Let’s image the simplest, most boring volcano possible. It consists of a magma balloon fed by a straw. The ground is perfectly flat above the balloon. Magma comes up the straw at a constant rate. It would be easy to predict such a volcano. The balloon inflates until it bursts. Eruption happens. Because the supply is constant and the balloon always has the same capacity, then every eruption should be evenly spaced and have about the same size. It would be like a clock. Wouldn’t life be easy that way? Volcano tourists could catch up to a volcanic eruption just as it starts. Evacuations could be planned with time. There would be no uncertainty or loss of control over the situation. But volcanoes are not clocks, obviously, and Kilauea can tell us why.

Contrary to most scientific depictions and models Kilauea is NOT a balloon. Magma is actually stored in a vast, intricate system of interconnected chambers, conduits, and sheet intrusions. A complexity that is probably shared with other volcanoes in the world. But in this case, more active, very shallow, and well monitored.

In this article, I’m going to talk about the shallow portion of Kilauea’s magma system. As opposed to the deep system, that I’ve looked into in the past. This is the most dynamic part. Inflating and deflating, feeding intrusions and eruptions, and collapsing into craters and calderas, it is a never-ending action. The depth of the plumbing decreases slightly outward from the summit.

At the top of the volcano lies Halema’uma’u Crater, the beating center of Kilauea. A column of magma rises up under Halema’uma’u, often it makes its way towards the surface, making a semi-permanent lava lake, the fiery heart of the volcano. The top of this lava lake lies at an elevation of 1 km above sea level. Dyke intrusions that propagate from the walls of this great magma shaft run very close to the surface, running no more than a few hundred meters underground. The Great Crack of 1823 was like this. The Great Crack formed during a catastrophic draining of a massive lava lake atop Kilauea. Speleologists have descended into the dyke. The main passage of the 1823 Great Crack dyke lies 120 meters underground, is a crack 10 meters wide, and travelled a distance of over 30 km from the summit of Kilauea.

In 2018, there was a magma chamber that collapsed. This magma body was 2 km under Halema’uma’u, or about 1 km below sea level. Actually, it may have been two stacked chambers that collapsed. First, the Halema’uma’u conduit started collapsing, it was followed by a 1.5×1.5 km area of piston collapse, then a larger 2.8×1.7 km area. There may have been deeper levels that were untouched. So maybe the magma chambers under Halema’uma’u structure into stacked levels of a Christmas tree-shaped complex.

Deeper than these magma chambers we encounter the two rift conduits. The Southwest Rift Conduit and the East Rift Conduit, which run about 2 kilometres below sea level. But what are the rift conduits exactly? A long-lived lava flow will over time, develop into an organized system of melt transport. It will grow a main lava tube, which carries material from the erupting volcano. This is somewhat like the aorta does with blood, the main artery that distributes blood from the heart across our bodies. The lava flow keeps growing through new breakouts and inflation thanks to the material supplied by the tube.

The two rift conduits of Kilauea are like lava tubes, only 3 km underground, 2 km bsl. They get the magma from Halema’uma’u, the central heart of Kilauea. This material is distributed into the two rift zones. The dykes and sills of the rift can be compared to the lava flow, some may have solidified long ago, others are still molten and can inflate or deflate, store up magma, and even evolve magma into new chemical compositions that are quite common in the distal sections of the rift plumbing (for example, the notorius Fissure 17). New breakouts of a lava flow can be compared to the new dykes and sills which grow from the rift conduit. The processes that happen deep inside a rift zone are hidden under a great thickness of rock, but are also similar to those that happen on the surface of lava flows, that we can directly observe and understand.

When Kilauea inflates with new arriving magma, if you head to the Kilauea monitoring page of the Hawaiian Volcano Observatory there, you will see how the two rift conduits flare-up with earthquakes. The conduit expands and its walls start to fracture from the pressure. Thus, when Kilauea swells with magma, the rift conduits inflate and experience seismic activity. They are hydraulically connected to the heart of Kilauea so they are very sensitive. Brief subtle reductions in the pressure of Kilauea’s central magma chamber under Haluma’uma’u, called DI events, are enough to shut off seismic activity in the rift conduits, but earthquakes will usually resume as soon as the DI event ends a few days later.

The following map shows earthquakes along the two rift conduits during selected periods of inflation. Red lines mark the fissure swarms, the directions that dyke intrusions follow. Initially the rift conduits start near Halema’uma’u, and then head southward. The summit of Kilauea (Halema’uma’u) is detached from the rift system, which, like a hungry living creature, yearns for the south flank of the mountain, the fundamental source of spreading that opens room for dyke intrusions. That is why at first the rift conduits take a southward direction, advancing for 8 kilometres in order to reach the rift axis. Then they align and run parallel to the dikes. They connect the summit with the rift zones, and because of this, some publications call them the connector conduits.

Lets start with the southwest side of Kilauea. In this flank many dykes propagate from Halema’uma’u. There is another highly important location, however. An intrusion locus 3 kilometres south of Halema’uma’u, from which sills and dykes often start. The following map shows the intrusions from Halema’uma’u in red and from the upper rift conduit in yellow.

Traditionally, it is said that Kilauea has two magma reservoirs, the Halema’uma’u magma body and the South Caldera magma body. This is the simple interpretation that I was critiquing earlier. I’m of the opinion, that we should be talking of a complex of sills and dykes that radiate from certain area. Those areas are Halema’uma’u and the upper Southwest Rift Zone conduit. The intrusion complex under Halema’uma’u sometimes coalesces into a magma chamber of variable size that can collapse afterwards.

The so-called South Caldera magma body is probably a vast plexus of dykes, sills and inclined sheets that propagate from the same intrusion locus 3 km south of Halema’uma’u that is in turn supplied from the summit. South Caldera sills have been modelled to be about the same depth a the rift conduits, 2 km bsl. The base of dykes, like in 1981, also seem to run about 2 km bsl.

Below is an interferogram (a deformation map) for an intrusion on August 23, 2021. This is one of the most convincing candidates for a new sill being intruded. It was sudden, short-lived, with propagating seismicity, and caused by inflation of a flat sheet of magma. Otherwise, it may be difficult to distinguish new sill intrusions from inflation of the already existing ones; or worse yet, from the signal of dikes intruded at the same time as the sill, and concealing the sill inflation. It follows seemingly the same path as dykes in 1974 and 1981.

There is yet a third location that feeds intrusions in the Southwest Rift Zone that must supply the Kamakai’a Hils. The Kamakai’a Hills are a cluster of cones that is offset from the two main fissure swarms of the Southwest Rift and that often erupts evolved basalt and basaltic andesite magma. This basaltic andesite must have evolved from some stale sheet intrusions. The location is around the distal end of the Southwest Rift Conduit. It must be a small system of sills and dykes far from the summit of Kilauea, that being in the periphery of Kilauea’s system develops evolved magma compositions. The Kamakai’a Hills may have last erupted sometime around 1800 AD (Kamakaia Waena lava flow). In the past several decades, I don’t think there has been any magma intrusion here.

The other rift of Kilauea, the East Rift Zone, is even more complicated and intriguing. Like the SWRZ it has multiple fissures swarms, and evolved magma pockets (including a pocket of dacite magma that was accidentally drilled at the Puna Geothermal Venture in 2005, some 45 km away from the summit!). But it also has unique aspects, in fact, so unique that no other volcano featurs this structures so far from the main summit, I’m talking about the pit craters, and the shield volcanoes that develop features of central volcanoes (the 1200 AD Kane Nui o Hamo, and the historical eruptions of Mauna Ulu and Pu’u’o’o). The pit craters, large collapses up to a km across, have been particularly overlooked. These incredible structures, are they collapsed central magma chambers to their own systems of intrusions? The position of the dikes, radiating in tight strands from them, could indeed suggest this to be the case. But in any case, they do appear to occupy nexus positions where the connectors meet the various dike systems.

The pit craters are found along the East Rift Connector and make up the 15-km-long Chain of Craters. They are Keanakako’i, Luamanu, Hiiaka, Pauahi, Aloi, Alae, Makaopuhi, and Napau. Some of them collapsed in the late 18th century, probably in 1790. Each seems to have its own fissure swarm. They are very small in volume; the largest two, Napau and Makaopuhi, may have at times held 0.2 km3 of magma each. They are not to be confused with the shield volcanoes. The shield volcanoes, like Pu’u’o’o, also have pit craters from their shallow shafts of magma that collapse during dike intrusions and eruptions, yet these sit atop shallow cones of magma or pyroclastic material. The pit craters, instead, have no remnant whatsoever of any sort of shield structures in their locations.

They can also be divided into 4 pairs. Makaopuhi and Napau basically share the same fissure swarm. Makaopuhi will usually send a short dyke eastward for no more than a few kilometres, at the same time, another dyke will start growing also eastward from Napau which can reach a much greater distance of over 20 km. Aloi and Alae often work together making double dyke intrusions. The last double Aloi-Alae dyke event was in 2007 and was well observed. It will usually happen that Aloi will send a dyke westward into an elevated area known as the Koae Fault Sytem where dykes never erupt, while Alae will send a dyke eastward to lower elevations where it erupts more easily. The same is true for the Hiiaka-Pauahi pair, Hiiaka sends the dykes mainly westwards into the Koae Fault System where it cracks open the ground but doesn’t erupt, while Pauahi sends dykes eastwards, often erupting. They complement each other. The last pair or trio is Keanakako’i-Luamanu which does very small double intrusions that do not advance very far.

When a dyke intrusion occurs in the Upper East Rift, earthquakes start near one of the craters, while immediately the summit starts deflating, there is an hydraulic connection. Within hours lava may breach the surface and gush out making fountains inside or near the crater. But it may also stay underground and not erupt. The pit can fill with a lava lake that later may perhaps drain back into the dyke. Fissures and cracks will start opening up at increasing distance from the pit crater, and as mentioned earlier, it can happen that one dyke goes west while another goes east. Such paired dyke intrusions are typical of Kilauea’s Upper East Rift.

Dyke intrusions can also originate from certain spots of the East Rift that are located further east than Napau Crater. It usually takes longer for such dykes to grow and reach the surface. The deflation of the summit can also lag several days behind the intrusion or eruption start. Magma erupted first will be chemically evolved, sometimes basaltic andesites.

There is a sweet spot for magma intrusions under Honualua volcano, where the Puna Geothermal Venture is now located. This place is 45 kilometres away from the summit of Kilauea but nonetheless magma can travel all the way here through the labyrinth of magma pathways that underlie Kilauea’s rifts, although pressure changes may take several days to go or come from here if the path has not been cleaned up beforehand. This is the location where the dyke intrusions of 1924, 1955, and 1960 commenced. Honualua may have also supplied the evolved basaltic andesite that came out of Fissure 17 during the 2018 eruption of Kilauea. Small earthquakes have been happening at this location ever since proper seismic monitoring was established.

In 2005, while drilling an injection well at the Puna Geothermal Venture on Puu Honualua, a pocket of dacite magma was encountered at a depth of 2400 meters underground. The magma had a silica content of 67%, far higher that typical Kilauea basalts. It is probably related to this same focus of intrusion activity under Honualua volcano.

There are two other main storage areas under the East Rift Zone, probably consisting of sill and dyke complexes. One is centred at Puu Kamoamoa, which is also the location where Pu’u’o’o grew in 1983-2018 (although Pu’u’o’o itself was formed from a Napau dyke). The other is located downrift under Lilewa Cone, which is upslope of Highway 130. During the 2018 eruption the two storage areas deflated. Later, they underwent reinflation. The Lilewa area twice experienced episodes of rapid inflation (seen in the JOKA GPS), first in the two months or so after the 2018 eruption, and then in March-May 2020, plus slow gradually decaying uplift between the events. The Kamoamoa area underwent massive scale inflation from the end of the 2018 eruption until October 2020, affecting an area of about 20 x 7 km, and reaching about half a meter of uplift at the centre.

 

Conclusion

In this post, I’ve tried to give a quick look into the magma architecture of the Kilauea volcano at its most fascinating complexity. I sometimes think of Kilauea as if it were fourteen volcanoes in one. Additionally, Kilauea itself is also in complex, possibly direct, interaction with Mauna Loa.

Kilauea is far more than it seems of the surface; it is a true volcanic gem. Sometimes, quite literally, as some of the recent fountains have shown. Happy holidays to everyone! Let’s relish the beauty of lava:

Relevant links

The seismicity of Kilauea”s magma system

The December 1965 Eruption of Kilauea Volcano, Hawaii

Comprehensive High‐Precision Relocation of Seismicity on the Island of Hawai‘i 1986–2018

Two hundred years of magma transport and storage at Kīlauea Volcano, Hawai’i, 1790-2008