This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. Along the coast of western Norway, and for thousands of square kilometers inland, the rock is gray banded gneiss, the sort of rock exposed on continents everywhere. Geologists have studied it for well over a century. Scattered through that gneiss are pods of a darker, denser rock, and some of the crystals in those pods have specks inside them too small to see without a microscope. Those specks can only have formed under pressures found about 100 kilometers down, far below the bottom of the crust, down inside the Earth's mantle. That demands an explanation, because continental crust is buoyant, which is the reason continents stand high and ocean floors sit low. If pushed down into the mantle, it ought to bob back up, or refuse to go at all. And yet here is a piece of it that went to mantle depths, stayed long enough for its minerals to recrystallize, and then came back up to the surface.

Let me define a couple of words that will come up in the podcast. The first is eclogite. When a rock is heated and compressed, its minerals rearrange themselves into whatever is stable under the new conditions. So the same starting material gives you one set of minerals at 10 kilometers down and quite a different set at 50. Take ordinary basalt deep enough and it reorganizes into an attractive-looking rock containing red garnet and a green pyroxene mineral called omphacite. That is eclogite. It is a dense rock, even denser than the mantle. In Norway, it occurs as pods and lenses within the gray gneiss.

The second is coesite. Coesite is a high-density polymorph of quartz, meaning it has the same chemical composition as quartz, silicon dioxide, but packed into a more compact crystal structure. Coesite forms only at pressures above about 27,000 atmospheres, equivalent to a depth of 90 to 100 kilometers. Metamorphism above the pressure at which quartz turns into coesite is called ultrahigh-pressure metamorphism.

Torgeir Andersen is professor of geology at the University of Oslo, and for several decades he has worked on the Western Gneiss Region of western Norway, the world's largest and best-exposed area of ultrahigh-pressure rocks. Torgeir Andersen, welcome to Geology Bites.

Torgeir Andersen: Thank you very much. It's a pleasure to be invited.

Oliver Strimpel: I said that geologists have been studying the rocks of western Norway for a long time, but coesite was only discovered in 1984. Why did it take so long, and how was it eventually discovered?

Torgeir Andersen: Well, it took a long time because all these ultrahigh-pressure minerals, like the coesite and the diamond as well, are tiny little inclusions sitting inside the other minerals, the eclogite-facies minerals that we all know quite a lot about, like the garnet and the omphacite and so on.

Oliver Strimpel: Just a word about the term facies, which is a set of minerals that forms together at a particular pressure and temperature. So eclogite facies means the conditions at which eclogite's minerals become the stable ones.

Torgeir Andersen: So it takes a very, very careful mineral study of a number of eclogites to actually find these small, magical ultrahigh-pressure minerals that define this ultrahigh-pressure metamorphism. So it took a long while, but it was discovered in western Norway, as it was in the Alps, at the same time.

Oliver Strimpel: And what was it that enabled them to distinguish it from quartz?

Torgeir Andersen: In order to actually prove that you have coesite, you need Raman spectroscopy, that is, checking exactly the length of the bonds in these ultrahigh-pressure minerals, because they have a tendency to back-react when the rocks are decompressed. So quite commonly the coesite has back-reacted to normal quartz, and during that process it expands a bit and exerts pressure on the rocks in which it sits. So it creates a radial crack system around these inclusions, which is often how you first suspect that there has been coesite. You see these strange radial cracks, and then occasionally a little part of the coesite is preserved. It has a slightly higher refractive index, so you can actually see it with a microscope and suspect that it is coesite, but then you need to do a Raman examination in order to actually confirm it. And actually, the same goes for the diamond: you also should have a proper Raman examination to have them confirmed.

Oliver Strimpel: So apart from coesite and diamond, which you've already mentioned, are there any other minerals whose appearance in a rock would unambiguously point to ultrahigh pressures that we're talking about?

Torgeir Andersen: Not really. The normal eclogite minerals, like garnet and omphacite, are still stable at ultrahigh-pressure conditions, and it's not proven that you have ultrahigh pressure if you have these two minerals that are required to be an eclogite. However, you can do chemical analysis of these minerals and try to calculate pressures and temperatures from equilibria between the different minerals, and quite commonly you will find pressures that could correspond to ultrahigh pressure. In order to have coesite in the rock in the first place, you need to have quartz, free SiO2, free quartz. If you don't have quartz in the rock to start with, you will not have coesite. And the same goes for diamond: you have to have carbon in the rock. So if you don't have carbon in the rock, you will never have diamond, even if you pressurize to 40 kilobars.

Oliver Strimpel: If the reactions that form these minerals only take place at these tremendously high pressures, why don't they go into a reverse when the rocks come back up to the surface and the pressure drops?

Torgeir Andersen: Well, as I already mentioned, they quite commonly go into retrograde metamorphic conditions. We mentioned just the radial cracks around the coesite. But I guess the reactions take a long time, which is why we're able to see the coesite at all, still surviving. One of the conditions that is required is that the decompression, that is, the exhumation of the rocks to lower pressures, goes fast. We often talk about isothermal decompression, that is, that the eclogites are decompressed very rapidly without cooling, so they retain the temperature they had at eclogite-facies conditions, or can even increase in temperature as they are approaching the surface. But that is really a precondition that is required: rapid exhumation.

Oliver Strimpel: So as I said in the introduction, crust is buoyant compared to the mantle. So how could such rocks find their way all the way down to the mantle?

Torgeir Andersen: Yeah. This is one of the main things that we have been working on in the Western Gneiss Region, because we have the outcrops necessary for understanding quite a lot of what has been going on, both during the burial to these extreme depths and extreme pressures and during the exhumation. In order for any eclogite to return back towards the surface, or even get exposed at the surface, they need to sit inside lighter rocks that have a buoyancy with respect to the mantle. The mantle has a density of something like 3.1 or 3.2 tons per cubic meter, whereas the eclogites can have up to 3.6. So they should sink. So they need a life belt in order not to sink, and the life belt is the granitic rocks in which they are sitting, which don't want to go into the deep mantle, because they will always have a buoyancy with respect to the mantle.

Oliver Strimpel: So what was it that dragged them down to those great depths in the first place, and if it was net buoyant?

Torgeir Andersen: Yeah, they have to be attached to a thick down-going slab during subduction. And this slab is most commonly an oceanic lithosphere, where the entire crust has a mafic composition. It has a higher average density than continental crust, and during subduction it can be completely eclogitized, in which case it will sink and drag down with it the continent to which it was connected in the first place. So it sinks down by this negative buoyancy of the oceanic plate that was attached to the continent. So that's the reason why you can drag it down.

Oliver Strimpel: I see. So it's dragged down by this heavy anchor, if you like, and it's pulled down behind the eclogitized slab. And in that case, why doesn't it continue subducting, going deeper and deeper without bobbing back up to the surface?

Torgeir Andersen: After some subduction of the continental crust, this will still be buoyant with respect to the mantle around it, and the down-going slab doesn't have the strength to pull this continental crust further down. So the slab, the subducting slab, breaks off or detaches from the continent. At that point, the continental part of the down-going slab will have a positive buoyancy with respect to the mantle, and the subduction movement will be reversed.

Oliver Strimpel: So after the slab breakup, then, how exactly does this bit of continental crust that now finds itself very buoyant and in the wrong place, essentially, what exactly happens to bring it to the surface?

Torgeir Andersen: What happens is then that it starts to rise along the former subduction interface with the upper plate. In addition, it has been heated up a bit, so in addition to the reverse movement, it can also flatten vertically. So you go from stretching in a vertical manner to flattening in a vertical manner. So the initial stage in this down-going slab is that it flattens and moves back up towards the surface. So it's an interplay between vertical shortening and reverse movement on the former subduction interface.

Oliver Strimpel: So it basically comes back up along the way it went down. But are there any other geological contexts in which rocks that have undergone such ultrahigh-pressure metamorphism can reach the surface?

Torgeir Andersen: Well, we all know about kimberlites, which come up in kimberlite pipes in a very, very, very fast way and bring with them high-pressure diamonds. This can also take place if you have an eclogitized lower crust that starts to melt and is affected by large-scale extensional tectonics and rifting. There is one place in the world where the youngest ultrahigh-pressure eclogites are found. It's in the Woodlark Rift system, north of Australia, Papua New Guinea, and the islands are called the D'Entrecasteaux Islands, and in these the youngest coesite-bearing eclogites have been found. They are five, six, maybe a bit more, million years old, and they returned from some depth that is capable of producing coesite to the surface. So that means that for the exhumation velocity, we're talking centimeters per year rather than millimeters per year, which is a more common velocity for this exhumation. So that's an extreme variety, and I don't think there is any other example that is known of such a diapiric type of exhumation mechanism for ultrahigh-pressure eclogites.

Oliver Strimpel: So the diapir is a convection process, is it, that brings it up to the surface?

Torgeir Andersen: If you have a continental crust, and you bring it down and you start to melt it, of course granitic melt compositions will have a very high buoyancy at the depths we're talking about. And it's well known that granite can rise diapirically into its overburden. And I guess this is a similar type of mechanism in the case of the D'Entrecasteaux area. In addition, it sits in a big extensional rift zone, so there is a combination of tectonic stretching and maybe also this buoyancy.

Oliver Strimpel: You've carried out a great deal of research on the Western Gneiss Region of Norway. What does it look like today, and can we see lots and lots of eclogite there?

Torgeir Andersen: There are hundreds of eclogites in the Western Gneiss Region, of various sizes, from several kilometers in size to small lenses. And some people are still doing detailed work, because it's one of the key areas to really figure out what is going on at depth under big mountain belts, like the Caledonides were and like the Himalayas are today.

Oliver Strimpel: Eclogites form in rocks of basaltic composition, and we already said that the Western Gneiss Region is continental crust. So where did the basalt that's inside this continental crust come from?

Torgeir Andersen: Every rock in the Western Gneiss Region is Proterozoic in age. They go from about 1,860 million years down to 900 for the youngest. During this long history in the Proterozoic, there have been a number of magmatic events. The main rocks are granodiorites and granites, but there are also quite big gabbroic intrusions and dolerite intrusions, mafic dikes, and a large number of these mafic rocks have now become eclogites.

Oliver Strimpel: So in the Western Gneiss Region, can we tell from the structural geology how the subduction and the subsequent exhumation took place? Can we actually see faults along which the ultrahigh-pressure rocks return to the surface?

Torgeir Andersen: The reason we can look at all of these rocks in the Western Gneiss Region is primarily due to a very large extensional detachment. This is a normal-type shear zone, normal meaning that the hanging wall moves down with respect to the footwall, and the detachment itself has a displacement of many, many tens of kilometers, maybe 100 to 150 kilometers. And this normal displacement, where the Caledonian nappes and rocks in the hanging wall of the detachment have moved back towards the west with respect to the footwall of the detachment, which has moved relatively eastwards, constitutes a deformation zone that can be three to four, five kilometers in thickness, all characterized by very strongly deformed rocks that we refer to as mylonites. In this case, they are extensional mylonites, because we can identify a number of structures inside the mylonites, from outcrop scale to microscope scale, that give us information about the relative movement, top to the west, along this thick zone of mylonite. So the shear strain along this detachment zone is remarkably high. For every meter of thickness of mylonite, you can have up to 20 meters of down-to-the-west movement. So when you add that up, it becomes many, many kilometers, or 100 kilometers or thereabouts. So we can quantify the movement by measuring the strain in the detachment, and we need to have this amount of detachment to bring the rocks back from these big depths.

So that's the evidence for how the rocks can return back to the surface. And then, of course, it's 400 million years since the eclogites formed, and during that period of time there's been a lot of erosion as well. So you remove the top by erosion, but you cannot erode away the rocks that are covering the eclogites completely. So you need this extensional tectonics in order to have exhumation.

Oliver Strimpel: You said the ultrahigh-pressure metamorphism took place around 400 million years ago, and across the region, the pressures and temperatures recorded in the rocks rise steadily from east to west. Was that gradient always thought to be Caledonian?

Torgeir Andersen: In the Western Gneiss Region, we have this beautiful gradient that was figured out already 40 years ago by the pioneers. At that time, they didn't have ages of the eclogites, so they thought that this gradient was a Precambrian phenomenon, because the rocks are Precambrian. So it was only when radiometric dating showed that the eclogite-facies metamorphic rocks were formed during the Caledonian orogeny, rather than at some Precambrian orogeny, that we realized that this was a phenomenon related to the Caledonides and not some Precambrian orogeny.

Oliver Strimpel: Let's talk a little bit about the broader tectonic context that resulted in the rocks that we see in the Western Gneiss Region of Norway. Here the Caledonian orogeny involved the collision of Laurentia to the west and Baltica to the east, closing the Iapetus Ocean between them. So what are these ultrahigh-pressure rocks we see? Were they part of Laurentia or part of Baltica?

Torgeir Andersen: The Western Gneiss Complex, the gneisses and the mafic rocks sitting in them, were definitely part of Baltica. All the age characteristics are typical for the rest of Baltica.

Oliver Strimpel: I'm trying to piece together a picture of how this collision and the subduction zone took place. So if the ultrahigh-pressure rocks are Baltican rocks, were they subducting towards the west under Laurentia, or how did that happen?

Torgeir Andersen: Everybody agrees that during closure of the Iapetus Ocean, the subduction direction has been almost always towards the west.

Oliver Strimpel: Do we see ultrahigh-pressure metamorphic rocks associated with other former or presently active subduction zones around the world?

Torgeir Andersen: Yes. Almost all of these ultrahigh-pressure rocks are somehow related to subduction of oceanic and in part continental crust. This is the only way that it's possible to get rocks down to the pressures required.

Oliver Strimpel: So we can see ultrahigh-pressure rocks in yet other former subduction zones?

Torgeir Andersen: Absolutely. You have them in Newfoundland. They occur along the Variscan systems in Central Europe, with both coesite and diamond occurrences in the Black Forest areas, and in Europe and Russia, Siberia. So I don't know any areas in the world where ultrahigh-pressure rocks have not been formed in a subduction environment.

Oliver Strimpel: So in all these places, we've taken the pressures recorded in the rocks as a measure of how deep they went. I'm wondering whether that assumption is always valid. I mean, couldn't the pressure on deeply buried rock vary for reasons other than the weight of the rock above it?

Torgeir Andersen: Yes. Numerical modelers have shown that you can create pressures that are larger or smaller than the pressure from the weight of the overburden of rocks.

Oliver Strimpel: Let me just explain how that could happen. In a liquid, the pressure at any depth is simply the weight of what lies above. But solid rock has strength, and strong rock can hold a pocket inside it at a pressure above the weight of the overburden, rather like the lid of a pressure cooker. The extra pressure can come from tectonic squeezing, or from rock that expands as it reacts or melts in a sealed pocket.

Torgeir Andersen: Many people don't like at all the idea of overpressures. They say rocks are not strong enough to sustain overpressure, whereas others have used numerical models and physical parameters to argue that overpressures can exist. I think personally that it is possible to create overpressures, and I have been coauthor on at least one paper where we have argued that that happened in the Western Gneiss Region, in order to explain an occurrence of diamond. However, when you have a large area like the Western Gneiss Region, we can map out a gradient of metamorphism that gives increasing temperatures and pressures as you go from east to west: 10 kilobars in the east and 30 kilobars in the west; 500 degrees in the east, 650, 700 degrees in the west. So there is a definite metamorphic gradient. In order to have a big pressure increase, you need to have an increasing thickness of rocks.

Oliver Strimpel: So the gradient across the region is a matter of depth. Where does the overpressure come in?

Torgeir Andersen: If you then suddenly get a diamond that is totally out of this gradient, and you cannot find a structure between the normal-gradient rock and the diamond-bearing rock that can explain juxtaposition from different depths, then you either have to believe that all the rocks went to the pressure of the diamond, and then only the little pod with diamond in it didn't back-react, or you have to have an alternative explanation. That is when the overpressure hypothesis becomes a possibility.

Oliver Strimpel: So if overpressure can play a significant role, how much does that change the reconstruction of tectonic history that we've been discussing? I mean, can we trust the depths that we've been inferring?

Torgeir Andersen: It doesn't change the subduction story very much. The subduction has happened, obviously, and it makes it easier to understand variations in pressure. As long as we have a gradient, as we have in the Western Gneiss Region, we have a good way of explaining things. But when you have a little isolated outcrop with a diamond or a coesite, and it doesn't really match with its surroundings, then it is more problematic.

Oliver Strimpel: So the subduction itself isn't really in question. The pressure differences involved are several kilobars, equivalent to some tens of kilometers of rock. That makes a difference for an isolated outcrop, but not really for the big picture. And so these rocks really did go to those great depths.

I gather that although you're retired, you're still working on the Western Gneiss Region of Norway, but at the surface rather than at depth. The same great faults, or detachments, that carried these rocks back up from the depths also opened up basins on the surface. So while the rocks were rising from 100 kilometers down, rivers were filling those basins with sand and gravel, right on top of the faults that were bringing them up. And you're trying to reconstruct what the landscape looked like at the time, which would be in the Devonian, around 400 million years ago.

Torgeir Andersen: Yeah, in the Devonian, mainly. So we are working with that at the moment, but it is a bit premature to make it public what exactly we are working on. I think it's very exciting.

Oliver Strimpel: Torgeir Andersen, thank you very much.

Torgeir Andersen: It's been a pleasure.

Oliver Strimpel: To see pictures and illustrations that support this podcast, go to geologybites.com, where you'll also find a subject matter index of all the episodes. There you can also give me feedback, which I welcome, as well as sign up to get my emails about new episodes.