Musing over the mantle

16 June 2020

Interview with 

Huw Davies, Cardiff University

Share

Since the time of Mohorovičić we’ve learned a lot more about the Earth’s interior. Geologist Huw Davies from Cardiff University is setting up a project to learn a lot more about the mantle, and he spoke to Katie Haylor, explaining firstly what the crust is.

Huw - Well, the crust is a silicate rock, basically rock. And yeah, primarily, has a slightly higher level of silica than maybe some other rocks.

Katie - And if it were possible, I don't think it is at the moment, to drill down into the mantle from the crust. What would that look like? What's the environment like?

Huw - So the mineralogy would change. So the composition would change, simply. So basically the crust really results from melts produced from the mantle. So elements that like to be in melts have ended up being in our crust. So one thing we would see is clearly slightly different minerals as we got into the mantle, and that's what leads to this different seismic velocity we just heard about, but also we will be getting hotter. And that's part of the reason why it's actually very difficult to drill into the mantle because it becomes a very hot environment. And basically our drills can't sustain that very well.

Katie - How far down has anybody got?

Huw - So there's two aspects of that. So the deepest hole is in the Kola Peninsula in Russia, and they've drilled 12 kilometres, so that's in continental crust. But there the reason they've been able to drill so deep is because it's a very, very cold part of the earth. And actually that's in the continental crust and actually to get to the mantle they probably have to drill another 30 odd kilometres. So where we've drilled closer to the mantle is in the oceanic crust, which is a lot thinner, but there we've only managed to drill about a kilometre or two, and we would need to drill another four or five kilometres to reach the mantle. So the only places where we've seen the mantle is where tectonic processes have brought it up to the surface, or we've had little fragments brought up by volcanic eruptions.

Katie - I'm glad you mentioned that because where do tectonic plates come into this? How do we end up with these really dramatic geological events like earthquakes and volcanoes?

Huw - Well it's kind of strange in a sense. The surface of the earth as we discovered sort of, well, we sort of got to the argument sort of in the fifties, but really was won in the sixties, of the surface of the earth is broken up into these tectonic plates. So basically large regions don't suffer much deformation. And then the movement is all occurring at the boundary of these plates and that can occur in three different ways. They can be moving side by side, like in the San Andres Fault. They can be approaching each other, converging, like in Japan where one plate goes under another. Or they can be diverging separately in like in the mid-ocean ridges deep in the ocean where we get magmas.

Katie - Can you break that down a little bit in terms of what leads to quakes and what leads to volcanoes?

Huw - So when the plates rub against each other, in the case of side by side, so the plates are basically rubbing against each other. And the earthquake is because you get a stick-slip mechanism. So the plates kind of stick, but the forces of the mantle are moving the plates, and then at some point, the fault can't sustain the force any longer and then it jumps and that would be the earthquake event. And basically the movement will have caught up with the movement in the mantle, and then it repeats itself. But the biggest earthquakes of all are in the convergence zone. So these are where the plate goes down beneath each other, for example, in Japan, in Indonesia and beneath South America, for example, and it's the same idea, basically, again, the plate sticks and then it moves suddenly to catch up with the forces that was always wanting to pull it down.

Katie - So in a volcano, we get an eruption and magma. What's going on there compared to what we were just talking about?

Huw - Most magmatism occurs beneath your ocean floor, where the plates are moving apart. So we get hot rock coming close to the surface. And interestingly, as it gets closer and closer to the surface, the pressure on it gets less. And interestingly, the melting point of rocks gets less as the pressure decreases. So at some point the rocks are hot enough, but not under enough pressure. So they melt and they produce in fact, the oceanic crust, but those typically are beneath the ocean floor so we don't see them very much. But the case where the plates are colliding against each other, we also get volcanoes. And these are the dramatic volcanoes, like in Japan, in the Andes, in Indonesia. And in this case, what we've got is the ocean floor, which is as a converging plate, is jumping stick-slip, as earthquakes, gets down into the interior, and the water of the ocean that has penetrated into the crust - in fact, when it formed the lavas, when it was spreading way back at the ridge - so all this gets carried down and the water gets carried down inside the earth, and then it gets released, and water also reduces the melting point of the mantle, and that's how we get melts in that part of the world. And that's why actually those magmas tend to be the most explosive because that's actually that water forming bubbles as the pressure's released exactly the same idea as we get with a Coke bottle or something, when we open the top and loosen the pressure. And then the third scenario is where we get just purely hot rock coming up towards the surface, just the heat sort of melts the rock. And that's a place like Hawaii is an example of that region.

Katie - I'm pretty sure you said magmatism, and magmatism-magnetism, I guess you gotta be a bit careful between those two!

Huw - Yes. Magmatism coming from magma, magnetism of course is to do with magnets yes.

Katie - How much do we understand about the mantle then? Where does the knowledge gap exist? Because you're about to start a pretty big project on the mantle, is that right?

Huw - Yes. So we understand a fair bit given that it actually isn't something we can really get our hands on very easily. And a lot of it comes from seismology, and we just heard earlier about how seismology helped to tell us the difference between the crust and the mantle. Well today, we kind of know the onion-peel structure of the earth very well, all the way down to the core-mantle boundary, but we can take that further and we can now do CAT scanning type ideas using the seismic waves. So we have some sense of the 3D structure. Resolution is relatively poor. We have some sense of the down-going movement, where the plates go down, but the bit that's less well understood is, clearly if a thing goes down something has to come back up and it's the up-flow that we don't understand as well. And that's going to be the primary focus of our project.

Katie - So how do you go about trying to look at that then?

Huw - The centre of the project we can say will be a model of the dynamics. So the inside of the mantle creeps sort of at the same rate as our fingernails grow. So we don't see them when we look at them, but we know they've grown after a few months. So that's the same thing for the mantle. We can model it on a geological timescale like a flowing liquid. So we'll have this model and we'll apply the plate pushing histories that we know to the surface. And then it will make predictions for example of how the structure would be at present day, which we can compare with the seismology. And we'll also have looking at chemistry of the magmas that come out and they'll give us different constraints. And we will also look at magnetic signatures of rocks at the earth surface, which will tell us something about the old magnetic field, which tells us about the core, which the mantle sort of puts a constraint on. So we'll bring this model of how the earth's moved, and then we'll be constraining it with all these other different techniques and hopefully understanding how the Earth's flown.

Comments

Add a comment