This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. About 80% of Greenland is covered by ice. While we’ve been able to map the geology of the 20% that is exposed, our existing geological maps of the glaciated interior have relied on extrapolation from the exposed regions and educated guesswork. But in April 2024, Joe MacGregor, a research physical scientist at the NASA Goddard Space Flight Center, and his colleagues published a new geological map of Greenland. The map incorporates new data about the geology of the rocks below the ice, derived from seismic, gravity, magnetic and topographic surveys. The map features three hitherto unknown provinces that lie beneath the central part of the ice cap north of about 70°, as well as an extensive network of long subglacial valleys. How does the new data enable us to learn about the rocks buried by up to three kilometers of ice? And what is the significance of the three new provinces and the network of valleys? Joe MacGregor, welcome to Geology Bites.

Joe MacGregor: Thank you for having me, Oliver.

Oliver Strimpel: What was the best map we had of Greenland’s geology before yours appeared? And how was that one compiled?

Joe MacGregor: The best one that existed before ours was compiled by Peter Dawes, a British geologist working for the Geological Survey of Denmark and Greenland, and he published it in 2009. It represented the best understanding that we had at the time of how the various provinces that had been mapped around the periphery of Greenland might connect within the interior. He described it as conjectural, and he left it at that. And so we have attempted to revise that map across the 80% of the island of Greenland that is covered by ice, for which we have almost no direct geological information.

Oliver Strimpel: I said in the introduction that your map uses data from seismic, gravity, magnetic, and topographic surveys. Let’s talk about how we captured this data and what kind of information each data type provides. Taking seismic data first, did you use natural sources, i.e., earthquakes, or artificial sources of seismic waves? And where were the seismometers placed?

Joe MacGregor: We used natural sources of seismic signals, so those include earthquakes and other seismic sources, for example calving of icebergs around the periphery of Greenland. And those seismometers are mostly located on land around the periphery of Greenland. There are a handful that have been placed on the ice, typically located next to major stations. But we’re dealing with a very small number of seismometers compared to the set that is distributed across the United States. And the observations that they make of these earthquakes and other seismic sources allow us to generate tomography of the subsurface in Greenland and around the Arctic.

Oliver Strimpel: Roughly how many seismometers are we talking about here to cover the whole of Greenland?

Joe MacGregor: So currently it’s on the order of two dozen. It’s relatively few, and it gets sparse as you go farther north in Greenland, understandable given the population centers. And so that leads to a degree of bias in how we’re able to interpret what the seismic structure is, particularly in what we think is the more complicated northern region, because you rely on the ray paths between wherever those sources were and those stations to infer what the velocity might have been between them. So the more stations you have, the more ray paths you’re going to have, and the wider variety thereof, to infer that seismic structure.

Oliver Strimpel: So that really goes to the resolution of the tomographic images you can generate, I suppose.

Joe MacGregor: Yes. So the deeper you want to sense, typically the coarser the resolution will be. We’re interested in the near-surface geology, so the tomographic velocity anomaly pattern in the upper five to 10 kilometers of crust. That’s relatively hard to resolve, given the nature of the sources that exist and the ray paths that they travel. But yeah, that’s the challenge we’re faced with. So we’re getting an estimate that’s of the order of 50 kilometers square, across an island that’s, roughly, in rectangular terms, 2,000 kilometers north–south and 1,000 kilometers east–west. To be absolutely clear, although some of my co-authors have generated these seismic tomography maps, we didn’t generate a new one for this study. We took what we thought was one of the best existing ones.

Oliver Strimpel: When you try and interpret one of these tomographic images, do you interpret the seismic velocity as reflective of variations in temperature of the crust or mantle, or is it a combination of temperature and composition?

Joe MacGregor: So it’s really a combination of the two. The density, the temperature, and the lithology can all factor into what generates an anomaly relative to some reference field that we attempt to interpret in this context.

Oliver Strimpel: And before we start talking about the other data types, were there any areas, just looking at the seismic, that appeared to be significantly different or hotter than other parts of Greenland?

Joe MacGregor: There are places in Greenland, particularly as you head towards the east coast, the Blosseville Coast, where there are known basalt outcrops, where it does in fact appear that the seismic velocities are lower, and that would be consistent with warmer rock, potentially a shallower mantle, depending on how you view that structure.

Oliver Strimpel: Okay, let’s talk about the gravity surveys. How were those captured, and what does a map of the gravity variations, or gravity anomalies, tell us about the bedrock?

Joe MacGregor: So those were captured in two ways. The first is from satellite-derived gravity measurements, and here we used the GOCE mission that was run by ESA. That allows us to resolve large-scale gravity anomalies. That’s also been combined with aerogravity measurements, and that is one of the things that I’ve specialized in, which is having gravimeters on board aircraft, and we travel across Greenland mapping the gravity anomaly as we cross the ice. That’s a challenging measurement to make, because gravity and its variations are fundamentally about measuring the vertical acceleration and the minute variations associated with changing rock composition beneath. And when you’re doing that measurement from the air, you’ve got the added complication of the turbulence that the aircraft is experiencing, and that can knock you all sorts of different ways. And you have to record how the aircraft is moving in space very precisely, with GNSS antennas, so that you can then remove that noise to get at the gravity signal beneath. Where you see higher gravity, that’s indicative of denser rock, or, depending on the scale that you’re looking at, potentially a deeper crust–mantle boundary. And that allows you to resolve the large-scale structure of the rocks that you’re flying over.

Oliver Strimpel: But unlike the seismic tomographic images, presumably if you’re actually flying pretty low in an aircraft, you can get better spatial resolution.

Joe MacGregor: Yes, that depends on a number of factors: how fast your platform is traveling, how high it’s traveling, and then of course the quality of your gravimeter. In the ideal scenario, if you want to cover a lot of ground, you’ve got a low and slow aircraft. And in some of these measurements we have had that. So the field that we worked with was the combination of these satellite gravity measurements and some of the aerogravity measurements that have been made, not just by NASA over the past couple of decades in Greenland, but also by various Danish institutions, who are of course interested in Greenland, and in some cases German institutions as well.

Oliver Strimpel: So just looking at the gravity data alone for a moment, were there any regions that stuck out in the sense of having anomalously high gravity or low gravity?

Joe MacGregor: Yes. In the vicinity of the Caledonian orogen, we do often see a lower gravity. That’s in northeast Greenland, and that signal seems to propagate a little inland toward the center of the island. Conversely, in the northwest of Greenland and in the far north of Greenland, we see positive gravity anomalies. Those appear to be associated with a different type of orogen, perhaps a different rock composition, and in one case, apparently a basin, or a past basin.

Oliver Strimpel: And how are the magnetic surveys performed?

Joe MacGregor: The magnetic map that we used was generated in a similar way to the gravity map, but of course with different sensors. There are magnetometers on board satellites that can give you large-scale magnetic anomalies; we’re talking order 100 kilometers. And then those have been combined with aeromagnetic measurements. Those were also made by NASA in the past two decades, and by other institutions. In the case of magnetics, it’s not so much a concern how fast you’re flying, but it is helpful to get low if you’re trying to sense signals closer to the ice–rock interface. In this case, we used a data set that was generated by the US NOAA, the National Oceanic and Atmospheric Administration, in 2017.

We also flew an aircraft with NASA that was particularly well suited to the challenge of making airborne magnetic measurements. As you might imagine, you would like your magnetometer, which is very sensitive to magnetic fields, to be as far away from your metallic plane as possible, and to have that plane be well characterized. NASA used a P-3 Orion that it had inherited from the US Navy. That aircraft was originally designed as a submarine hunter, flying very low and having a magnetometer boom that allowed them to detect any rapidly changing magnetic signals that might be indicative of a submarine.

Oliver Strimpel: Just to give roughly an idea of the extent of the survey and how big a project it was, roughly how many traverses were made across the Greenland continent?

Joe MacGregor: To compile the data across Greenland, NASA flew somewhere north of 800,000 kilometers since 1993. That’s mostly in reference to ice thickness, so not all of those surveys had magnetometry. But there are other predecessor surveys, collected in the ’80s and early ’90s, that did collect that kind of information. So there’s a long history here, again not just the US, of going out and collecting these sorts of airborne measurements to unravel the mysteries of Greenland.

Oliver Strimpel: Can you give me an idea of the resolution of these magnetic surveys?

Joe MacGregor: Sure. So that’s a function of the altitude that the aircraft flies above that ice–rock boundary. Ice is very kindly nonmagnetic, which makes it easier to interpret the signals. And you can think of us in most cases as gaining exceptional information along the flight line, getting an independent data point on the order of every 500 meters. And then we have to combine, interpolate, all of those crisscrossing flight lines, which had all sorts of competing scientific priorities, into a larger grid that is typically interpreted on the scale of order 10 kilometers.

Oliver Strimpel: And is it fairly even coverage, or do you have spots that are very well covered and then big gaps in other places?

Joe MacGregor: There are still gaps in the Greenland ice sheet, which is roughly 2 million square kilometers in size. There are gaps of the order of 10,000 square kilometers in size. There are several of those, but it’s increasingly good, especially as compared to Antarctica, an ice sheet which is seven or eight times larger and a more challenging beast to survey.

Oliver Strimpel: And finally, what about the topographic surveys? How were these performed, and did they capture the topography of the rock surface below the ice or the topography of the ice surface?

Joe MacGregor: Great question. So those are collected using radar sounders. Radar sounders are essentially an evolution of what some of your listeners may be more familiar with, which are radar altimeters. In the ’50s or ’60s, pilots were flying aircraft over parts of the Greenland ice sheet and discovered that sometimes their radar altimeters didn’t work very well relative to what they could see. They seemed to be returning much larger values in terms of the distance to the surface than what was actually apparent. And what someone eventually realized is that that is because ice is relatively transparent to radio waves. Polar ice is typically quite dry and quite cold, all conditions that favor low attenuation of those radio wave signals. So that’s been subsequently exploited throughout the world over ice to measure ice thickness relatively easily. The velocity through the ice at radio frequencies is pretty well constrained, and so that allows us to make widespread ice thickness measurements more or less whenever we’re flying an aircraft over the ice for a scientific survey.

Oliver Strimpel: When you say you can use the radar to tell you what the thickness of the ice is, are you basically comparing the reflection from the top of the ice with the reflection from the bottom of the ice and then taking the difference?

Joe MacGregor: Exactly, yes. And you mentioned, are we using radar to observe the top of the ice? We are, but we have another tool to measure that surface elevation very precisely, which is a laser altimeter. NASA has developed a series of those over the years, and that motivated most of our scientific surveys of Greenland and Antarctica, which is to collect observations of changing ice thickness or ice elevation, to interpret in terms of the broader response of these ice sheets to ongoing anthropogenic climate change. This began in the early ’90s and has continued more or less ever since, and has resulted in now two satellites that are laser altimeters. One was called ICESat, which orbited from 2003 to 2009 and collected laser altimetry data in a way that we’d never had before. And the second is called ICESat-2, which launched in 2018 and is continuing to produce excellent-quality laser altimetry data across both ice sheets, and allows us to have a very precise understanding of their total mass change.

Oliver Strimpel: When we spoke earlier, you mentioned stereo photogrammetry. Is that relevant to the topographic survey?

Joe MacGregor: Absolutely. We mentioned that we collect all these crisscrossing flight lines. We get incredible data along the track of the flight line, but we remain relatively ignorant as to what is going on between the flight lines. So that’s where stereo photogrammetry can come in. It’s not as accurate as laser altimetry, but it fills in the gaps extremely well. And so when those two are combined, we get a happy medium of the accuracy of laser altimetry and the wide coverage of stereo photogrammetry.

Oliver Strimpel: When you measure the topography of the rocks below the ice and then the topography of the surface itself, does the topography of what’s happening below project itself onto the surface of the ice, even if it’s kilometers above it?

Joe MacGregor: It does indeed. It’s been known for a long time that when ice flows over a bump or a valley, even if it’s a kilometer thick, two kilometers thick, it will leave a surface impression of that bump or valley at the surface. Now, that valley might be 100 meters deep, and the resulting bump might only be a meter in amplitude at the surface. But therein lie the benefits of stereo photogrammetry and its ability to resolve relative surface elevation change, and also of these ice sheets acting as superb low-pass filters of these signals. As our ability to resolve relative changes in surface elevation has improved, we now, through this study, have the ability to see how the ice sheet is responding to all the bumps across Greenland, and directly compare it to places where we have ice thickness data, and to see that those line up. And that allows us to look at a map of the surface of the ice sheet, which at its thickest is just three kilometers thick, and then be able to infer where all the valleys are, even if we haven’t measured them all directly. What we learned from that, for example, is that these fjords that have beautiful glaciers that calve off into the ocean, that we see at the edge of the ice sheet, in some cases more or less propagate back into the ice sheet subglacially, in a straight line, for up to 300 kilometers.

Oliver Strimpel: So as I said in the introduction, your map shows three new geological provinces under the ice. How do you define a province?

Joe MacGregor: We describe them generically as regions, and they may yet be associated with some of the known provinces that have been mapped at the periphery of Greenland. How we define a province is not intended to be any different from how geologists describe it elsewhere, which is a common set of geologic features that are contemporaneous with each other, that are reconcilable at large scale. So for example, large igneous provinces: we mapped part of the North Atlantic Igneous Province in the east of Greenland. Large basins, such as the Independence Fjord Basin in the north. Orogens, such as the Inglefield orogen in the northwest and the Rinkian orogen farther south. And cratons and other large-scale features that have common types of rock, whether it’s igneous rock associated with igneous provinces, or sedimentary successions associated with basins.

Oliver Strimpel: But when you decide to put a boundary between one region and another, are you looking at specific features and saying, okay, I think here we have a combination of data sets that indicate we’re dealing with an igneous rock, say, and in another one here we’re dealing with sedimentary rock? Or is it more that you just put all your data together and just look for the greatest amount of contrast, considering all the data as a whole, without taking a view as to what the data is actually telling you about what’s underneath?

Joe MacGregor: So it’s really a combination of the two. In the case of igneous provinces, those are typically going to be associated with large positive magnetic anomalies and relatively slow seismic speeds; it’s a little more nebulous with the gravity anomaly, and the topography could be rough. That’s the simplest case. In most other cases in Greenland, typically with very old rock and provinces that were formed 1 billion years ago plus, except in the case of the very youngest ones, we took our geophysical fields and looked at those fields island-wide, including at the periphery, where we have a much better understanding, obviously, of what the geology is. And we looked at the pattern that we saw there: the magnitude of anomalies, whether they had a particular texture, a common wavelength or azimuth. And then we looked at how that propagated inland and searched for contrasts in those fields, and we did that independently for each of those fields. So what we ended up with was a mishmash of lines in some places, and relatively coherent, consistent boundaries between most fields in other places. More often than not, the coherent boundaries were in southern Greenland, and they were much more complex in northern Greenland, which appears to be an amalgamation of several relatively ancient provinces.

Oliver Strimpel: Can you describe a little bit more about what characterizes each of these regions? You mentioned the simplest case that would indicate an igneous rock. In terms of what you actually found in that northern region that you chose to divide into three regions, would you be able to give them a qualitative description, and then maybe even go so far as to say something about how they were assembled and what their origin is?

Joe MacGregor: We purposefully stayed away from that latter question, as to how they were assembled, and instead focused on what distinguished them geophysically, so that we have a baseline for future interpretation. Recall that the Dawes 2009 map, our reference map, essentially used no geophysical data to constrain what was happening across 80% of the island. So here we’re trying to take that next step, where we look at different regions and say, okay, it seems like in the west we often have a shallower Moho boundary. We seem to have faster shear-wave velocities in the southern two-thirds of Greenland. And the gravity anomalies and the magnetic anomalies, which typically have a finer resolution: that’s where you really start to see a lot of the strange contrasts across northern Greenland that give you a clue that this is a complicated region subglacially in terms of the geology, not just as is apparent from the periphery of the ice sheet.

Oliver Strimpel: When you looked at it after you generated the map and you defined what you think were the boundaries of the regions that created the greatest contrast, was there any particular data type that, in retrospect, seemed to map most closely into each of the different regions, that you could say characterized those regions best if you were just to pick one?

Joe MacGregor: If I had to pick two data types, and revealing my bias as a geophysicist here, I would pick the gravity and magnetics. Those give us the best combination of resolution of what is going on in the subsurface beneath the ice and insight into what the nature of those rocks really is. Topography, the bedrock topography, both observed and inferred, is an indirect indicator at best. And the depth to Moho and the shallow shear-wave velocities, those are helpful context, but they’re relatively coarse resolution.

Oliver Strimpel: We believe that the hotspot that is currently contributing to all the volcanism taking place in Iceland previously traced a path below Greenland as it drifted to the west as part of the North American plate. Were you able to identify a channel of basaltic rock marking the hotspot in your data?

Joe MacGregor: Not directly. It’s widely accepted that 50 to 60 million years ago, the Icelandic hotspot passed through Greenland, but where exactly it passed through Greenland on this northwest-to-east traverse remains an issue of debate and is still not clearly resolved. There are hints that perhaps the hotspot didn’t break through the crust, but it left a thermal signature lateral to the path of the hotspot. There are other studies that have suggested that perhaps there’s elevated geothermal activity in particular regions. We did notice that there are co-located gravity and seismic and, in particular, magnetic anomalies in this region that we called Central Region B, that is in east-central Greenland. It’s subglacial. It does not appear to have any known outcrop, and it is an area that three decades ago someone described, based on some preliminary magnetic and gravity data, as the Central Greenland Magmatic Province. And we went back and forth within our group as to whether or not we would give these specific genetic indicators to these regions, but ultimately we decided to leave them as generic, so as not to lead the witness too much.

But this is an interesting region. Whether it has high geothermal activity remains to be seen. It just so happens to be next to a glaciological feature that is unique in Greenland, and indeed relative to Antarctica as well. That feature is called the Northeast Greenland Ice Stream, and it starts almost at a point close to the middle of the ice sheet and then expands as it slowly heads downstream for several hundred kilometers, until it reaches the ocean and becomes this very crevassed and beautiful glacier in far northeast Greenland called Zachariæ Isstrøm, a glacier that’s actually changing very rapidly today, which we’re quite concerned about in terms of future mass change of the ice sheet. So what starts the Northeast Greenland Ice Stream? There’s nothing like it anywhere else. There are hints that there may have been similar features elsewhere in northeast Greenland, but it’s the only one we see today. And it’s a quite interesting coincidence that it happens to be located right next to this anomalous subglacial geologic area.

Oliver Strimpel: One of the things that you and your team are interested in was to learn about the flow of ice on Greenland, and in particular to see if there’s a relationship between the bedrock and the flow of overlying ice. And you’ve just mentioned a specific correlation that you think strongly suggests there might very well be. You developed a new visualization technique to facilitate this, which you called flow-aware hillshade. Can you tell us about that?

Joe MacGregor: Sure. So this is a hillshade of a digital elevation model, which is a stereo photogrammetric model that we talked about earlier. Instead of lighting the DEM from a single azimuth, it takes into account the direction of the flow of ice, so that instead of being limited to a single azimuth, it can vary relative to the direction of ice flow. In this case, we found that orthogonal to ice flow seemed to work best, and that allows you to visualize these bumps, as the ice is flowing over features, relatively cleanly.

Oliver Strimpel: And what did you find by means of this visualization technique?

Joe MacGregor: What we found was clear evidence that there are large networks of very long (sometimes hundreds of kilometers) subparallel valleys across the ice sheet. And they’re often quite straight, and they have a predominant southwest–northeast azimuth, although not exclusively. They often reveal features that you can see at the margin of the ice sheet that come out as, say, a river or a fjord. So you can look at this flow-aware hillshade and then see, when you get towards the very edge of the ice sheet, if you’re following a single trend line, then, oh, there’s a river. And so what that tells us is that we’re dealing with something that’s glaciologically useful, but it’s also revealing a long-term erosion pattern that is present across the ice sheet in this systematic way.

Oliver Strimpel: So you were able to see the very long valleys best by examining the ice flow on the surface, rather than looking at the radar signal, for example.

Joe MacGregor: So what the radar does is it allows you to get the depth of that valley precisely. However, we don’t have blanket coverage of the entire ice sheet. So what happens is the radar only occasionally crisscrosses many of these valleys, particularly in the deep interior. And in some cases, some of these valleys, not really the longest ones, but some of the more prominent ones, curiously, we’ve never flown over. But it’s clear that they’re there, and they have an effect on the flow of ice above them. Kilometers above, they leave an imprint on the flow of ice. I wouldn’t say that they have a strong influence on how the ice is flowing, but they could potentially in the future, because valleys are places where any water that is generated beneath the ice is going to tend to end up and then flow toward the periphery. And that could ultimately influence how fast that ice is flowing, because water, we know, is very effective at lubricating the base of the ice and ultimately leading to sliding.

Oliver Strimpel: Do we think that these long valleys predated the ice and really have nothing to do with the ice?

Joe MacGregor: That is our working hypothesis. The fact that there are long networks, that they are subparallel to each other and straight, and also the radar observations, which tend to observe more V-shaped valleys than U-shaped valleys, suggest that these valleys probably predate the ice sheet. And they may be indicative of some sort of tectonic imprint on the rock in which the valleys formed that led to preferential paths of erosion for these valleys.

Oliver Strimpel: Yeah, I’m wondering, do these represent sutures between different terranes, or major strike-slip faults that were weaker and so somehow or other gouged out by the ice?

Joe MacGregor: We did consider a hypothesis like that initially, but our conversations with geodynamicists suggest that the stresses that would be required to generate patterns like that are likely far too large to explain them directly.

Oliver Strimpel: But it is very remarkable that you have these parallel valleys, thousands of kilometers long and roughly parallel to each other. It’s not really something we see anywhere else in the world, is it?

Joe MacGregor: Yeah, it’s not very common, and that’s why we felt that it was very much worth noting.

Oliver Strimpel: Do you have a new mission to Greenland in the works?

Joe MacGregor: We do. Actually, I’m part of a new mission called Snow4Flow. We will be measuring snow thickness and ice thickness in key glaciers that are relatively understudied, both in southeastern Greenland but also elsewhere in the Arctic, so that we can better project how these regions are going to change in the context of ongoing anthropogenic climate change.

Oliver Strimpel: Joe MacGregor, thank you very much.

Joe MacGregor: Thank you, Oliver. It was really nice to join you and talk about this study.

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.