Scientists measure the rate of ascent of volcanic magma

Plinian volcanic eruptions are notoriously destructive. These very powerful eruptions often occur after long periods of quiescence and are preceded by relatively short periods of seismic restiveness. Volcanoes that tend to show this kind of behaviour include Mount Vesuvius in Italy, Mt. Pinatubo in the Philippines and Mt. St. Helens in the USA. Professor Donald Dingwell of Ludwig-Maximilians-Universität (LMU) in Munich, together with Professor Jonathan Castro of the University of Orléans in France, has now been able experimentally to measure the speed with which molten rock rises during a Plinian eruption. The two scientists studied rocks that erupted from the volcano Chaitén in Southern Chile in May 2008. Their experimental analyses revealed that the magma must have ascended from the interior of the volcano to the surface within a period of only four hours. These results raise the disturbing prospect that it may not be practically possible to give adequate warning and carry out orderly evacuation procedures prior to this type of eruption.

The first description of a highly explosive volcanic eruption dates from the year 79 AD. In that year, the Roman author Pliny the Younger observed the famous eruption of Mount Vesuvius, which buried the city of Pompeii under enormous amounts of ash and pumice. Pliny's description led later students of volcanology to name eruptions of this type after him. Plinian volcanoes are characterized by long periods of quiescence, during which they show very little activity. Moreover, the rare eruptions are preceded by quite short bursts of tectonic activity, signalled only by minor earth tremors and increased emission of gas. During the build-up that precedes the eruption itself, magma rises to the surface within a very brief interval, and is expelled from the volcano at high pressure in a huge explosion.

More than a dozen Plinian volcanoes are found in the Andes of South America, yet scientists observed a typical Plinian eruption there only last year. On May 2nd 2008, the volcano Chaitén in Southern Chile suddenly began to spew large quantities of ash and rock fragments into the air. The erupted material eventually gave rise to an ash plume some 20 km high. The town of Chaitén, 10 km away, was covered by a layer of ash several centimeters thick and had to be evacuated. "This eruption was particularly noteworthy, because the volcano had been quiescent for over 9000 years", says Professor Donald Dingwell, Director of the Department of Earth and Environmental Sciences at LMU Munich. "The best estimates suggest that the last eruption took place in the year 7240 BC."

Together with the Jonathan Castro from the University of Orléans in France, Dingwell has now been able to calculate the velocity with which the volcanic material must have risen within the magma chamber. The researchers collected samples of pumice from the eruption debris, and these were then subjected to a series of laboratory analyses in Munich, while being heated to a temperature of 825 degrees centigrade at high pressure. After a certain time under these conditions, characteristic crystalline margins begin to develop around the feldspar crystals in the pumice. Dingwell and Castro systematically varied the temperature and pressure, and measured the time it took for these crystalline margins to grow. "The interesting thing is that we found none of these crystalline margins in the natural samples themselves", reports Dingwell. "From that we can conclude that the material must have risen so quickly that there was no time for them to form."

The researchers were surprised by the results of their analyses. Their calculations suggested that the rock fragments they had collected had ascended from the earth's interior to the crater floor in less than four hours. To accomplish this, the material must have risen at a rate of about one meter per second. "This figure is very disturbing, because it implies that a Plinian eruption can develop with astonishing speed", Dingwell points out. "In such a case, it would be well nigh impossible to give adequate warning of an impending eruption, in particular if the period of activity preceding it also happened to be very short." This was precisely what happened at Chaitén. The inhabitants of the town felt the first perceptible earthquakes on the evening of April 30th. The first ashfall arrived on the next day, and on May 2nd there was a violent eruption, followed by the appearance of a huge cloud of ash over the mountain.

"The problem with such short periods of heightened activity is that they may, but do not necessarily, forecast an eruption", explains Dingwell. "In the case of Chaitén we knew that we were dealing with a highly explosive volcano. What we did not know was what kind of activity would give notice of an impending eruption." Normally, patterns of volcanic activity are observed only locally, for instance by geophysicists who measure seismic waves, or by geochemists who analyse the gases emitted in the vicinity of a volcano. "Our study is something entirely new and complements the local observations by using a well-founded experimental and theoretical approach", says Dingwell. "In our view, this will become an important option in future investigations of the behaviour of volcanoes." (am)