Skip to main content

Can We Forecast Caldera Collapses?

The formation of cauldronlike volcanic depressions is enormously destructive—but it may also be predictable

Mount Pinatubo in a satellite photograph.

Credit:

Getty Images

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


In 1991, hundreds of people died as a result of the caldera collapse of Pinatubo volcano in the Philippines. Could such a collapse have been forecasted if there had been appropriate sensors?

This was what an international research team set to find out. A caldera is a large, cauldron-like volcanic depression formed in response to magma chamber drainage during large eruptions of magma. The loss of support of the rock column above the magma reservoir may initiate the collapse of the overlying ground, resulting in caldera formation.

Caldera collapses are rare events (only seven have been recorded over the last century). They are also particularly destructive; they can produce catastrophic changes in the shape of the volcanic edifice, its environment and its volcanic activity.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


Left, before the collapse on October 31, 2006  and right, after the first collapse on April 17, 2007. From April 10, the geometry of the caldera structure did not change significantly. The deepest part of the caldera is 340 meters; for comparison, the height of the Eiffel Tower is 324 meters. Credit: OVPF and IPGP

A recent study I published in Scientific Reports with a number of colleagues describes 48 collapse events associated with the 2007 incremental caldera collapse of Dolomieu (La Réunion Island, Indian Ocean) over more than nine days. The collapse resulted in an estimated 340 meters of surface subsidence. The study detected the first collapses at depth with a broadband seismometer (a sensitive instrument that records a ground motion over a wide range of frequencies) prior to the surface rupture and Dolomieu caldera formation. These observations demonstrate the feasibility of forecasting caldera collapses for a volcano with a high aspect ratio (that is, the thickness versus the width of the magma chamber roof).

Identifying the occurrence of the first collapse at depth is of major importance in evaluating the triggering factors and in anticipating the caldera surface formation. This detection can help to forecast future caldera collapses and subsequent events, such as explosive eruptions or atmospheric impacts. In the past two decades, only four caldera collapses have been monitored by dense geophysical sensor networks: the 2000 Miyake-jima, Japan; the 2007 Dolomieu, La Réunion/France; the 2014–2015 Bárðarbunga, Iceland; and the 2018 Kīlauea, Hawai‘i.

Laboratory experiments on caldera formation highlight the importance of the aspect ratio of the magma chamber roof at the onset of caldera formation. For low aspect ratios (£1), the caldera collapses as a coherent piston and caldera surface subsidence is observed from the onset of the caldera formation. This was likely the case during the Fernandina caldera formation (in the Galápagos archipelago) in 1968, where the roof aspect ratio was around 0.3. For high roof aspect ratios, laboratory experiments and numerical models have predicted the occurrence of precursory collapses at depth before the onset of the surface subsidence; however, observations obtained from adequate broadband seismometers were lacking.

In the case of the 2000 caldera formation at Miyake-jima and the 2007 Dolomieu collapse, the aspect ratios were high: between 1.9 and 3.8 for Miyake-jima and around 4.25 for Dolomieu. Therefore, precursory collapses at depth observed at both Miyake-jima and Dolomieu were likely due to the high roof aspect ratio.

Piton de la Fournaise is a basaltic shield volcano in the southeastern part of La Réunion Island, in the western Indian Ocean. Its summit cone has two summit depressions: the Bory crater in the west and the Dolomieu caldera (about 1 kilometer in diameter) in the east. In 2006, the August eruption (August 30 to December 31, 2006) completely filled the Dolomieu Caldera and caused its lower, eastern border (2480 meters of elevation) to overflow. On April 5 at 20:48, the Dolomieu caldera collapsed—three days after the beginning of the largest historical eruptions of the volcano. This collapse was associated with an earthquake of magnitude 4.8.

For the first time, tilt signals related to changes in ground slope associated with the first collapse at depth were observed on a broadband seismometer installed on the volcano; these occurred approximately 20 hours before the surface rupture of the Dolomieu caldera collapse. This work illustrates the importance of having at least a single broadband seismometer deployed in volcano monitoring.

The 1991 Pinatubo (Philippines) caldera collapse was also associated with a high roof aspect ratio. Unfortunately, no broadband seismometer was available at that time on this volcano. Early detection of the onset of a caldera collapse can provide crucial information to understand caldera formation and thus to minimize risks for the nearby population and visitors.