Podcast: Watching volcanoes

People watching a volcano erupting

18 May 2009 by Richard Hollingham

Predicting when a volcano will blow has to be one of the holy grails of geology. Geologists have got much better at this than they used to be by constantly monitoring active volcanoes and studying ancient ones to figure out how they work.

Science writer and broadcaster Richard Hollingham visits Dr Sue Loughlin and Dr Kathryn Goodenough at the British Geological Survey (BGS) in Edinburgh to talk to them about their work.

Sue, a former director of the Montserrat Volcano Observatory, which monitors Soufrière Hills volcano, explains how so-called 'earthquake swarms' underneath this volcano made eruptions much more predictable in the 1990s than they are now. Whilst Kathryn studies rocks from ancient volcanic eruptions - including the extinct volcanoes of Glencoe and Edinburgh - to build up a picture of how volcanoes work.

To assist those who find text-based content more accessible than audio, a transcript of this recording is available below.

Female voice: The Planet Earth podcast, presented by Richard Hollingham.

Richard Hollingham: This is the sound of a volcano.

[loud knocking sound]

I'm with Sue Loughlin at the British Geological Survey in Edinburgh. Sue, what on earth is this we're listening to?

Sue Loughlin: Well, we're listening to earthquakes at the Soufrière Hills volcano, Montserrat, in the West Indies. We're listening to earthquakes building up, getting more rapid and close together before a lava dome collapse in 1997.

[loud knocking sounds increase rapidly in tempo, followed by a continuous noise that fades away]

This is seismic energy that's been converted to a sound wave, but each of those knocks that you heard is a single earthquake. It's what we call an 'earthquake swarm', when you get many, many earthquakes coming very rapidly one after enough, and getting closer and closer together as the pressure builds up inside the volcano.

Richard Hollingham: Now we're hearing this in just a few seconds, that was over a period of, what, hours?

Sue Loughlin: This particular one was over probably a couple of hours. But, yes, that was speeded up 100 times from the real life event.

Richard Hollingham: Does this mean you can predict when the volcano is going to erupt, because these knocks are getting closer and closer and closer together?

Sue Loughlin: Well these particular earthquakes, called hybrid earthquakes, we detected a lot of them in Montserrat during the course of the ongoing eruption at the volcano. And they were very useful in the late 90s, because they often did have cycles of activity, and as they got more numerous and closer together, as you just heard, that usually said the pressure was building and something was going to happen. But unfortunately these days those hybrid earthquakes have largely disappeared and we don't get them anymore, so it's much more difficult to forecast what the volcano is likely to do.

Richard Hollingham: So what is going on under... under the earth... under the mountain of the volcano, if you like, prior to the eruption? What's happening?

Sue Loughlin: Well what we heard there was basically a build-up in pressure, and this is all driven from a magma chamber deep underground, probably several kilometres deep underground, and magma is being supplied to that chamber, and as the pressure builds in the chamber it is driven up towards the surface. And again, if the pressure is strong enough just below the surface, then it breaks out... the magma breaks out as lava onto the surface. And in Montserrat this builds a lava dome, because the lava is very thick and sticky and it can't flow. So the whole eruption process is driven from a magma chamber at depth.

Richard Hollingham: So what do you do now, then? Your plan to be able to predict volcano eruptions from this knocking, that's out of the window now. What do you do know? How do you go about looking for signs, perhaps, of when a volcano might erupt?

Sue Loughlin: Well there are many different ways, but one way you can try to find out what's happening in the magma chamber is to take blocks of the lava, for example, and study it in a laboratory, to look at signs and clues for what's going on in the magma chamber. You can use seismic experiments; there was just one recently in Montserrat to try and locate the magma chamber; find out where it is, how big it is. Then you can have some kind of a guess to how long this eruption might go on for, cause currently there's no sign of an end. But most of what we know of magma chambers has come from study of old volcanoes, where the magma chamber is actually exposed at the surface and geologists can study it. So it's really a combination of looking at what's happening now at an active volcano and also taking what we know about how magma chambers work from the study of old volcanoes, putting that all together, to come up with models of how we think this volcano, and other volcanoes like it, are operating.

[Crunching footsteps]

Richard Hollingham (out of breath): Fortunately, if you work in Edinburgh, you don't have to go very far to find a volcano. I've come up onto the hill above the British Geological Survey with Kathryn Goodenough. Set the scene for me here, because the view is spectacular, I can see Edinburgh Castle, there's Arthur's Seat and then behind us there are more mountains... um, we're looking at volcanoes?

Kathryn Goodenough: That's exactly right. Because here at Edinburgh we've got seven hills and all of those seven hills are volcanoes. Ancient volcanoes. And then all around us as well - we've got a good view today - we can see even more volcanoes, out and on the coast to the east, away to the north and the south, the hills of Edinburgh, the Pentlands. And these volcanoes are hundreds of millions of years old.

Richard Hollingham: So the ground... I was going to say the earth, but it's rock solid... under our feet, we're on a volcano?

Kathryn Goodenough: Exactly, what we're standing on here are volcanic rocks, and these were actually erupted from one of those ancient volcanoes, maybe 400 million years ago, or thereabouts. So here we're actually standing on the remains of volcanoes that don't look that different now from how they looked all that time ago. But if you go further north into the Highlands, then you can see much more eroded volcanoes, and up there we can actually see the roots of the volcano - the magma chamber.

Richard Hollingham: You've brought with you a piece of rock, now this is from, what, Glencoe? And it's got stripes, stripes within the rock, almost as if lots of different bits of rock have been sandwiched together.

Kathryn Goodenough: That's right. And this rocks interesting because Glencoe is a massive volcano that was erupting something like 400 million years ago and it's one of the best examples in Scotland where we can see the deposits from pyroclastic flows. Now pyroclastic flows are these incredibly hot flows, or columns almost, of ash and rock fragments, and so on, that you get when a volcano explodes, like Montserrat for instance, or indeed the famous 1980 eruption of Mount St Helens produced pyroclastic flows. And pyroclastic flows are devastating, they destroy everything in their paths, so it's something we really need to try and understand more about. Now, at Glencoe, we can actually go and study the deposits of these ancient pyroclastic flows.

Richard Hollingham: And you can relate those to an active volcano today? Even though these happened millions of years ago?

Kathryn Goodenough: That's right, because essentially the same processes tend to form the same types of rocks. And in Glencoe we can go and look at those deposits of the ancient pyroclastic flows, we can look at what happened in the volcano before they formed, and so on, so we can understand more about the history that led up to them. Whereas, perhaps somewhere like Montserrat today, it's quite hard to go and look at anything other than the surface of these deposits from the volcano.

Female voice: This podcast has been brought to you by the Natural Environment Research Council.