This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. When I went on a trek in the Himalaya a few years ago, a friend of mine told me that I'd be starting out on rocks scraped and metamorphosed from the Indian continent, and then, near the top of the pass, I would cross onto very different rocks that once formed at the bottom of the now-disappeared Tethys Ocean. Eagerly expecting such a momentous transition to be marked by a major feature in the landscape, I was surprised to find no such thing. In fact, it was all I could do to see the transition, faintly marked by a somewhat darker rock overlying a somewhat lighter rock on a rocky mountain face. So why did the bedrock geology have no effect on the landscape there? Is the opposite true sometimes as well, and what does it depend on?

Bob Anderson is chair of the Department of Geological Sciences at the University of Colorado Boulder. He is a geomorphologist who has studied most aspects of the landscape, focusing recently on alpine and Arctic landscapes in which ice plays a prominent role. Bob Anderson, welcome to Geology Bites.

Bob Anderson: Many thanks for inviting me to this conversation, Oliver.

Oliver Strimpel: Let's start with a puzzle I encountered in the Himalaya. Why did the dramatic shift in bedrock geology have no apparent effect on the landscape?

Bob Anderson: I can translate that into asking what about a rock really matters for its susceptibility to erosion. The erosion of a point in the landscape requires that the rock break down into small enough pieces to be transported by the available transport processes, and the conversion of a rock into such pieces is in turn governed by a number of processes, including physical and chemical weathering processes. A rock is composed of minerals that differ in their properties. Some dissolve easily, like salt, and others are converted to secondary weathering products like clays; feldspar transforms into clays. Some of these minerals are more recalcitrant to these chemical processes and stick around.

But in the environment you're talking about, at the top of the Himalayas, the more likely processes involved are physical processes. A rock heats up and cools on a daily basis. The water in its pores freezes, attracting yet more water to freeze at that same point, which generates an ice lens in the rock that can blast it apart, by what we call frost weathering. In this case, at the top of the Himalayas, you won't be dealing with trees, whose roots break up rock as well. And therefore it's the density of flaws in a rock that really matters. These can be joints in the rock, or fractures that can come from the processes by which the rock formed in the first place, or they can be due to proximity to a fault, for example.

So it's the density of flaws in a rock. Bedding, for example, is a flaw. It's a characteristic of a rock that will differ from, say, a shale, which has very fine bedding, to a sandstone, which might have bedding every meter or decimeter. So the density of that bedding, the fineness of that bedding, matters in the case of a sedimentary rock. In the case of an igneous or metamorphic rock, the flaws are often more widely spaced, and therefore those rocks tend to be harder, or more resistant to weathering, or to transformation into pieces that can be transported. So I can imagine that in the place you're talking about, the metamorphic and igneous rocks that you were seeing in that section might not differ very much across that geologic contact in terms of the density of flaws.

Oliver Strimpel: So is the erodibility of a rock that you've just really been talking about something that is intrinsic to the rock, or do rocks behave differently depending on their environment?

Bob Anderson: Yes, the proximity to a fault can matter a lot, in that the straining of a rock during the earthquake cycle that ultimately allows slip on that fault can strain the rock not just right at the fault interface, but nearby, out to hundreds of meters to a kilometer. So there's a sort of damage zone, which in the terms I've just been using would translate into the density of flaws generated in a rock, governed by its proximity to a fault.

A great example of this is in the Sierras. The Sierras are basically just a big amalgamation of plutons that accumulated in the Mesozoic and have gently been brought to the surface by erosion of the overlying volcanic massif, both on the east and the west of the Sierras. It's granite, and one might imagine, therefore, that, OK, it's all going to erode at about the same rate, or be susceptible to erosion in the same way. But that couldn't be further from the truth. If you go to the western Sierras, let's say Yosemite Valley, the rock is very intact. The climbing routes on the walls of El Cap are tens of meters apart. So the density of flaws, in other words, in that particular granite is sparse, and that allows the rock to stand up in kilometer-tall walls. It manifests itself in other ways that we'll talk about in a moment.

That contrasts a lot with the eastern Sierras. On the eastern margin of the Sierras, there's a frontal fault, a Basin and Range edge, where there is dip-slip motion on the fault that generates the topographic front. That's an abrupt front on the eastern side, and the rock near that fault in the eastern Sierras is highly broken up. You don't see very tall walls; they're just rubble heaps, with meter-scale or smaller blocks separated by flaws, the cracks associated with the straining near the fault. So that manifests itself geomorphically as a lack of tall walls in the east, and rubble-pile-like, angle-of-repose-like hillslopes leading down to the valleys.

In addition, it manifests itself in the signature of the glaciation. A glacier damages the landscape across which it flows by sliding at the bed, and if it has rocks embedded in the sole of the ice, those rocks can scratch the bed and essentially abrade it, like giant sandpaper. If the rocks that are embedded in the sole of the ice get to be relatively fine, you get very fine sandpapering of the landscape, ultimately polishing the landscape. That's the process of abrasion. But if the rock is broken up into small enough bits by pre-existing fractures, those rocks can be plucked from the bed wholesale. So big blocks, meter-scale blocks, decimeter-scale blocks, can be plucked from the edges of bumps in the bed, and that gives you a very different kind of look, a sort of hackly or rough look, to the glacier bed.

So the way this manifests in the Sierras is that you don't find glacial polish on the east side, near the frontal fault. The rocks that were at the bed of the glacier got plucked and ended up amalgamated into huge, 100-meter-tall ridges, or moraines, that illuminate the footprint of the glacier back at the Last Glacial Maximum, for example. So you have a lack of tall walls, a bunch of blocks that end up as moraines, and a very hackly, rough-looking glacial bed in the eastern Sierras. On the west, it couldn't be more different. You don't have blocks that are susceptible to being plucked from the bed. The flaws are too sparse, the glacier cannot pluck, and therefore it's left with its only other technology, which is to abrade. And it abrades so well that the rock gets polished. The western Sierras' Yosemite Valley is famous for its glacial polish, and it's actually quite anomalous, though we think of it as the iconic glacial valley. The fact that the rock is so intact manifests itself as this polished bed, which is actually very rare. So: same rocks, completely different manifestation geomorphically, because of the density of flaws.

Oliver Strimpel: So other than its erodibility, are there other aspects of the bedrock itself that can impact the landscape?

Bob Anderson: Yes. Rocks are made of different minerals, and those minerals break down chemically in different ways. In some cases, the chemistry of the soils that are generated from the breakdown of those rocks can manifest itself in the landscape as different vegetation types. A really great example of that is the bristlecone pines that are so famous in the White Mountains of California. They tend to grow in a very limited geologic context: they like dolomite. Wherever you see dolomite way up high, it tends to be colonized by bristlecone pines. So yes, rock type can manifest itself as different flora in the landscape.

Oliver Strimpel: The landscape is really incredibly varied. Does geomorphology attempt to explain it in all its various manifestations?

Bob Anderson: Yes, indeed. We think about the problem of the shape of the Earth at all scales, from the continents, and why they pop up a good fraction of a kilometer above sea level, whereas the seafloor, two-thirds of the Earth, is significantly lower. Going from that largest of scales down to smaller and smaller scales, you go to mountain ranges, which are big, and the big high plateaus of the world, like the Tibetan Plateau, the Andean or Puna Plateau, or even the western United States, which could be considered a plateau. Its mean altitude is about a mile high, which is about where I am here in Boulder, Colorado.

And then you get down to smaller scales, where individual faults play significant roles, bringing hard rock up against soft rock. That's our backyard here. We have the very hard, crystalline Precambrian rocks of the Front Range that were thrust up 60 million years ago against the shales of the Interior Seaway, which are very susceptible to erosion, which is what we've been talking about. So there you have an erodibility contrast that sets the shape of the front and the magnitude of the front of our range. Then, down to smaller scales than that, you get valleys, and the sort of geomorphic unit which is the catchment, or, hydrologically speaking, the locus of all points that drain through a particular point in the landscape; then the hillslopes that bound the streams, and the outcrops that dot the hillslopes. All of those are within the geomorphologist's domain.

And that makes it very much fun. We get to interact with people who know a lot about the geophysics of faulting and of mantle motion, which drives even larger-scale, thousand-kilometer-scale, hundreds-of-meters-amplitude subsidence and uplift of rock, what you had on another podcast here called dynamic topography. All of those processes generate topography; they bring rock into harm's way. That's the deep Earth causing the rock to come into harm's way. And by harm's way, I mean that once it gets above sea level, it gets subjected to weathering processes, and basically to all of the processes that are driven by the solar engine. We have the deep Earth engine driving the mantle and the lithosphere around, and then we have solar power driving the atmosphere, a much lower-viscosity fluid, around, due to differences in heating on a sphere and the rotation of that sphere. That generates the atmospheric currents that deliver water and snow to the surface of the Earth, which then run down by gravity and cause all the geomorphic work to be done. So yeah, we get to work on all of those things as geomorphologists.

Oliver Strimpel: Are there some unifying themes that geomorphologists use that really run through all this very diverse set of topics?

Bob Anderson: Yes, there are, thank goodness. The equations that govern our models of landscape evolution all begin with statements of conservation: conservation of mass of rock, conservation of mass of ice, mass of sediment, volume of water. One can translate that into mathematical statements, so we can lean pretty heavily on statements of conservation.

I'm also enamored of thinking of geomorphic systems as conveyor belts. A conveyor belt goes in one direction, and in, say, an assembly line, the things on the conveyor belt may have stuff happen to them. They may be taken off the belt, or be added to along the way, or somebody tightens a bolt on them. Well, hillslopes work that way. Rock breaks down and is added from below onto the conveyor belt, that is, the mobile layer of soil, which goes in one direction. It goes downhill, driven by gravitationally organized processes of flow or of individual particles moving, and along the way more soil is added to it, and stuff happens to it, like it weathers. Some of the feldspars might convert to clays, for example. Where does that material go? Ultimately it is delivered to what's at the bottom of a hillslope, which is the river. And guess what? The river is going in one direction as well, also due to gravity. It's going down-valley, and the river can be considered as a conveyor. Material is added to it: the hillslopes are line sources that contribute sediment to it, and individual tributaries can deliver sediment to it at a point. So it has line sources and point sources, and again, stuff can happen. The sediment gets sequestered on a point bar and gets weathered before it gets transported again, or grains bash into one another and become a smaller set of grains.

And ultimately, you know what happens to the sediment in the river? Well, it gets delivered to the ocean, and at the ocean, in some instances, the sediment gets embedded in a littoral cell, which is essentially a river of beach along which that sand is being wafted by wave after wave after wave, every 10 seconds, getting pushed a little bit further. Typically the waves have a dominant direction, so it too is going in one direction down that littoral cell. And again, stuff gets added to it by beach cliffs and by rivers. So one way to view geomorphic systems is as a set of linked conveyor belts, along which stuff gets added and stuff happens to the material along its path.

In my own work recently, I've been paying attention to edges in the landscape. Often we think of erosion as being just the lowering of the landscape, just lowering it vertically. But exactly how does that happen? Consider a cubical block of bricks. Which bricks in that pile would be the easiest bricks to dislodge? Well, the ones at the edge, right? Not from the middle of the pile, not from the middle of the top surface, but from the edge of that pile. Erosion works this way. It always grabs the most vulnerable pieces of rock from the surface, and these are at the downstream edges of ledges in a riverbed; they're at the down-glacier edges of bumps in the glacier bed, just as we were talking about with the plucking process. Even at the atomic scale, it's the atoms at the edges or corners of a crystal that are easiest to remove. They have the least number of bonds to break in order to be freed from the crystal lattice. So in all of these cases, the lowering of the surface, what we would call erosion, occurs by the lateral translation of edges across the landscape. At a point, nothing happens, nothing happens, nothing happens, then boom, the surface lowers by the height of the ledge as that ledge passes, and then nothing happens, nothing happens, nothing happens. It's a very spotty erosion of the surface: lowering of the surface by translation of edges.

Oliver Strimpel: In discussing these processes, you've talked about the rates at which things happen in the landscape, especially erosion, and determining these rates must really be crucial in your studies. When geochronology came of age, with the radiogenic dating of minerals such as zircon using uranium and lead isotopes, we could finally pin the stratigraphic geological time scale, which was a relative scale, onto an absolute time scale, which had a revolutionary impact on almost all areas of geology. Is there an equivalent in geomorphology?

Bob Anderson: There absolutely is. What I'm referring to is cosmogenic radionuclides. The two that I use the most, and that have been most useful in the last three decades, are beryllium-10 and aluminum-26. These are cosmogenic, meaning they're produced by rare interactions of cosmic rays with atoms in surface materials. They're radionuclides in that they have half-lives; they decay to other atoms, in this case with half-lives of the order of a million years. If you recall carbon-14, which is in fact a cosmogenic radionuclide born of nitrogen in the atmosphere, it has a half-life of about 5,700 years. So the number of those carbon-14 nuclides decays to essentially zero after a few of those half-lives; let's call it forty or fifty thousand years. And an awful lot of the Earth's surface has evolved on time scales that are longer than that. That only gets us back one glaciation. There have been tens and tens of glaciations over the last three million years, and we can reach those with the dating tools of beryllium-10 and aluminum-26.

The geomorphic community was not using these radionuclides until about 1990, and the reason is that they're very hard to measure. They are produced in really small quantities, and let me give you a sense of what a small quantity is. These are nuclides that are born of cosmic-ray impacts with the nucleus of an atom. For example, beryllium-10 is made by a cosmic ray impacting an oxygen nucleus; that creates beryllium-10 plus some shards. And aluminum-26 is born of a cosmic ray impacting a silicon atom, to create aluminum-26 plus shards. The rate at which these are produced is of the order of, well, here in Boulder, about 20 atoms per gram of quartz per year. To tell you how rare that is, you sort of have to understand how many atoms there are in a gram of quartz, and it's a lot. So to allow you to visualize that, if an atom were a grain of the same kind of size that you typically see in sand dunes, about 0.1 millimeter in diameter, that production rate of 20 atoms per gram of quartz per year would mean you were producing 20 new sand grains out of a pile of sand grains knee-deep, covering the entire state of Colorado. Twenty of those sand grains would be altered in a year. So it's just an unimaginably small number, and yet we can measure these now. The technology has developed so that we have the measuring tools, which are accelerator mass spectrometers. We have two of them in the United States, one at Lawrence Livermore National Lab and the other at PRIME Lab at Purdue, that can measure these rare nuclides.

I've just talked about how beryllium-10 can be produced from oxygen and aluminum-26 from silicon. Silicon and oxygen make up quartz, SiO2. So both of these nuclides are produced in quartz, which is one of the most ubiquitous minerals on the Earth's surface. So every one of the quartz grains on the Earth's surface is effectively a clock. It actually has two clocks working in it: one is the beryllium-10 concentration, and the other is the aluminum-26 concentration. That's pretty much magical. It means we have a ubiquitously available dating material that can be used as a clock on the Earth's surface, and we can now date surfaces that are rocky. We simply collect the rock, isolate the quartz from the rock, boil it down to the impurities in that quartz, one of which is beryllium-10 and another aluminum-26, and then go and measure their concentrations at these accelerator mass spectrometers. So we can date a surface, we can date a moraine, we can date marine terraces, we can date river terraces. We can even date the age of sediment in a cave.

And we've developed the mathematical machinery over the course of the last 30 years to the point where we can assess erosion rates in the landscape, which are typically really low. I mean, just to give you a sense, a millimeter per year is a screamingly fast rate in the geomorphic community. I get really excited by millimeter-per-year kinds of rates. Ten millimeters per year is outlandish, like what the Indus River is doing as it carves through the Himalayas, and that's really rare. So most rates are very small fractions of that, and you're not going to measure them by looking at the landscape, putting a pin in the landscape and coming back four years later at the end of your PhD. That's just not going to happen. So we now have a tool that puts timing on the landscape that was simply inaccessible to us before 1990.

Oliver Strimpel: Can you give an example of a landscape that we were able to date this way, with the result that our understanding of it underwent a fundamental change?

Bob Anderson: We dated the glacial polish I was telling you about in the Sierran landscape, and we were able to document the deglaciation of the range, and that was what led us to this insight into how erodible the rock was. It made us think about the erodibility of the rock in different parts of the valley. In most cases, we could document that yes, the glacier was out to its moraine around 18,000 years ago, and then receded, to be completely deglaciated, lost from the landscape, by about 14,000 years ago. And we have found that that is very common across the western United States: the Front Range in Colorado, the San Juan Range in southern Colorado, the Uinta Range in Utah, the Wind River Range in Wyoming. These ranges all show a deglaciation history that would otherwise be completely inaccessible to us without the use of beryllium-10 concentrations in the eroded rock.

Oliver Strimpel: I can certainly appreciate how satisfying it must be to explain patterns in nature, but to borrow a term you used when we talked earlier, are you content to address such parochial problems? Or do you see the solutions to such problems as also addressing much larger questions?

Bob Anderson: Certainly I see the solutions as addressing some larger questions. There are questions of how the Earth's surface has responded to climate in the past, which has varied significantly over the last couple of million years as we go from glacial to interglacial. And now, of course, this massive human experiment, our human-caused climate change, is being foisted upon our Earth's landscapes, and that raises many questions about the sustainability of agriculture and how water is going to work in the landscape. I mean, we now study what's termed the critical zone, the zone between the top of the canopy and the base of the groundwater system. It's the system within which life evolved, and it's the system that sustains life, and many parts of that critical zone are being tweaked significantly by climate change. We're also engaged in thinking about how the cryosphere has changed. This is essentially the canary in the coal mine of climate change: glaciers in the landscape, permafrost in the landscape, sea ice. There's significant focus on the cryosphere, and there is growing interest in planetary surfaces. So there are very societally important questions that geomorphologists are embedded in, as they should be.

But I will admit from the start that I'm driven by, call it, a sense of awe of the landscape. I absolutely love patterns in the landscape: patterns of ripples in sand and of sand dunes. I'm interested in cracks in the landscape. You know, we talked about flaws: flaws in rocks, mud cracks, frost cracks, frost wedges and so on. I love it when a pattern that is visible in paint at the centimeter scale is also manifest in the landscape when seen from space, as in the Arctic, with frost-wedge polygons that are the size of city blocks but have the same pattern; they obey the same rules. I think the connections between processes at atomic scales and processes at continental scales are some of the things that excite me the most.

Oliver Strimpel: Bob Anderson, thank you very much.

Bob Anderson: It's been my great pleasure.

Oliver Strimpel: For more about Geology Bites, as well as pictures and diagrams that illustrate this podcast, you can go to geologybites.com.