Sizing Up the Next Eruption

Earth Observatory Blog

Sizing Up the Next Eruption

A type of volcanic crater called a caldera forms when a large eruption drains a magma chamber beneath a volcano, causing the ground to collapse. These eruptions can eject anywhere from 10 km3 to over 1,000 km3 of magma, devastating the local area and sometimes cooling the entire planet.

Although rare, one or two caldera-forming eruptions happen every century, such as the 1991 eruption of Pinatubo in the Philippines and the 1912 eruption of Katmai-Novarupta in Alaska. In between these catastrophic eruptions, however, caldera volcanoes have many small eruptions that pose a hazard only to those in the immediate vicinity.

Many caldera volcanoes around the world are currently in a state of unrest, suggesting that an eruption is potentially brewing, which leads to a few important questions: can we tell how big will the next eruption at these volcanoes is likely to be? Will it pose only a local hazard, or could it be large enough to have a regional or even a global impact?

To attempt to tackle this question, we have been looking at Rabaul, a caldera in Papua New Guinea. Rabaul demonstrates nicely the scale of the different types of eruptions. The caldera itself forms a flooded bay about 8×14 km3 and is actually the result of multiple eruptions (see photo below).

Panorama of Rabaul Caldera, in Papua New Guinea, taken from the Rabaul Volcano Observatory on the caldera rim

The most recent of these is known as the “1400 BP” Ignimbrite, as it occurred roughly 1,400 years ago (it has now been more precisely dated at between 667 and 699 AD), and it discharged at least 11 km3 of pumice and ash. This actually makes it rather small for a caldera-forming eruption, although still big enough to make a rather large hole in the ground and to blanket the surrounding area with deposits up to 30 m thick.

After the 1400 BP Ignimbrite explosion, volcanic activity has continued, but has been less violent. There have been 17 periods of activity since the earliest historically recorded eruption in 1767, with eruptions occurring from multiple vents spread out across the caldera. The most visible remains of these eruptions are the twin cones of Vulcan and Tavurvur, which have both been built up over several successive eruptions. The 1994 eruption destroyed a large part of Rabaul town; the surviving part of the town can be seen at the bottom of the panorama above. However, it is clear from the size of the two cones relative to the size of the caldera as a whole that a caldera-forming eruption would be orders of magnitude more destructive.

Deposits from the 1400 BP eruption

Building up enough magma for a caldera-forming eruption takes time. Deep beneath the volcano the mantle melts, producing basalt. Basalt is a fairly runny magma, and when it reaches the surface it tends to erupt as lava flows or fire fountains, like happens at Hawaii and Iceland. If this basalt gets trapped deep underground, then it can begin to transform through a process known as fractional crystallisation. As the magma cools, crystals begin to form, and the remaining liquid becomes stickier and more explosive. At Rabaul, fractional crystallisation turns basalt into a type of magma called dacite, and it is this magma that erupted explosively at the surface both during the 1400 BP eruption and more recently from Tavurvur and Vulcan.

A close-up of Tavurvur

We can use the deposits from repeated eruptions of Rabaul to try and piece together what goes on beneath the surface of a caldera. For example, in the deposits from Tavurvur and Vulcan we can see streaks of basalt mingled in with the dacite that makes up the bulk of the erupted magma. The basalt and the dacite didn’t have the time to mix properly, suggesting that the basalt only entered the magma chamber shortly before the eruption.

By comparing samples taken from different eruptions, we can see if any patterns emerge. Do we see an injection of basalt before the 1400 BP eruption? In fact we don’t, which suggests that Rabaul’s plumbing system changed after the 1400 BP eruption. By going back further, and looking at other eruptions from before the 1400 BP Ignimbrite, we can build up a better picture of how caldera volcanoes work.

All photographs are taken by Dr Gareth Fabbro.

Geographic Area: