This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. Since almost all of our modern understanding of the Earth has something to do with the theory of plate tectonics, it is natural to ask when plate tectonics started. This is a very challenging question, since so few rocks from the early Earth survive to the present day, and many of those that do have been heavily altered by intervening metamorphic or hydrothermal events. I discussed this topic in an earlier podcast episode with Peter Cawood. One key manifestation of plate tectonics is the relative motion of tectonic plates. My guest today recently succeeded in detecting such motion occurring fully a billion years earlier than any previous such detection. Alec Brenner studies Archean rocks using paleomagnetic and petrographic methods. He is a postdoctoral associate at Yale University. Alec Brenner, welcome to Geology Bites.

Alec Brenner: Thanks for having me, Oliver.

Oliver Strimpel: Let's start with your results. Exactly what motions did you measure, and when did they occur?

Alec Brenner: We looked at a piece of Earth's crust that's been around for over 3.5 billion years. Ancient pieces of crust like that are called cratons, and we were specifically looking at the Pilbara Craton. It's part of Western Australia today. By making magnetic measurements of some of its rocks that are 3.5 billion years old, we measured the latitude where those rocks formed over time. And those measurements showed us that at the time, about 3.48 billion years ago, the Pilbara was at about 53° latitude, give or take somewhat. That's roughly where Berlin, Germany, is today. But when we then did the same measurements on rocks that were just a few million years younger than those, those slightly younger rocks showed that the Pilbara was at quite a bit higher latitude, about 77°, give or take. That's about where Svalbard is today. So a pretty big shift in just a few million years. And that means that the Pilbara had moved rapidly from one latitude to another during that time, right? And the rate turned out to be about several tens of centimeters a year, in line with the very fastest of modern plate motions. And what's especially important here is that our colleagues have previously measured rocks of the same ages in another ancient piece of Earth's crust, called the Kaapvaal Craton, in South Africa. And those measurements of those South African rocks showed that that craton, the Kaapvaal, was at the equator 3.5 billion years ago and was not moving at the same time. That's a big contrast to the Pilbara part of Australia, which was at higher latitudes and moving around. And that means that we've found two pieces of crust that were moving relative to each other 3.48 billion years ago. And fundamentally speaking, that requires that there's a plate boundary in between those two pieces of crust. That's differential motion, and that's a hallmark of plate motions.

Oliver Strimpel: And you know that that was actual translational motion, rather than just one giant rotation of a single plate?

Alec Brenner: It's a little more complex than that, in the sense that all plate motions are rotations. But yes, the fact that we see one piece of crust moving and the other isn't allows us to constrain that they aren't part of one global single plate, one global shell that doesn't have plate boundaries in it, as has been proposed pretty often for this time in Earth history.

Oliver Strimpel: You said that the one plate that you actually worked on, the Pilbara, was moving fast, but the other plate was not really moving at all. Is that actually something different from what we see today, in the sense that we wouldn't really find any plates that are static over the timescale you described?

Alec Brenner: It's quite similar to today in that respect, actually. There are plates that you can find that aren't really moving all that quickly. A good example is South America. It is moving in longitude somewhat, but paleomagnetically, using the technique that I used to study this problem, you can't really see that it's moving, because it's not moving in latitude, and it's also not rotating; it's just moving in longitude, if that makes sense. So that motion is invisible to us. And there are also a few other places on Earth where you can see no motion occurring. But yes, that's roughly in line with what you might expect.

Oliver Strimpel: So, as you said, you use paleomagnetic methods to measure this motion, and paleomagnetism has actually come up in previous episodes of Geology Bites, in particular the one on supercontinents with David Evans. Can you refresh us on how paleomagnetic methods enable us to reconstruct past plate motions?

Alec Brenner: Yeah, sure. Earth's got a liquid iron–nickel metal outer core, and that's electrically conductive and also convecting, and so it generates a strong magnetic field that surrounds Earth. And it's in the shape of a dipole, right? The shape of a bar magnet, with a north pole and a south pole. And what's handy is that most rocks forming on Earth's surface contain tiny grains of magnetic minerals, like magnetite and hematite and pyrrhotite. And those grains can act like microscopic compasses: when they form, they can freeze in place the magnetic field where they form, and therefore record it for us to measure later. And so, much like you can use a compass to navigate your way around the world today, we can measure the directions of these magnetic signals preserved by tiny magnetic mineral grains in rocks to piece together how those rocks, and the pieces of crust they sit on, have shifted over Earth's surface over time, piecing together things like the supercontinent cycle and the formation and closure of oceans. For this study that we're doing, we don't have very much context beyond just single points in time for individual plates. But what's important is that we can measure the latitude at which those rocks formed. And that, it turns out, is very straightforward to calculate from how steeply up or down the magnetic field is pointing at Earth's surface. If you're at the equator, the magnetic field is horizontal to the Earth's surface, pointing north. If you're at the north pole, the magnetic field points straight down into the Earth, and if you're at the south pole, it points straight up out of the Earth. And it smoothly varies in between.

Oliver Strimpel: But how do you cope with the fact that the actual magnetic pole is usually wandering around, even if it's not reversing?

Alec Brenner: Yeah, we cope with that with a lot of samples. Yes, the magnetic field does vary, through what we would call in the business paleosecular variation. I've heard it analogized to a drunk person hanging on to a pole in a subway car: they kind of dance around it, as they're unstable, but on average they're still standing up straight. And the same thing applies to Earth's magnetic field. It's chaotic, and it varies quite a bit, but on average it's still oriented along Earth's rotation axis. And the reason that works is that Earth's rotation organizes all of the convecting motions inside the core along that axis, on average, over tens of thousands of years. And so if you take a lot of samples that are preserving the field at slightly different points in time, over several tens of thousands of years in Earth's past, you can average out all of that variability and get at something a lot more coherent, an average result. So we sample a lot of rocks. In this case, we took several thousand samples for this study, to be able to average all of that out properly.

Oliver Strimpel: Wow, several thousand? And how much rock do you need to drill out for each sample?

Alec Brenner: Each sample is, in our case, a core of rock that's about 2.5 centimeters, an inch, wide and a few inches, like 10 centimeters, long. And we can also grab a block that we've oriented. But the main thing that we have to do when we're sampling is make sure that we're recording the orientation of the sample as we found it, because we're measuring the directions of the magnetization vector in these samples. And so we have all these fancy orienting devices in the field with us to measure that. But for any geologists or geology enthusiasts who have been out in the field and randomly seen, oh, that outcrop has a cluster of holes about that wide in it: that's people like us. That's paleomagnetists, usually. And the number of samples there reflects, again, our effort to try and reproduce our data and make sure that we're getting that average field orientation.

Oliver Strimpel: Let's talk a bit about the actual kind of rock that you're sampling. Are these rocks lava flows that took place on an ocean floor and left magnetic stripes, like we see today on either side of the Mid-Atlantic Ridge, for example, and then are you using uranium–lead in zircons inside that lava?

Alec Brenner: Yes and no. They are, yes, lava flows that erupted onto the ocean floor about 3.5 billion years ago. They're mostly pillow lavas, so exactly the sort of thing you might see in, say, Hawaii today. They aren't necessarily in the same setting that would have produced magnetic stripes, or magnetic lineations, on the seafloor, because these formed in a setting that's not a mid-ocean ridge that's spreading apart. The setting would look more like maybe something like Hawaii or parts of Iceland today. We don't have very much context, but that's at least something that we do know. And yes, the way that we date these rocks is using uranium–lead systematics in zircons. The lavas themselves are very poor in zircon; these are mostly basalt lavas. And what's fortunate is that in between some of these lava flows, you get sedimentary horizons, or felsic volcanic horizons, that are rich in zircons that can be dated using uranium–lead methods. And so throughout this stack of lava flows that we've measured, throughout this stratigraphy, there are different points that other colleagues of ours have dated, and so we can interpolate what time all of those different rocks formed.

Oliver Strimpel: So it seems like a huge question is how do we know that the magnetic signal you're measuring today was in fact imprinted so very long ago, and is not the result of some intervening effects that overwrote, or in some other way masked, the original magnetic imprint? And I wonder how intense such effects would have needed to be in order to erase the original signal.

Alec Brenner: Right. As you've alluded to, magnetizations that are preserved by rocks can get reset in a few different ways. The main one is thermally. If you heat a magnetic material up to a high enough temperature, the magnetization that it's holding onto gets overwritten with a new one. You can actually try this at home by putting a fridge magnet in your oven and baking it, and you'll actually neuter the magnet; it will make it quite a bit weaker. For the mineral that we're measuring in nature most often, that's magnetite, the temperature required to completely reset any magnetite grain is about 585 °C. That's almost 1,100 °F. And there's another issue also. Thermal processes are not the only thing that can reset a magnetization. Rocks can also undergo chemical reactions between minerals and fluids, for example, that might dissolve or grow grains of magnetic minerals. And so we have to be really careful that the magnetizations that we are measuring date to when those rocks formed, or at the very least that we can date when those magnetizations happened in the first place. And that goes for anyone studying the very earliest, oldest rocks; as you pointed out, they have all been heated and altered somewhat by billions of years of later geological events. And so it doesn't matter if you're a paleomagnetist like myself or a geochemist: you have to make sure that the signals that you're measuring actually date to when you think they do. And we really have our work cut out for us. We have developed a whole battery of tests in order to establish the authenticity of that magnetic signal.

Oliver Strimpel: Can you describe what these tests are?

Alec Brenner: Sure, yes. They're called field tests in paleomagnetism, and they tell us whether magnetizations in a rock predate or postdate some geological feature of interest. They take advantage of the traditional crosscutting relationships that geologists use for mapping and relative dating of different units. One of those is a fold test. So let's say you measured rocks in a bunch of different areas that are all the same age of rock, but they've been folded or tilted in different ways after they formed. What you should see, if the magnetization is older than the folding event, is that, you know, two different rocks that originally had one coherent magnetization, and then got folded: if you correct for that folding, their magnetizations should line back up again, roughly speaking. And so, therefore, the magnetization predates the folding, which, again, you can date using other mapping and uranium–lead methods. Another one of these tests is a baked contact test. This is something you can try wherever there's a magma body, like a dike or a pluton, that's intruded into some older rock. And the trick there is to measure the magnetization inside the intrusion, just outside of the intrusion, and then gradually further and further away from it in those surrounding older rocks. And what you should expect to see, if nothing has been overprinted, is that the intrusion magnetization and the magnetization immediately outside of it are in one direction, and then, as you work your way further and further away from it, you should expect to see it gradually give way to some other magnetization in the surrounding older rocks. That means that the intrusion magnetization dates to the time of that intrusion, and the other magnetization, in the country rocks around it, is older. There are more examples of tests than that, but those are the two tests that we've used in this study, and our rocks fortunately passed them with flying colors. That means that they're still holding on to magnetizations that are older than a folding event that's been dated to about 3.4 billion years ago, and also older than magma intrusions, dikes, that have been dated to 3.44 and 3.46 billion years ago. And again, just for reference, the rocks themselves that we're measuring are 3.48 to 3.45 billion years old, so that's not much older in the grand scheme of things. And so we can have confidence that what we're looking at are truly ancient signals, rather than some later noise superimposed on top of them.

Oliver Strimpel: And these tests rule out both physical, or metamorphic, events as well as chemical overprinting?

Alec Brenner: Yes and no, actually. Yes, it doesn't matter where the magnetization came from; these tests can still apply. They're still useful for figuring out what the relative ages of the magnetization and geological events are. That said, we've actually done quite a bit of legwork on these rocks to determine how they got magnetized in the first place, and they're actually not holding on to the original thermal magnetizations that these lavas would have picked up when they cooled on the ocean floor. We've actually determined that they got remagnetized in seafloor hydrothermal systems. This is where hot seawater percolates through the rocks and makes them undergo chemical reactions, and those reactions, as I mentioned earlier, both dissolve old magnetite grains and grow new ones. So that's a little disquieting. But what's fortunate here is that even though these hydrothermal magnetizations are overprints, they're still incredibly old. In fact, they actually happened just after the rocks formed, within a few million years at most. And we've been able to determine that by mapping the hydrothermal deposits that were formed by these systems, with help from all sorts of different datasets. A major one is regional-scale spectral images, taken by planes and satellites, of the field area that we're in. And that all allows us to pinpoint where in the stratigraphy, where in the stack of lava flows, and therefore when, in the relative timing of things, these hydrothermal circulating fluids percolated through each of the rock units that we sampled and reset them. So we're still able to use them. We're leveraging that alteration as a signal itself.

Oliver Strimpel: Okay, let's get back to your results and what we can learn from them. You show relative motion between two plates about 3.5 billion years ago. Does this require full-blown, modern-style plate tectonics, with subduction and so on? Or can we explain your observations with alternative models?

Alec Brenner: No, it does not necessarily require modern-style subduction. Fundamentally, stepping back and looking at our results, all that we can say completely confidently, geodynamically speaking, is that we see one piece of crust, the Pilbara, moving relative to a different piece of crust, the Kaapvaal. And this rules out a situation, as I've mentioned before, that has often been proposed for this time in Earth history, where Earth's whole surface is one giant global shell that's rendered stagnant and unmoving by its lack of plate boundaries. And that's because our data require a plate boundary of some kind in between the two pieces of crust that we can see, again, moving relative to each other, each on a separate plate, 3.48 billion years ago. What we don't know is where, or how many, or what kind of plate boundary that might have been. There are different scenarios, depending on what that might have looked like, that feature moving plates that are not modern plate tectonics. So of course, modern plate tectonics, as you mentioned, is possible. That's where plates actively move themselves as they're pulled into the mantle by sinking at subduction zones. But you can also maybe get plates moving as they're more passively, sluggishly pulled along by the mantle as it drags on their underside. And you can also envision an Earth that might be stagnant most of the time, but intermittently might undergo these really brief but rapid episodes of plate motion that resurface big portions of Earth's crust. And our data can work with all of those situations that I just mentioned, so we can't necessarily, with what we have in hand at the moment, say anything more than that. We are working on additional data from a longer time span in the Pilbara, and those data may suggest that what we're looking at is that last, episodic, intermittent case that I just described. So, again, that would be quick bursts of plate motion punctuating these long, tens-of-millions-of-years-long periods of quiescence, static periods with no motion. But that's just a work in progress.

Oliver Strimpel: But you can say pretty confidently that it wasn't just a big stagnant lid all the way around the Earth, perhaps a bit like Mars or Venus today.

Alec Brenner: Correct. Our data do require that there was some kind of plate boundary to accommodate that differential, that relative, motion between two different pieces of Earth's lithosphere at the time.

Oliver Strimpel: Do your results enable us to say anything about what else was happening in the Earth around 3.5 billion years ago?

Alec Brenner: Yeah, I guess I'll make two points there. One is that paleomagnetic data are direct observations of what Earth's core is doing, because that's what makes the magnetic field in the first place, and there are things that we can look for in our data that keep tabs on that. A great example would be a geomagnetic reversal. This is where Earth's north and south magnetic poles switch places, and this happens pretty frequently on the modern Earth, a few to a few tens of times per million years or so, one to 10 times per million years. And excitingly, we actually found a reversal in our data that happened about 3.46 billion years ago. You can see its signature in the Pilbara data, and in the data from the Kaapvaal Craton as well, at the same time. And that's actually the oldest example of a reversal that's ever been identified. And so you wouldn't be wrong to conclude that Earth's core could circulate in a pattern that makes a modern-like magnetic field, which also conveniently enables us to do this paleogeographic analysis that traces how things moved around on Earth's surface. The other thing that I'd point out here is that I'm working on another, also unpublished, dataset, also from the Pilbara, that likely tells us about how the atmosphere circulated 3.4 billion years ago. That's the age of a bunch of deposits in the Pilbara called evaporites. Evaporites form inside sediments that are exposed to arid air and arid surface conditions. And what we're finding is that those evaporites formed at roughly the same subtropical latitudes, like 15 to 35° or so, that most of Earth's deserts form in today. And that latitude relationship is actually governed today by how the atmosphere circulates from equator to pole. And so it's a great illustration, I think, of how powerful paleomagnetism can be as a technique, because it can tell you about dynamical processes on Earth deep in the past, all the way from the core to the mantle and the crust and even the atmosphere.

Oliver Strimpel: Wow, that's amazing. So the subtropical desert regions are the result of a certain kind of atmospheric circulation pattern that is of the type that we have today.

Alec Brenner: That's what it appears to be. The type of circulation pattern here is called the Hadley circulation, after, I believe, George Hadley, who came up with it in the 1700s. Basically, warm air rises and drops all of its moisture at the equator, and then makes its way northward at high altitude, and eventually sinks back down to Earth's surface at roughly 30° latitude. And that's responsible for the subtropical jet stream. It's responsible for the trade winds at those latitudes. And also, because all of that air had dropped its moisture at the equator, it's dry, and so when it gets back to Earth's surface, it dries out that part of the Earth. Hence, if you look at a map of the Earth and look at where all the large deserts are, the Sahara, the Kalahari in Africa, Australia, Arabia, the Middle East, the desert Southwest and parts of northern Central America, these are all deserts in that latitude band for a reason. It's where that air is coming down from high altitude and drying things out. And it looks like we can see the same thing happening 3.4 billion years ago.

Oliver Strimpel: It's striking also that the time period you're studying coincides with the time when we see evidence in the rock record of the very first life forms. I talked about this, and about stromatolites in particular, in an earlier podcast episode with Martin Van Kranendonk. Have you seen any evidence for life associated with the rocks that you've been studying?

Alec Brenner: Yeah, it's kind of perfect, actually, that you mention Martin here, because it turns out the stromatolites he's worked on for the last few decades, among other things, are actually right in between the rocks that I'm studying for this paper. They're the oldest well-established fossils on Earth, and some of them even grew in a hot spring field, which Martin and his colleagues were responsible for documenting, that was fed by one of the hydrothermal systems that gave our lavas their magnetic signals. And so that means that now we have a little bit of direct tectonic context for those life forms: they grew on a tectonic plate either in motion already or about to be. And that's really intriguing, because plate tectonics is intimately linked with all sorts of processes that foster habitable conditions. It plays a role in keeping surface temperatures stable for billions of years, by modulating the amount of CO2 in the atmosphere. It creates environments with sources and gradients of chemical elements that living things can metabolize and build from, for instance in hydrothermal systems. And to be clear, I should point out that the fossils that Martin's worked on, and the plate motion I found, are probably not the oldest examples of those things, full stop, right? Other, even older examples almost certainly came before them. But still, I find it profound that, in the endless quest of geologists to look further back in geologic time, the oldest records of life, of habitable, energy-rich environments for that life, and of the tectonic plates that might have generated such environments, all just coincide pretty much perfectly. I can't help but imagine that the early Earth might have dealt life a really great hand to get started and develop.

Oliver Strimpel: You spent quite a bit of time in the Pilbara region of Western Australia collecting all those thousands of samples you talked about. What's it like to spend time in the wilderness there?

Alec Brenner: On a practical level, it's a place you might call ruggedly beautiful, I guess. It's hot, it's dry. Pilbara comes from a local Aboriginal word meaning dry country. It's very hilly terrain that's pockmarked by all these spiky bushes of grass everywhere, scraggly desert trees, the occasional kangaroo, sometimes an eagle if you're lucky. And under all that desert is this deep sense of antiquity that you can't really escape from. You're always aware of it. The rocks, of course, are crazy old, and they look it too. They've been out in the weather for billions of years, and that's stained them all these weathered hues of red and orange, much like a lot of the continent of Australia. But you also get hints of a history that's a lot closer to home. During my fieldwork, I've had the great privilege of stumbling onto prehistoric rock art sites: petroglyphs, chippings in the rock faces, that depict people, human figures interacting, animals (you'll see dingoes and kangaroos and emus all depicted there), parts of the landscape, maps of river systems. And having had a chance to talk with some of the modern descendants of those ancient artists, they were clearly all trying to understand their place in the world around them, tens of thousands of years ago. And in a weird way, that's kind of what my colleagues and I do out there too. We're trying to understand our place in the world around us and how we got here, just using different tools, tens of thousands of years later. And walking around the Pilbara really gives you an immense sense of timelessness and heritage: some of the oldest art made by human beings and the oldest known life forms are right there next to each other. And it makes you feel small, in all the right ways.

Oliver Strimpel: Alec Brenner, thank you very much.

Alec Brenner: Thanks, Oliver. It's been my 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.