Here is the third part of my Kilauea series that was promised, a bit more delayed than I would have wished though. Many things have happened at Kilauea since the previous part. A sill intrusion took place in the Upper Southwest Rift in August, then on September 29, about a month after the sill, lava erupted inside Halema’uma’u Crater and has been filling it up ever since, meanwhile earthquakes never cease raging under Pahala, a M 6.2 earthquake took place not too far from the Pahala swarming and at similar depth in October. However what I’m here to talk is about the plumbing of Kilauea
Knowing the plumbing of a volcano is something really difficult, because we can’t see it, it is underground. The magmatic structures are invisible, or are they? Here I piece together different types of evidence and observations that helps unravel a little about how the interior of Hawaii and Kilauea is structured.
One way of knowing the plumbing is by looking at earthquakes. These happen when one block of rock slips past another. The movement of magma can cause earthquakes by applying pressure against the rock that surrounds it, known as volcano-tectonic earthquakes. There are also other types of earthquakes thought to be caused directly by the movement or the volumetric changes of magma, like long-period earthquakes and tremor.
In Hawaii the distribution of earthquakes is also related to the residual gravity anomaly. Residual gravity highlights the differences in the strength of the gravitational field depending on the location, and corrected for altitude and other factors. Anomalies in the gravitational field can be used to find underground masses of denser material. For example the volcanoes of Hawaii have a high gravity anomaly, because the interior is thought to contain heavy material, like gabbro, that has formed from solidified magma, causing the stronger gravity attraction, and revealing the internal structures.
The deep rifts under Hawaii
When the location of earthquakes occurring under Hawaii is plotted there are some interesting things that show up. Most activity seems to be concentrated in a band oriented WSW-ENE that runs across the southern flanks of Kilauea and Mauna Loa. The activity happens at two levels, 0-12 km and 30-50 km depth. Below 50 km there are very few earthquakes in Hawaii, probably the rock is too hot here to break. The band does not show in the 12-30 km range either.
In the map below, 6-10 kilometre-deep earthquakes are shown in yellow-green while those that are 36-50 km deep are shown in purple-pink. A main WSW-ENE band can be distinguished. The deeper band has the same direction but it is displaced more to the south. Another N-S band of earthquakes shows up on the southeast flank of Mauna Loa which merges with the main WSW-ENE band.
Clusters of earthquakes are also seen under the summits of Mauna Loa and Kilauea with contrasting depths. They are long-period, LP, earthquakes presumably related to gas bubble growth in magma. Mauna Loa’s LPs are centred at a depth of 45 kilometres. In contrast Kilauea’s LP activity has been identified be 5-10 kilometres deep using relocation methods, much more shallow. The main vertical conduit of Mauna Loa must be very robust and deeply rooted compared to Kilauea.
The band of earthquakes is well aligned with the East Rift Zone, the ERZ, of Kilauea, and as we will see is parallel to the contour lines of residual gravity. This is why I think the earthquakes occur in front of a deep rift zone of which the East and Southwest Rifts of Kilauea are their shallow expression. The term ERZ was created to refer to the linear zone of intrusions and eruptions extending east of the volcano. Kilauea however occupies a rather central location behind the band of earthquakes, which goes both WSW and ESE from the volcano, so the term East Rift doesn’t quite work. So in order to refer to it I will be calling it the Ka’u-Puna Deep Rift.
The deep spreading of Kilauea has been well studied in the past. Magma is continuously flowing into the deep rifts under the volcano, which inflates the rift zone, and causes the south flank of the volcano to bulge seaward, it moves up, east and south. Over time the flank is contracted until can no longer resist the strain. It breaks. Then the whole flank shifts seaward by as much as meters, and at the same time subsides as it falls back to its former relaxed shape. The cycle of rift swelling and flank contraction then repeats again. That is why Hawaii is a prolific source of large earthquakes, having produced 55 Magnitude ≥ 6 earthquakes since 1823, on various faults, including those in front of the Ka’u-Puna Rift axis.
The largest earthquake was an M 7.9 in 1868. The Great Ka’u Earthquake. It is easier to understand what happened when taking into account the Ka’u-Puna Deep Rift. Shown below are 5-3 km deep earthquakes, in white, that follow the Ka’u-Puna structure, the other N-S rift is also marked in yellow, which I here call Kapapala Deep Rift. In 1868 a large dike intrusion preceded the earthquake which would. The location of the intrusion must have extended from the summit of Mauna Loa, which erupted before the earthquake, to the eruption fissure of 1868 marked in red, which erupted afterwards. The Ka’u earthquake was observed to cause a 1-2 meter subsidence of the entire coast of Hawaii Island, from Kalae in the southwest to Kapoho in the southeast. The event was disastrous. A tsunami obliterated many settlements along the southern coast of Hawaii which were never rebuilt. Still today this coast remains largely unpopulated. The death toll was 77 people.
The rupture would most likely have extended all across a reverse fault in front of the Ka’u-Puna Deep Rift, from Kalae to Kapoho, relieving the strain accumulated throughout the rift flank, it was a full rupture. The 1868 intrusion occurred at the western end of the rift setting off the earthquake. The flank then slipped away. Because Kilauea is also on top of the rift its Southwest Rift Zone snapped open and erupted immediately after the slip, the summit of Kilauea drained and collapsed.
The deep rift is usually thought to extend to a depth of only 8 kilometres under the rift zones of Kilauea. However I think this could be an incorrect assumption. Looking at the earthquake distribution I find it more likely that the deep rift extends down to more than 50 kilometres and that it extends under both Kilauea and Mauna Loa. I do think that the Ka’u-Puna Deep Rift is a most important structure to Hawaiian volcanism producing most large earthquakes, it is perhaps responsible for the apparent connection between Kilauea and Mauna Loa too.
Inflation of the rifts of Kilauea has been modelled and shows that in the 1961-1982 period it averaged 0.18 km3/year, this is nearly 6 cubic meters of magma every second that end up under Kilauea! During the 1983-2018 Pu’u’o’o eruption the spreading slowed down. Nonetheless, considering that there may be further regions of magma accumulation at depth, then most of the magma supplied to Hawaii actually remains underground to make intrusions, probably dikes and sills.
The Pahala Swarm is an enigmatic clustering of earthquakes under the town of Pahala. In the past 3.5 years seismic activity has skyrocketed spectacularly in this location. Shown below are 26-38 km deep earthquakes under this area. The location of 41 km deep tremor is also marked, most Hawaii tremor events originate from that spot. The swarm has a somewhat rectangular shape and the long axis is parallel with the Ka’u Puna Rift, probably because it is caused by it. As I mentioned earlier as it gets deeper the earthquake band seems to shift southward. This could be because of two reasons. First, the rift might widen as it becomes deeper. Second, dike intrusions in the East Rift Zone are known to be inclined, they dip southwards at 80º degrees, if this slope is extrapolated then, as you go down, the East Rift Zone ends up where the deepest earthquakes are happening.
One interesting possibility is that Kilauea is a satellite conduit of Mauna Loa. A conduit meant to feed magma along the Ka’u-Puna Deep Rift. The Pahala Swarm and the deep tremor source could be involved with this pathway. None of this however is clear, rather an interpretation of how the structure works, and it could be otherwise too.
Apart from the Ka’u-Puna Deep Rift and the Kapapala Deep Rift, there is also a third more faint linear swarm of earthquakes. It extends from the Southwest Rift Zone of Mauna Loa towards Hualalai. I will refer to it as the Ka’u-Kona Deep Rift. Large earthquakes have also occurred here, including a M 6.9 in 1951. This swarm is less active, however it is nonetheless quite dangerous since the Kona coast is heavily populated and a future large earthquake could have devastating consequences.
I find it interesting that the Southwest Rift Zone of Mauna Loa looks almost like a symmetry axis, with the Ka’u-Kona Deep Rift and Hualalai on one side, and the Ka’u-Puna Deep Rift and Kilauea on the opposite side. The residual gravity map of the island also seems to reflect this structure. The strongest positive anomalies are near the summits of the volcanoes, which is expected since there is a lot of shallow intrusive activity there, but the anomalies extend parallel to the seismic lineaments. Mauna Loa’s intrusive core is oriented mainly southward along the Kapapala Rift. Hualalai’s anomaly is offset from its summit and elongated along the Ka’u-Kona Rift. And finally Kilauea shows a strong alignment with the Ka’u Puna direction. Overall the shape of the residual gravity field over the southern half of the island is somewhat triangular, centered on the south of Mauna Loa and with three vertices marked by Hualalai, Kilauea, and by the Ka Lae Ridge, the submarine rift zone of Mauna Loa.
Right below, the surface rift zones of the three volcanoes are shown. Initially near the summits, shown in orange, their direction is somewhat different from the seismic lineaments, but as they extend away they align with them.
My interpretation of the earthquakes, gravity, distribution of fissure eruption is that there is a tremendously complicated magma structure down there. Magma makes pathways along rift zones, accumulates in sills and dikes, and connects various volcanoes of the island.
This is not unique to Hawaii, the Pacific seafloor is full of ancient volcanoes many of which are tens of million of years old. There is something however that they all have in common, rifts/dyke swarms. The volcanoes can be shaped like ridges or stars, because eruptions happen along linear zones which intrusions propagate through. Hawaii is however a magnificent opportunity to see these structures live and breathe.