How to predict a supervolcano eruption

Thursday, 2 February 2012

by Rachael Bayliss

A study of pre-eruptive magmatic processes and their timescales may improve our understanding of the events that precede supervolcano eruptions.

Credit: iStockPhoto

LONDON: Supervolcanic eruptions, which can devastate entire continents by pouring out poisonous gas and scorching magma for months, may be predictable.

Supervolcanoes, which form large cauldron-like structure called calderas, are capable of ejecting thousands of cubic kilometres of magma in mere hours and can have a significant global impact. Having advanced warning before these events take place could save millions of lives in the future, but until now, scientists have said that they are very difficult to predict.

"Our work, supported by studies of other caldera systems elsewhere, suggests that the chamber growth can undergo marked acceleration in the decades prior to eruption," said Timothy Druitt of the Universite Blaise Pascal in France, and lead author of the study published in Nature today.

"The point is that this late-stage major growth occurred on a very short timescale compared to the repose period. This is in one sense shocking, as it means that these monsters can re-activate and build up to eruption on the human time scale."

The explosive build-up

Until recent years it was thought that it took hundreds of thousands of years for huge amounts of magma to gradually accumulate within supervolcanoes before an eruption could occur. Volcanic events of such magnitude are rare and therefore no supervolcano has ever been monitored during this pre-eruptive phase. This means that many of the precise precursors and indicators to such an explosive event have gone undocumented.

Druitt's team was able to re-create an essential timeline of the melt processes taking place in the magma-reservoir within the final century before an explosive eruption. "At Santorini we think that the chamber grew by at least 10% in that time, pressurising as it did so," said Druitt. "On the other hand, establishing that this accelerated growth occurs can enable us to try and detect these deep magma movements by surface geophysical techniques."

The clue is in the crystal

Druitt and his colleagues studied chemically zoned crystals in volcanic rock from the caldera-forming eruption of the Santorini Volcano in Greece during the late 1600s BC. They used a new combination of techniques that enabled them to retrieve data for multiple timescales from each of over 300 crystal samples; these were from all four stages of the eruption process and allowed them to build a more complete picture than ever before.

The zoning textures of the crystals were analysed using a broad range of experiments and it was the vital differences in the composition of the cores and outer layers of the crystals that lead the team to uncovering the complex processes taking place in the magma reservoir prior to the eruption. Until recently, similar studies had focused on Zircon crystals and what they indicated about the longevity of the melt.

Predicting a supervolcano eruption

It is hoped that the results of this study will allow scientists in the future to have a much clearer insight into what is happening deep underground at known dormant, but potentially active, supervolcanoes such as the ones at Yellowstone in the U.S., Campi Flegrei in Italy and Taupo in New Zealand.

"Caldera volcanoes commonly experience crises of seismic and ground deformation, but in general only one in ten such crises will result in eruptions. Our work suggests that the chamber growth events leading to eruptions may be big ones that we are able to recognise," said Druitt.

"Essentially, we have always had problems determining how rapidly a 'restless' caldera such as Yellowstone or Campi Flegrei might become 'primed' for a major explosive eruption. In this regard, the research is particularly important as it constrains this timescale, and because the time-scale is so short," commented Bill McGuire, a professor geophysical and climate hazards at University College London in England.

Volcanology and sedimentology expert Ray Cas of Monash University in Melbourne, who was not involved in the study, added, "The advance this particular paper made was in understanding the origins of the very particular zoning. Remarkably, different layers in the [crystal] rims must involve multiple diffusion batches." This explains how such a vast quantity of magma can form over such a geologically short period of time. Cas went on to discuss that the best indicators of an imminent eruption will still be episodic periods of ground movement, along with an increase in seismicity, changes in both the composition and rate of gas releases and most telling of all initial minor explosive events that would most likely occur within the preceding year of an event.

Both Druitt and Maguire stressed the need to maintain and expand the current monitoring programmes on restless caldera volcanoes. "Some caldera volcanoes such as Yellowstone and Campi Flegrei are extremely well instrumented; others in more remote regions of the globe are not, but should be," said Druitt.