Dozens of volcanoes erupt every year. Donald Dingwell wants to know how eruptions occur and why they happen when they do – so he simulates them in the laboratory.
A volcanic eruption can blow the top off a mountain. How then can such massive explosions be modeled in the laboratory? – Don Dingwell well remembers the skepticism he encountered among colleagues when he began to experiment with volcanic rocks and to simulate the eruptive process on a laboratory scale. “Around 1989 we began to explore the idea, in 1993 we built the first equipment, in 1995 we set off the first explosion, and in 1996 we had our first experimental breakthrough,” he recalls. Leading international science journals published his reports on the spectacular experiments, and magazines and newspapers featured stories about ‘the man-made volcano in the basement’ of the Bavarian Institut of Experimental Geochemistry and Geophysics in Bayreuth, where he then worked.
Shortly afterwards, Dingwell, born and educated in Canada, moved to Munich to take up the position of Director of the Section for Mineralogy, Petrology and Geochemistry where he later cofounded and led LMU’s Department of Earth and Environmental Sciences. And his toolbox has since grown. Now four model volcanoes have been built in his lab and, as he remarks in his hybrid Bavarian-Canadian accent: “Experiments are now as commonplace in volcanology as they have always been in physics, chemistry or biology.”
So what can experimental volcanology tell us? The world map that Dingwell has on his computer monitor during our conversation is littered with black dots. “Each of them marks an active volcano,” he says. “Every year 50 to 60 erupt, many of them near densely populated or agriculturally important regions.” Furthermore, many of the larger eruptions have repercussions far beyond their immediate surroundings. The dense plumes of ash emitted by the Icelandic volcano Eyjafjallajökull in 2010 significantly disrupted air traffic in the North Atlantic for several weeks. On the order of 100,000 flights were cancelled, and 10 million passengers were affected. The airlines suffered revenue losses amounting to between 1.5 and 2.5 billion euros. “The need for reliable forecasts of volcanic eruptions – and prior assessment of the damage they can do – is becoming ever more urgent.”
‘The man-made volcano in the basement’
That is why many of the most dangerous volcanoes are virtually paved with sensors designed to detect changes in behavior that might signal an imminent eruption. For volcanic eruptions seldom occur out of the blue. They are usually the culmination of long drawn-out processes, during which increased activity is signaled by enhanced seismic activity, emission of gases or geomorphological changes. “Volcanoes send out signals, which geoscientists record with their various instruments. So we have a lot of data to work with,” Dingwell explains. “With the aid of experimental volcanology, we want to discover how to interpret these signals. And we want to learn more about what actually drives eruptions. We would like to understand the eruption mechanism in detail – the physical processes that take place in the vent immediately prior to, during and after the explosion. Above all, we want to quantify them.”
In other words, Dingwell and his colleagues are primarily interested in the build-up and explosive release of energy that powers eruptions. What physical processes are taking place in the instant when magma – a complex mixture of solid crystals, gas bubbles, and foaming molten rock at a temperature that may exceed 1000°C – abruptly shoots with enormous force from the vent of a volcano, and countless particles of fragmented rock are borne aloft, sometimes rising thousands of meters into the stratosphere?
With the aim of gaining insight into the real process, Dingwell and his coworkers try to reproduce the decisive instant in their downsized volcanoes in the basement of the Institute. One of the projects they work on is financed by a grant from the European Research Council. “What we do here is basic research,” says Dingwell. “We simulate volcanic eruptions – under controlled conditions of pressure and temperature, using rock samples collected from a wide range of volcanoes, whose precise composition and chemical evolution we carefully analyze – before, during and after each experiment.”
The model volcanoes have a room to themselves in the basement. They don’t look much like their natural counterparts, although their size is quite impressive in a laboratory context. Each consists of a dark-gray cylindrical vessel, topped by a stainless-steel cylinder about 3 m high and with a capacity equivalent to that of two rainwater barrels. The bottom chamber to which it is attached is made of a special steel alloy of the sort used in gas turbines and tank barrels, which can withstand the enormous pressures and high temperatures that develop during experiments. The whole assembly is mounted on a steel scaffold about 4 m high. Inserted between the two receptacles are so-called rupture disks – thin metal plates which are pierced by the force of the explosion.