanites that Haleakala currently erupts suggest that about 75 % of the original magma could have crystallized, 25 % reaches the surface. The even higher enrichment of ultramafic nephelinites from the volcanic fields of Oahu would suggest that barely 3-7 % of the original magma has survived the journey there. Alkaline magmas also tend to concentrate water. I don’t if this is true for CO2 too.

Volcanic field eruptions of Oahu. Ultramafic nephelinite is frequent here. From Wikimedia by Mbz1.

To me it makes a lot of sense that some hawaiian volcanoes would carry melt more efficiently than other. For example Mauna Loa and Kilauea have mature plumbings, with central conduits, and magma chambers, and will carry the melt much more efficiently towards the surface than the mushy volcanic fields of Oahu. This should be reflected somehow in the chemistry. I think that alkalinity is the way this is reflected. In this theory not the source but rather the volcanic structures and processes would produce the different chemistries.

Both Kilauea and Mauna Loa are also known to have produced rare eruptions of alkali basalts from their flanks. This could happen when the magma bypasses the summit plumbing and instead rises through the rifts, which Loihi seems to demonstrate quite well, by erupting tholeiitic basalts from the summit, and both tholeiitic and alkali from the flanks. This I think argues against the classic mantle plume model for the magma generation. Instead it is more consistent with the different pathways taken.

Magma rhizomes

If Kilauea volcano feeds laterally from Pahala, 40 kilometres away, then it is also likely that other older Hawaiian volcanoes did the same. I have investigated this question and reached some interesting theories. The theory I’m going to show here might seem unnecessarily complicated, but it is the best way I know of explaining the different types of volcanoes in Hawaii, their timing, the relationship to each other and how the Mauna Loa-Pahala-Kilauea connection works. It explains A LOT.

A lot of insight came from comparing Kilauea to Piton de la Fournaise, a volcano near Madagascar.  These two volcanoes are often considered as analogues to each other. There is one structure that I find particularly interesting. The northeast rift zone. This NE rift connects the two volcanoes of the island, the older Piton des Neiges to the northwest, and the younger Piton de la Fournaise to the southeast. The rift erupts lavas more alkaline than those of Piton de la Fournaise, meaning that it could be a deep structure that bypasses the summit. Earthquakes under the rift run at a depth of 20 kilometres, and marks a path leading to below the flank of Piton des Neiges edifice, here it descends to a depth of 30 kilometres, and when also considering that a lot of CO2 has been observed to be emitted by this rift, it does seem likely it represents the feeder path of Piton de la Fournaise. Does the younger volcano feed from the older one?

The two volcanoes of Reunion are connected by the NE rift of Piton de la Fournaise, Deep earthquake activity and gas emissions show that the NE rift is a magma pathway probably carrying malt towards Piton de La Fournaise. Created in Google Earth.

I imagine the rift zones as being similar to the rhizomes of plants, and the volcanoes being stems, a volcano can grow a rift laterally and then a second volcano might sprout from this rift system. So then what if some volcanoes can act as feeders to other volcanoes? What if this happens in Hawaii? It turns out this could provide the answer to many questions about Hawaiian volcanism that I had.

Volcanic Complexes

The construction of Hawaiian volcanoes doesn’t follow a perfect progression. One would expect that as the Pacific Plate moves over the Hawaii Hotspot it would make one volcano, then another, and another, and so on. But no. Actually it is more sequential, with one group of volcanoes growing during a certain period and then activity moving to the next group. I will call these volcanic complexes. I will call each complex after the volcano that I think was the first in each of them.

This is the timing of the main tholeiitic stage of each Hawaiian volcano of the past ~6 million years:

Loihi: Entering tholeiitic stage. New complex?


Hualalai Complex:

  • Kilauea: 0.1 Ma-present.
  • Mauna Loa: 0.47 Ma-present
  • Hualalai: ?-0.1 Ma.


Mahukona Complex:

  • Mauna Kea: ~0.7-0.4 Ma
  • Kohala: 1.1-~0.5 Ma.
  • Mahukona: ?-0.6 Ma


Lanai Complex:

  • Haleakala: 1.9-1.25 Ma
  • West Maui: 1.8-1.4 Ma
  • Kahoolawe ?-~1.15 Ma
  • Lanai: ?-1.3 Ma


Koolau Complex:

  • East Molokai: Somewhere before 1.75 Ma
  • West Molokai: ?-1.9 Ma
  • Koolau: 3.2-1.8 Ma


Kaena Complex

  • Waianae: 3.9-3.3 Ma
  • W and E Kaena: ?-3.6 Ma


Lihue Complex:

  • Kauai: 5.4-3.9 Ma
  • Niihau: 5.5-4.7 Ma
  • Kaula: ?

There are other reasons to think each of these groups are separate structures. The Lihue Complex forms a topographic seamount that is separated from the other volcanoes of Hawaii. There was a large shield volcano located in what is now the western side of Kauai Island, this was the Waimea shield. It is thought to have had a giant caldera 20 km across. The island of Niihau are the remnants of the Waimea’s rift zone. The rift runs in an east-west direction connecting the three volcanoes of the complex, with Waimea in the centre, linking to the Lihue shield volcano on the eastern side of Kauai, and to the Kaula volcano in the west. All part of one large structure.

Lihue volcano complex. Niihau is the rift zone of Waimea Shield Volcano, partly destroyed by landslides. From USGS.

Complexes usually share the same shoreline break, except the Mahukona Complex. What is a shoreline break? The enormous weight of Hawaiian volcanoes pushes them down into the Earth. As they are active they grow and make an island that keeps getting bigger, but afterwards they start waning and the island sinks into the ocean. The shoreline break is the ancient submerged coastline that marks where the edge of the island once stood. Each group seems to have roughly the same shoreline break which basically means that they rose and fell together.

The main shoreline breaks are shown in colours, they mark the size of the islands at their greatest extension. Each shoreline break seems linked to a particular complex. The volcanoes are represented as circles.

This is particularly impressive for the Lanai Complex. These four volcanoes grew to form the largest Hawaiian Island, of all times? It is called Maui Nui, meaning Great Maui. They were also joined to the emerged remnants of the Molokai volcanoes. At the time it was even bigger than the island that Pūhāhonu, the so called Earth’s Biggest and Hottest Shield Volcano” created when it was erupting 13 million years ago. After Maui Nui became inactive it sank 2 kilometres into the crust, so that the saddles between the mountains were inundated, and the ancient landmass “broke” into the four islands of Molokai, Lanai, Kahoolawe and Maui.

Multiple volcanoes of each group often share landslide fans. For example the eastern side of the three Koolau Complex volcanoes collapsed in the Nuuanu-Wailau landslide complex, that with a volume of ~7000 km3 is among the largest on Earth. This makes sense, if the volcanoes grow together they should also reach gravitational instability together.

Nuuanu-Wailau landslide complex. From USGS.

So why the complexes? I speculate that each of these could be sharing the same magma feeder. Each group of 3-4 volcanoes would feed from one specific spot that was originally melted by the hotspot. Magma would travel sideways using the rifts. I have confirmed that the rift orientations of the volcanoes in each complex could potentially intersect with each other providing a network of magma passages between the various volcanoes of the complex. I will look more into this aspect in the next post where I will explain how the ancient rift of Mauna Loa, the Ninole Hills Rift connects with Pahala, which in turn connects to Kilauea through the Southwest Rift. Pahala being a junction in the rift network.

Each complex of volcanoes and its rift zones. Created in Google Earth.

I should also mention that there is a large active volcanic area that covers the northern islands, and an additional extension of the abyssal plains to the north. The North Hawaiian Arch Volcanic Field, it is among the few largest volcanic fields of the world. Some of its submarine eruptions are estimated at 40-70 km3. It generally erupts basanites or nephelinites. I don’t think it was created by the hotspot, while its exact age is not known it was already active when the Koolau volcano complex grew, because the debris of the landslides covers some of the lava flows. It still remains active, so it is fixed to the plate and not the mantle, and much of it is significantly off-track from the hotspot too.

One theory is that it is produced by flexure of the lithosphere around the load of the islands, so that it is pushed up and and undergoes decompression melting. I find the decompression effect very small though, about 1 km of uplift which is not much. In comparison lithospheric extension can cause the mantle to rise in the order of tens of kilometres. The eruptions also sometimes happen within areas that have actually subsided, like Kauai or Oahu.

So I find it most likely to be a tectonic feature. Volcanism strengthened enormously from Lihue to Lanai volcanoes, as the hotspot and the volcanic field met. So there may have been some tectonic-hotspot interaction in the recent pulse of volcanic activity of Hawaii.

Cyclic volcanoes

The idea that multiple volcanoes feed from the same magma feeder also provides an explanation to the different types of volcanoes that exist in Hawaii. Central volcanoes come in 2 types.

Most are covered entirely in tholeiitic basalts, or tholeiitic plus transitional, with only sometimes minuscule amounts of alkaline lavas, and they have low slopes.

Some have instead erupted large volumes of viscous alkalic lavas sometimes completely mantling the tholeiite lavas in all subaerial exposures, this gives them a more stratovolcano-like shape. It turns out that there is one stratovolcanish system in each the Mahukona, Lanai, Koolau and Kaena complexes. These are Mauna Kea, Haleakala, East Molokai and Waianae volcanoes. These four have erupted the vast majority of the alkaline lavas among the central volcanoes of Hawaii. They are the highest and steepest and must have at some point towered proudly above all others, majestically like Mauna Kea, the white mountain, does today:

Topographic shape of each complex. They are not to scale respect to each other but rather are intended to show the similarities in slope steepness patterns and relative heights of the volcanoes within each complex compared to the others.

The Lihue complex also has its own alkalic volcano, Kaula, which erupted phonolites, some of the most alkaline and silicic lavas of Hawaii. Kaula is very small though compared to its younger counterparts. Still, each complex has one.

In the Hualalai complex there are also some differences in the alkalinity. The most surprising is that while Kilauea is the more active, more powerful volcano, its lavas are actually slightly more alkaline than those of Mauna Loa. Alkalinity is related to the waning of the volcanic system and also as I’ve argued probably to lower temperatures and clinopyroxene/plagioclase crystallization. Why does the more active volcano erupt more alkaline lavas? There is also a difference in the amount of incompatible elements, which are higher in Kilauea. If the magma moves laterally, like from Pahala, and beyond, it might lose some of its heat, or crystallize more, interact with the rift, and make the magmas more incompatible element enriched, and more alkaline.

These gradients are roughly seen in all the complexes. In the Lanai complex, one end erupted the strongest tholeiitic lavas of Hawaii, the basalts of Lanai. The opposite end is truly the inverse with the most alkaline lavas of all Hawaiian central volcanoes, the basanites of Haleakala. Even in its peak tholeiitic stage Haleakala had twice as much incompatible element contents of Barium and Niobium compared to Lanai. As I will show these gradients also show in the isotope ratios erupted by each volcano.

Mauna Kea. A final volcano type. From Wikimedia by Nula666.

My theory is the following. The hotspot melts an initial volcano, like Lanai. The initial volcano develops a rift system which branches and propagates, the volcanic activity migrates with the propagating rift so that new volcanoes are created. Haleakala is an example of a final volcano, the one which is situated in the propagating end of the rift. When the next complex arises, the previous complex wanes. The waning complex will become increasingly alkalic due to the death and freezing of the plumbing system. These alkalic magmas will be provided mostly to the volcano at the propagating side of the rift system, the final volcano, like Haleakala, will erupt most alkaline lavas. If this is true then the volcano progression would be something like this:


W Kaena>E Kaena>Waianae

Koolau>W Molokai>E Molokai

Lanai>Kahoolawe>W Maui>Haleakala

Mahukona>Kohala>Mauna Kea

Hualalai>Mauna Loa>Kilauea

Magma would be transported from the initial volcano to the final creating a gradient in the chemical composition.

The complexes compared to the inferred rift propagation and hotspot track. Created in Google Earth.

This could explain some interesting landforms of Hawaii. Particularly the Hawaiian Moat. The moat or through is a depressed area of the ocean floor that is located on the eastern side of Hawaii. But why the eastern? This is might seem like a silly detail, however this detail had me quite intrigued for some time. The last 5 complexes have all grown eastward/southeastward, perhaps because they influence each other to follow similar directions. If the edifices of each complex build on top of each other then a lot more piled-up weight will end up on the eastern side, thus depressing mostly that part.

From USGS.

There are 4 prominent rift zones that emerge eastwards from the moat side of the islands. The Pauwela Ridge, the Hana Ridge, the Hilo Ridge and the Puna Ridge. Each emerges from one of the last 4 complexes. I think that these rifts are the propagating tips of the branching rift systems of each complexes, they are seen best where they erupt underwater and make steep ridges.


The isotopic ratios

Lastly there is one other geochemical characteristic of the magmas that could be related to the complexes. Hawaiian volcanoes show different isotopic ratios. When the volcanoes are plotted one complex after another and in the order in which the complex seems to have grown, in which the rift has propagated, there are some interesting gradients that show up. It seems to me that as magma is transported laterally it might be affecting the isotopic ratios too and causing their variation among Hawaiian volcanoes. I don’t know however which mechanisms could be affecting them.

Ranges of 87Sr/86Sr isotope ratios in the volcanoes of Hawaii.

206Pb/204Pb isotopic ratio ranges in the volcanoes of Hawaii.

In the next part I will be looking at how rift zones work and how Kilauea was born.

291 thoughts on “Kilauea II: Roots of the Hawaiian Islands

  1. Now I am on top of Storihrutur, I think there is a Webcam behind me, but I don’t know the URL


    PS it is spectacular!

    • Not unlike the way it looked 24 hours ago. The speculation was that there were two vents connected by a fissure. My comment at the outset yesterday was that it looked like a fissure eruption.

  2. 22.19, flow starts t be visible on the Natthagi webcam. Great view right now, it’s about two thirds of the way down to the bottom.

  3. There seems to be barely any variation in flow; can anyone confirm or refute this?

  4. Huge amount of lava is heading into Lower Meradalir, but there’s no sign of surface flow near the vent; there must still be some fair-sized tubes underneath somewhere.

    • I think someone had mentioned that there is a small tube at the front right of the cone – was visible on a recent drone overflight as a flow going into the rock, no whirlpool, but definitely noticeable. As well, during the first recent long flows, it was said that there was a lot of fountaining/flaring (rock meets lava type) at that spot as well.

      I couldn’t direct you to the correct video if my life depended on it, sorry (but it was recent, so you might be able to ferret it out).

  5. I am watching the very hot ground around the sides of the cone, and the smoking has increased the last hour and a half or so. I wonder if due to the pressure in the active cone, that some of the hot magma underneath has moved a bit closer to the surface? See

  6. I remember reading that Hilo Ridge was specifically a rift zone of Kohala volcano partially buried under more recent Mauna Kea lava, and not a rift zone of Mauna Kea itself. How does the theory factor this?

    Also, I wanted to browse the comments section to find thoughtful responses to this blog post and the theories within. But it seems the vast majority of comments are either specifically about what’s going on in Iceland, or (sometimes more commonly) are comments (otherwise without context) that, after spending moments trying to consider how they pertain to the blog post, I must conclude can only be talking to Iceland. It seems like I have to browse several pages at a time between reaching comments that are actually about this blog post. Yes, what’s happening in Iceland is very interesting, for sure, and I can see how the comments section of the latest VC blog entry may be a natural magnet for users mentioning new developments in that eruption, but might it be a good idea for users to start all their off-topic root comment threads with something along the lines of “(Off-topic.)” if it’s not about the blog post itself? Might make it easier to sift through what are now hundreds of comments by topic relevance.

Comments are closed.