How did Kilauea's crater form?

Where did all the magma go when Kilauea’s caldera formed? Perhaps as with a backyard sinkhole, you might ask where and how did the stuff disappear? Did an underground opening already exist (such as a lava tube), did something “dissolve” away (such as rotting vegetable matter), or did a kind of conveyer belt remove material and take it somewhere else (such as moving magma)?

We have this problem understanding how Kilauea’s caldera formed. It wasn’t by explosion, because no debris is piled around the caldera. It must have formed by collapse. And therein lies an unsolved mystery.

Geologic evidence indicates that the caldera appeared in about A.D. 1500, give or take several decades. It developed at the end of, or soon after, a 60-yr-long eruption from the `Aila`au shield just east of Kilauea Iki, which ended in about 1470 after flooding much of Puna with lava flows. The caldera formed at the start of, or shortly before, a series of explosions that radiocarbon ages date to about A.D. 1500.

That’s not terribly long ago. Both geologic and cultural evidence should exist for such a major volcanic event. Indeed, the events are likely chronicled in the Pele-Hi`iaka chants. But neither cultural nor geologic evidence tells us where the rocks disappeared when the top of the volcano sunk to form the caldera.

How much material was removed? The volume is uncertain but very large. If the entire caldera formed at one time, we estimate a volume of 6 cubic kilometers (1.4 cubic miles), equivalent to 780 million loads for a 10-cubic-yard dump truck. That estimate is based on the caldera’s dimensions when first surveyed in the 1820s and on guesses as to how much deeper it might have been.

Most geologic models assume very rapid removal of material to form a caldera. The argument is that a large void cannot support itself underground, so an empty volume of a few cubic kilometers (cubic miles) could not exist for long.

If so, a huge volume of magma must have almost instantaneously left the reservoir under Kilauea’s summit, leaving a void into which the summit fell. Where did the magma go?

We haven’t found it. Nowhere on Kilauea did a lava flow of huge volume erupt in about A.D. 1500. Could an eruption may have taken place offshore, along the Puna Ridge? Recent seafloor investigations found no large lava flow of appropriate age.

Maybe there was no eruption. Maybe magma stayed underground, migrating quickly from the summit reservoir to some place in the east rift zone. A conservative volume of 3 cubic kilometers (0.7 cubic miles) can be contained in a huge dike 60 km (37 miles) long by 5 km (3.1 miles) high by 10 m (33 feet) wide. We have no way to test for such a dike, so we must consider one.

Another possibility is that the eruption of the `Aila`au lava flow, with a volume of 6 cubic kilometers (1.4 cubic miles), created the caldera. This is appealing, because the caldera formed just after the eruption ended, suggesting a link between the two large events.

If this happened, then supply of magma to Kilauea must have slowed or stopped during the eruption, so the magma reservoir eventually emptied and then collapsed. Perhaps lessened magma supply during an eruption is necessary to form a caldera.

The `Aila`au eruption took place over 60 years. Did a void slowly grow within the volcano, not collapsing until it reached its final huge size? That would seem unlikely to most geologists. But what if we view the magma reservoir as a deep mine, with many interconnected passageways, rather than as a simple balloon? A mine has large void space, yet is stable to a point.

Since the Pu`u `O`o eruption started in 1983, the southern part of the caldera has been sinking 6-10 cm/year (2.5-4 inches/yr). Could this gentle sagging be an early stage of, or precursor to, another caldera collapse as the summit reservoir slowly empties?

Clearly there’s a lot we don’t know about how the caldera formed. We’ll eventually figure it out, but for now the origin of the caldera is as puzzling to us as the origin of your back-yard sinkhole is to you.

Activity Update

Eruptive activity at Pu`u `O`o continues. Most lava flows have been at the lower end of the rootless shield complex along the Mother’s Day lava tube south of Pu`u `O`o. Such flows have been small and short-lived but are gradually advancing toward the top of Pulama pali. On March 20 a new lava flow (the Kuhio flow) erupted from the south base of Pu`u `O`o; it remains sporadically active. Vents within the crater of Pu`u `O`o are incandescent and sometimes visible from Mountain View and Glenwood. No lava is visible from the Chain of Craters Road.

Two earthquakes were reported felt during the week ending on March 31. A magnitude 3.2 earthquake, felt at Miloli`i, took place at 6:17 p.m. March 27 at a depth of 6 km (4 miles) about 16 km (10 miles) east of the summit of Mauna Loa. The next day, residents of Honoka`a and Papa`aloa reported a jolt at 5:42 p.m.; this event had a magnitude of 3.1 and was located 8 km (5 miles) south-southwest of Pa`auilo at a depth of 10 km (6 miles).

Mauna Loa is not erupting. The summit region continues to inflate slowly. Seismic activity remains very low, with 1 earthquake located in the summit area during the past week.

This article was written by scientists at the U.S. Geological Survey’s Hawaiian Volcano Observatory and is republished by with permission.

Leave a Reply

Your email address will not be published. Required fields are marked *