This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. Scattered around the globe, within both oceanic and continental plates, are some massive accumulations of igneous rocks called large igneous provinces. What causes such huge volumes of magma to flow onto the Earth's surface? And what impact does their emplacement have on the plates where they erupt, on the atmosphere and climate, and on life? Richard Ernst has been studying all aspects of large igneous provinces for nearly three decades and has written the definitive textbook on the subject. He is scientist in residence at Carleton University in Ottawa. Richard Ernst, welcome to Geology Bites.

Richard Ernst: Thank you, Oliver. It's great to be here.

Oliver Strimpel: There are many places where volcanism has produced a lot of basalt on the surface of the Earth, not least mid-ocean spreading ridges and volcanic arcs. Is there an accepted definition of a large igneous province that distinguishes them from such other forms of volcanism?

Richard Ernst: As originally defined in 1990, the basic aspect was a large-volume intraplate event, typically with a short duration, and the minimum size was conceived as around 100,000 cubic kilometers. The intraplate setting distinguishes them from ordinary seafloor spreading and collision-related activity, and they can be of very, very short duration, oftentimes less than a million years, or they can be multiple pulses over a longer period of time.

Oliver Strimpel: But basically, it's the bulk of extruded magma where the definition lies, right?

Richard Ernst: That's the main thing. That's how they were originally recognized: it was these huge flood basalts observed in many parts of the world that could be kilometers thick of basaltic lava, with very little interflow sediment. So it was really recognized that these were a unique sort of thing, and then people took this to the oceanic realm and recognized similar kinds of thick accumulations of extrusive material.

Oliver Strimpel: Can you describe the structure of a large igneous province?

Richard Ernst: At the surface, you can see the huge flood basalts that might extend over many, many hundreds of kilometers, covering up to a million or more square kilometers. But as the flood basalts are eroded, you see more of the plumbing system. You see giant radiating dike swarms that can have a radius of up to 2,000 kilometers or more, and giant circular swarms that can have a radius of almost 1,000 kilometers. You also have massive sills, dolerite sill provinces associated with sedimentary basins. You also have the important group of mafic–ultramafic intrusive complexes, which are also an important part of the plumbing system. And then, as you go deeper, you get down to the magmatic underplate that's often present, where you might have up to 10 million cubic kilometers or more of material accreted to the base of the crust as mafic to ultramafic material. So the plumbing system is really quite extensive, and this indicates that these events are really even more voluminous than we realize from just looking at the extruded flood basalts.

Oliver Strimpel: But if I was standing on top of a large igneous province, on a continental flood basalt, would I recognize it from the plumbing system, or would I just see endless thick sheets of successive lava flows?

Richard Ernst: Yeah, that's the way you would recognize it. You'd see yourself sitting on top of basaltic volcanism as far as the eye could see: Hawaiian-style volcanism, but on steroids. If you have windows through it where the basalts were eroded, then you might see some of the dike swarm component and the other plumbing system components.

Oliver Strimpel: I have been to some of these large igneous provinces, for example on the western coast of Greenland and also in the Deccan. And when it's eroded through by a gully or something, you can just see layer upon layer upon layer, endless layers of successive lava flows.

Richard Ernst: Yeah, dramatically so. And even in the ones that have been deeply eroded, like the Mackenzie event of northern Canada, where the dike swarm component goes out to 2,000 kilometers, in the focal region of this overall radiating swarm you have the remnant of what must originally have been a very, very extensive flood basalt sequence, the Coppermine volcanics, a couple of kilometers thick or so.

Oliver Strimpel: What are the best-known large igneous provinces? Where are they, and roughly how many have been recognized?

Richard Ernst: One of the most dramatic is the Siberian Traps large igneous province, which extends over about 5 million square kilometers in Siberia. Another hugely dramatic one is the Central Atlantic Magmatic Province, at 201 million years ago. So those are two well-known ones. You mentioned the Deccan of India as well, at 66 million years; it is widespread in India. Those are the young ones, and there are lots more young ones. But also, as you go back through time and you look at the plumbing system component, we can trace them all the way back into the Archean, and up through the Proterozoic from about 2.5 billion years ago. From the end of the Archean, we have them occurring every 20 or 30 million years somewhere on the planet. And then there are the oceanic plateaus, which are only visible in the last 200 million years, where we still have current oceanic crust. But we know the oceanic plateau component continues back through time; we can see them as remnants in collisional zones, oceanic plateaus from closed oceans. If you add those in, we suspect that the rate of large igneous provinces back through time, back to about 2.5 billion years ago, might be as high as one every 15 million years or so. And then in the Archean, we're doing a new compilation based on all the new geochronology, et cetera, and it's looking like the rate is going to be fairly similar.

Oliver Strimpel: So you're saying that, really, all the way back to the Archean, we're talking about one of these being emplaced roughly every 15 million years?

Richard Ernst: Potentially as often as that, and certainly every 20 to 30 million years, based on the continental large igneous province record. To my mind, we talk about plate tectonics a lot, and it's a dominant mechanism on Earth, but this whole other process, this whole other voluminous magma production feature of Earth, is also incredibly important.

Oliver Strimpel: Could you talk about the geochemical makeup? Being basaltic, we're talking fundamentally about mafic composition, but there's also, as I understand it, a variation, between ultramafic at the one end and also some fairly silicic components of these large igneous provinces.

Richard Ernst: That is certainly true for the continental examples, where it's mainly mafic with some ultramafic component, but you can also have some more granitic compositions. As magma crystallizes, the residual liquid becomes more granitic, through fractionation of the magma. So you can generate some amounts of granitic-type magma from the mafic magma. But the other really fascinating component of continental large igneous provinces is the potential for the thermal source to cause massive partial melting of the lower mafic crust. If you melt a basaltic lower crust, you produce a granitic melt, and so you can get massive provinces of granitic melt associated with large igneous provinces, from the same heat source. These massive provinces are interesting because they can have a remarkable scale, 100,000 square kilometers or more. It's well known that you produce a lot of granitic-type magmatism in collisional environments associated with subduction, and as you go back in time, it can be hard to tell what setting you're in. So when you have a large granitic magmatic province of some age, is it evidence of a plate tectonic collisional zone, or is it an example of a large igneous province producing silicic magmatism?

Oliver Strimpel: Can you give an example of one of those?

Richard Ernst: The Karoo large igneous province, which is a massive event across southern Africa into Antarctica, and there's a felsic equivalent of it in southern South America, the Chon Aike event.

Oliver Strimpel: How long does it take to form a particular large igneous province? Or, to put it another way, do they have a typical magma extrusion rate?

Richard Ernst: The geochronology on the Siberian Traps event indicates that this 5-million-cubic-kilometer-or-more event, not including any underplate component, which would probably make it much larger, all came out within less than a million years. For the Central Atlantic Magmatic Province, which covers some 10 million square kilometers, uranium–lead dating suggests that most of it must again have come out within a million years or so, an incredibly short period.

Oliver Strimpel: Where exactly is the Central Atlantic Magmatic Province today?

Richard Ernst: The center of activity, from the North American perspective, is in Florida. From the African perspective, it would be the corner of Africa: if you close the Atlantic Ocean, you tuck it in adjacent to Florida there, and that's the center. But you have magmatism along the entire coast, all the way up to Newfoundland on the Canadian side. It's also extensive down into Brazil, into the Amazon Basin, and then into West Africa, all the way up into Mauritania, Mali, Morocco and Algeria, and up into the Iberian Peninsula as well. So it's a massive event, and everywhere we see it and date it, it seems to have an age of 201 plus or minus one million years or so.

However, there are also events that come out in multiple pulses that might span tens of millions of years. One of the important ones is the North Atlantic large igneous province, which is on Greenland, the units in Greenland you were talking about, and also in the UK, and offshore as well. The North Atlantic Igneous Province has a main pulse at 62 to 58 million years, and then a second pulse with the onset of rifting in the North Atlantic at 55. And Iceland, right in the middle of the North Atlantic, which is active today, represents the continuation of the North Atlantic large igneous province, at a much lower volume and flux rate. What happens many times is that you have these large igneous provinces with a huge volume of magma in a short period of time, and then there's protracted magmatism at a much lower pace for another 100 to 200 million years. So Iceland has pulses at 60 million years, then a second pulse at 55. Other events also have multiple pulses, oftentimes with the second pulse associated with the onset of rifting and continental breakup.

Oliver Strimpel: Are they unique to our planet?

Richard Ernst: They are not unique to our planet. On Venus, we see just a wealth of what we think of as large igneous provinces. On Venus, there's no plate tectonics, and so we see huge flood basalt events, lava flows continuing for 1,000 kilometers or more. We've also recently discovered huge dike swarms on Venus. Venus has no erosion and no plate tectonics. The lack of erosion is important because we see the surface expression of lava flows; we also see the surface expression of dikes laterally. We think of dikes moving up, and they do, but they can also move sideways in the crust for long, long distances. We know that from Earth, and we see it demonstrated vividly on Venus, where we can trace them for long distances, as swarms of dikes radiating for thousands of kilometers. We just recently traced a single one for almost 4,000 kilometers. And because of the absence of plate tectonics, it really lends support to a mantle plume model. Let me also mention Mars. Mars has graben systems, interpreted to be dike swarms 3,000 or 4,000 kilometers long, associated with and radiating from the area of protracted, plume-related volcanism called the Tharsis region, and also from another region, the Elysium region. So there are abundant radiating graben systems on Mars that are also interpreted to represent huge, laterally emplaced mafic dike swarms, associated with what we would call large igneous provinces on Earth.

Oliver Strimpel: Let's talk about what might generate the prodigious volumes of lava that form these large igneous provinces. I understand that there are two categories of model, each of them providing a mechanism that generates anomalously hot mantle material at shallow levels. The first involves shallow processes, by which we mean crustal and upper mantle processes, and the second invokes the deep mantle, with the arrival of a hot mantle plume. Let's start with the former explanations.

Richard Ernst: There are a number of shallow mechanisms. The lithospheric portion of the surface plate can drop off and descend into the mantle; that's a process called delamination. If so, then hotter material from the mantle below can rise to a shallower level and produce large volumes of melt. It's also been well recognized that with rifting extension, as you thin the plate, you also bring hotter mantle material to a shallower level, and that can also cause partial melting. There's another class of shallow models that links things to the supercontinent cycle. We know that the continents come together and break apart and come together, and that's happened several times in Earth history, the most recent time being Pangaea, which was together over the last few hundred million years and started to break apart around 200 million years ago. When you assemble all these cratons together, there's a region of mantle that is covered by thicker crust, by granitic crust, and it represents some sort of thermal blanket. So the temperature can then build up in the mantle underneath, potentially enough to generate melting.

Oliver Strimpel: I'm reminded of the episode of Geology Bites that we did with Peter Molnar, where he talked about the foundering of a keel below Tibet as an explanation of why it's so high today. And I think that is what you're referring to by delamination of the lithosphere: a heavier portion below sinking into the asthenospheric mantle below. And similarly, Damian Nance, in the last episode, also talked about thermal blanketing, as one of the ways in which you could get extra heating that would then lead to the demise of a supercontinent. So let's talk about the second category of model, namely the ones that involve the deep mantle and the mantle plume.

Richard Ernst: You have a plume that rises, and the top of the plume grows as it rises, because it is being filled from below faster than it can push its way up through the mantle. So the size of the plume head is predictable by the modeling. You get a plume head at the top that's many hundreds of kilometers across and that flattens, when it reaches the base of the plate, to something potentially 2,000 kilometers across. This kind of arrival and flattening can take place within a very short time, meaning that you suddenly get into a zone where massive melting happens very, very quickly, leading to large igneous provinces, and providing an explanation for the very rapid generation of huge volumes of melt. This mantle plume idea has also been useful for us with respect to radiating dike swarms. We see that in the Mackenzie swarm, which is one of the most magnificent in the world. It extends out over 2,000 kilometers in radius.

Oliver Strimpel: Where is that located?

Richard Ernst: In northern Canada. It continues from the Arctic coast of northern Canada all the way down to not too far from Lake Superior. Radiating swarms are the kind of thing you get from domal uplift above a mantle plume. A plume lifts up the crust by a kilometer or so and produces radial fractures that are filled by magma from the plume, and you get a radiating swarm. The swarm in the Mackenzie, and the swarms in many others, are radial for a distance and then, beyond a certain distance, swing, influenced by some sort of regional stress unrelated to the plume. We take that transition as a measure of the size of the underlying plume. So in the case of the Mackenzie, it transitioned from radiating to swinging into another regional stress pattern at about 1,000 kilometers, consistent with this mantle plume modeling.

Oliver Strimpel: Do these various models also help explain what we see on Mars and Venus, then?

Richard Ernst: Indeed, they do. On Venus, I mentioned this nearly 4,000-kilometer-long individual dike. It radiates away from a big volcanic center, radiates out 1,200 kilometers, and then swings about 50 degrees into a trend that's parallel to a rift zone that's 10,000 kilometers long. So we have a dramatic change in trend right at 1,200 kilometers, consistent with hundreds of other dikes that we are seeing as part of the radiating system, but this particular individual dike shows it so dramatically. On Mars, we see beautiful radiating swarms that seem to go out 3,000 or 4,000 kilometers, and the entire Tharsis Rise is larger than 5,000 kilometers across. So it's a big event on its own.

Oliver Strimpel: How does the model explain the initial radial pattern, and then, at a certain radius, the swinging of the dikes?

Richard Ernst: In order to drive a dike out laterally that far, you have to have a lot of magma, so you have to have magma continually being pumped into the starting end of the dike. The source area for dike magmas is typically near the plume center. Within a few hundred kilometers of the plume center, we can see in some cases gravity anomalies marking intrusions from which the dikes were pumped out. And they're pumped out initially in the direction given by the domal uplift, which imposes stresses on the crust that favor radiating openings. So the magma that's being pumped outward from the center region takes advantage of these preferred opening directions and happily drives out laterally, and keeps driving out sideways. And then, if it gets beyond the plume head, there is still regional stress. On Earth, it's related to plate boundary stresses. In continental areas away from spreading zones or collisional zones, there are intraplate stresses that have preferred directions. So the idea is that the magma keeps being pumped into the starting end of the dike, and the magma keeps moving at the far end. Once it gets beyond the domal uplift, beyond the influence of the stresses related to the uplifting plume head, the magma still wants to go, and it's going to happily go in the direction of whatever stress it experiences at that point. On Earth, that is presumably related to some sort of plate boundary, collisional or subduction stresses. On Venus, we see this one that parallels this huge rift zone for 2,000 kilometers of its 10,000-kilometer full length.

Oliver Strimpel: As we look through the geological record, we often see a temporal coincidence between the emplacement of a large igneous province and continental rifting. Can you give some examples of this, and do we understand why this is so?

Richard Ernst: We clearly see many examples where the arrival of a large igneous province is associated with attempted breakup, and often successful breakup, to open a new ocean. On the one hand, you have the Central Atlantic Magmatic Province, at 201 million years, which precedes the opening of the Atlantic by 5, 10 or 15 million years. There's a beautiful correlation there. On the other hand, you have the magnificent Siberian Traps large igneous province, which did try to break Siberia away from Europe. The west of Siberia has major rift zones that are part of the Siberian Traps event and represent serious attempted breakup, but it was ultimately unsuccessful. So this points out that there are two aspects that are relevant. You have to have favorable plate boundary stresses in order for breakup to be successful. In the case of the Central Atlantic, there was already a whole series of rift basins, at least on the North American side, called the Newark series of basins, which represented north–south rifting prior to the ultimate breakup, which happened 30 or 40 million years later, after the arrival of the Central Atlantic Magmatic Province. So in that case, you'd say there were favorable plate boundary stresses that were already inclined to cause a breakup, and then the mantle plume arrives, causes some uplift, causes radiating dikes, but also causes potential triple-junction rifting in multiple directions, going north to separate North America from Europe and Africa, and into the south to move South America away.

Oliver Strimpel: Ore deposits are linked to large igneous provinces. What kinds of ores are these, and do we understand what the connection is?

Richard Ernst: Yes, we have some good ideas. Most ores of copper, nickel and the platinum group elements, platinum and palladium in particular, are genetically associated with large igneous provinces. Part of the story has to do with the kind of magma you get, both mafic and ultramafic magmatism, in large igneous provinces. Ultramafic magmatism can often be associated with elevated levels of some of these elements, which can then get concentrated in a layered intrusion, for instance the Bushveld intrusion in southern Africa, a huge reservoir of many rich ore deposits. It's a single body that extends 400 kilometers across and is maybe 7 kilometers thick, a huge body of magma that differentiates and produces significant ore deposits at different levels. And it's part of a large igneous province from a little over 2 billion years ago in southern Africa. Then there are the Norilsk magmatic sulfide deposits in the Siberian Traps event. Those magmas seem to have interacted with sulfur-rich units at some point, causing sulfur saturation and the production of a separate magmatic sulfide liquid that can interact with the huge volumes of magma that might be moving past it in a magma chamber. The idea is that if you have a separate sulfide liquid adjacent to the basaltic liquid, as the basaltic liquid moves past it, its sulfur-loving elements, like nickel and copper and platinum group elements, jump out of the basaltic magma and into the sulfide magma, which progressively gets more and more rich in them. That's part of the reason why large igneous provinces are important for magmatic sulfides: they have the potential for huge volumes of magma to move past magmatic sulfides and continually offer them the opportunity to get more and more enriched in the sulfur-loving elements, nickel, copper and PGEs, which are very important in the green revolution.

Oliver Strimpel: We also see a correlation between the emplacement of large igneous provinces and major mass extinctions. Can you give examples of these, and how do we think this comes about?

Richard Ernst: Yeah. The best, perhaps, is the Siberian Traps event that I mentioned before, which covers northern Siberia and central Siberia. All of it seems to have been emplaced within less than a million years, precisely within a short time both preceding and postdating the well-dated timing of the Permian–Triassic mass extinction. So that's a superb example. I mentioned the Central Atlantic Magmatic Province, 201 million years ago. It precisely brackets the Triassic–Jurassic mass extinction. The Deccan large igneous province in India: again, most of it is less than a million years. It does seem to have two or three distinct pulses, with a very short duration for the main pulse, again bracketing in time the mass extinction associated with the demise of the dinosaurs. In that particular case, the dating of the Chicxulub meteorite impact in the Yucatán Peninsula seems to give a similar age, and so it looks like it's a joint effort between a large igneous province and a meteorite impact in causing that extinction.

Oliver Strimpel: I presume the extinctions are thought to be a consequence of the global environmental impacts of the enormous eruptions that create these large igneous provinces. But what exactly is so inimical to life?

Richard Ernst: There's an embarrassment of riches in that regard. There's rapid global warming, 15 to 20 degree temperature increases above present, and acid rain can be an important component. Modeling of the Siberian Traps event shows that there's CO2 and SO2, sulfur dioxide, coming out, but there are also the thermogenic gases. So there's a huge amount of gas released from sills, the intrusive component, interacting with sediments, and in the case of the Siberian Traps, there's the opportunity to interact with hydrocarbons, with oil deposits, and also the opportunity to interact with evaporite sedimentary deposits, which you get in deserts from the evaporation of water. People heated up some evaporite sediments to see what would be released from a sill at 1,000 degrees adjacent to such sediments, and they would release chlorine and bromine compounds, which are ozone-destroying. So ozone destruction is also a possibility. There's mercury release as well, poisonous mercury. It seems that mercury levels are higher during large igneous province events, so they are potentially a source of mercury poisoning.

Paradoxically, you can also have remarkable cooling episodes that are associated with large igneous provinces. If you erupt a large igneous province at the equator, there are a lot of flood basalts, and then they can be rapidly eroded within a million years or so, and that sucks a lot of CO2 out of the atmosphere. So if you're in a cooler transition time, it's enough to potentially pull you into a snowball Earth, where the oceans are frozen to the equator. An example of that is the longest snowball Earth in Earth's history, from about 719 million to 660 million years ago, the so-called Sturtian glaciation. And what occurred right at the absolute beginning of that? The huge large igneous province in Canada known as the Franklin large igneous province, 2 or 3 million square kilometers in Canada, whose precise age is 719 million years, about a million years before the onset of the Sturtian glaciation. And yet there's more. We have an event of the same age in formerly attached Siberia, called the Irkutsk large igneous province. And there's more: at the same time, in part of southern Africa and probably in adjacent Antarctica, there is another large igneous province, called the Mutare–Fingeren, that's of the same age. Also, in South China, there's a large igneous province of the same age, and probably in northern Europe, in Baltica, there's another large igneous province of the same age. Some of these other ones are also at the equator. So the Sturtian glaciation, the Snowball Earth event, was caused by a large igneous province, almost certainly the Franklin large igneous province, at a time when the Earth was already cooling down.

Oliver Strimpel: It's interesting that for some of these, you eject a ton of greenhouse gases into the atmosphere and you jack up the temperature by 10 or 20 degrees, yet for others, it's the subsequent effect of the drawdown of CO2 from all the erosion, leading to cooling, that is the major effect. I suppose it's a question of different timescales.

Richard Ernst: It's a little more complicated than that, because you also have sulfur dioxide release, and that initially causes greenhouse warming, but it can also convert to sulfate very rapidly and become solar-blocking. So you have the CO2 pulse initially from the gas, both from the volcanics and from the heating up of adjacent sediments by sills. The sulfur dioxide component, if it's large, takes you from an initial warming and starts some cooling. And then if you rapidly remove and erode the basalts at the equator, you more than compensate for whatever CO2 you put into the atmosphere, and you have a net loss. So you would expect a little bit of a pulse of warming right at the beginning, then, through the sulfate aerosols, a slight transition to some cooling, and then, if the erosion kicks in, even more dramatic cooling.

Oliver Strimpel: If three or four coincident large igneous provinces triggered the Sturtian glaciation in the Neoproterozoic, what about the Marinoan snowball episode that came soon after the end of the Sturtian?

Richard Ernst: There are some candidates, both for the interruption between the Sturtian and the Marinoan, which could have pumped a lot of CO2 into the atmosphere and may have taken us out of the Sturtian, and also at the end of the Marinoan, potentially ending that period. And then there's also the Gaskiers glaciation, at 580 or so, which seems to have been very short, and it may have been both started and ended by a large igneous province: started by a basaltic large igneous province component and ended by the felsic component. Can I mention one of the things? We have this wild idea that Venus had a massive climate change, from earlier habitable conditions to its modern climate, with no plate tectonics and no erosion. Could multiple large igneous provinces have happened at the same time on Venus, each of them increasing the global temperature by 10 or 15 degrees? Say you have three superimposed: you can increase the global temperature by 45 degrees, and that would lead to runaway ocean evaporation and to the end of plate tectonics, since water is considered important for subduction to continue. So we propose that this great climate transition, from Earth-like conditions on Venus to the current 450°C, self-cleaning-oven-type temperatures, was initiated through multiple large igneous provinces happening at the same time. And we have cheekily said that perhaps this is in Earth's distant future. There's nothing to stop a similar mechanism from happening hundreds of millions of years from now.

Oliver Strimpel: Richard Ernst, thank you very much.

Richard Ernst: Thanks so much, Oliver.

Oliver Strimpel: To see pictures and illustrations that support this podcast, go to geologybites.com, where you'll also find transcripts and 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.