Volcanoes - Plumbing the depths

A new understanding of what goes on under volcanoes
(Aug 14th 2003 - From The Economist print edition)

Dr. Keith Putirka HOW is a volcano "plumbed in" to the Earth? Does Mount Pinatubo have any U-bends in its pipe-work and how many twists and turns are there in the conduits underlying Mount Etna’s fireworks? The answer is fundamental to understanding how volcanoes work, and could be crucial to predicting when a particular volcano is next going to blow.

The traditional view is of a volcano being fed by a magma chamber that lies at the base of the Earth’s crust. A few pipes lead up to the surface, each ending in a volcanic vent. Magma (as liquid rock is known before it erupts) periodically spurts out of these pipes. But that idea is being challenged by a group of geologists who argue for a more complex plumbing system of "magma-mush" columns--a network of magma-filled cracks that finger their way up through the crust. Among these geologists are Keith Putirka, of California State University, and Christopher Condit, of the University of Massachusetts, Amherst. And they think they have some evidence to back the idea up.

The answer lies in the magma itself. Dr Putirka and Dr Condit reckon that lava (as magma is called once it has erupted on to the surface) from a particular volcano will have been altered chemically by its journey through the crust in ways that betray the details of that journey. In a paper in this month’s Geology, they apply their method to lava from the Springerville volcanic field in Arizona.

Dr Putirka and Dr Condit are especially interested in the composition of a mineral called clinopyroxene. When magma pauses on its journey through the crust, it begins to cool. Crystals, including crystals of clinopyroxene, then start to form in it. The exact composition of the clinopyroxene, though, depends on the pressure under which it is forming. And that, in turn, depends on the depth. In particular, more pressure causes the mineral to be richer in sodium. So the sodium in clinopyroxene acts as a barometer that tells you at what pressure (and hence depth) the magma paused on its way to the surface.

At Springerville, Dr Putirka and Dr Condit discovered two levels in the crust where magma had paused and pooled before bursting through to the surface. The first was in the middle of the crust, at a depth of around 26km (16 miles), while the second was much nearer the surface--12km down. Crucially, they found no evidence for magma pooling at the base of the crust (a depth of around 40km at Springerville), suggesting the traditional model is plain wrong. And further support for the magma-mush-column idea has come from Bruce Marsh, of Johns Hopkins University in Baltimore, who has been studying an exposed section of Antarctic crust that reveals a cross section of pipes and pools within an ancient volcano.

Since the depths at which magma forms pools under a volcano are likely to influence the type of volcanic eruption at the surface, this information could help to predict a volcano’s behaviour. Though little is known for sure, this being such a recent discovery, it seems likely that some plumbing structures will prove prone to producing infrequent, explosive eruptions, while others yield a steady trickle of lava at the surface.

That, of course, needs investigation. And Dr Putirka and Dr Condit are now doing so by studying other volcanically active areas. If they do uncover a relationship between a volcano’s plumbing system and the size and timing of its eruption, it could lead to more accurate long-range forecasts for volcanoes.