This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. If you were to name the world's greatest mountain range, it would probably be the Himalaya. It has the world's highest peaks and stretches uninterruptedly for over 2,500 kilometers, from Afghanistan in the west to Burma and even Indonesia in the east. And it gives its name to the Alpine-Himalayan orogeny, the giant orogenic belt associated with the gradual closure of the Tethys Ocean, which extends for over 15,000 kilometers from the Atlas Mountains and the Alps in the west, to the Caucasus that border Europe and Asia, to the Himalaya and even further east to Southeast Asia. But just north of the Himalaya lie several other great mountain ranges we hear a lot less about: the Karakoram, the Hindu Kush, the Pamir, the Gangdese, the Kunlun Shan and the Tian Shan. How can we explain the existence and geological makeup of these mountains? And are they part of the same geological circumstances that created the Himalaya?

Mike Searle has studied the Himalayan, Karakoram and Pamir mountain ranges for about 45 years, probably logging more hours in the field there than any other geologist. He has published hundreds of papers on the region and is perhaps best known for suggesting and documenting the evidence for a process called channel flow in the Himalaya, in which metamorphosed crustal rocks are extruded towards the south, sandwiched between a thrust fault below and low-angle normal faults above. He is Emeritus Professor of Earth Sciences at the University of Oxford. Mike has been a keen advocate and advisor for the podcast since its launch in 2020. He was my guest in the inaugural and the 50th episodes. Mike, I'm really happy to welcome you back to Geology Bites for this, the 100th episode.

Mike Searle: Well, thank you very much, Oliver. It's great to be back on Geology Bites, and thanks for inviting me again.

Oliver Strimpel: Where exactly are these Central Asian mountains? And let me just point out for listeners that the supporting web page on geologybites.com includes maps and a glossary, which might be especially useful for this episode.

Mike Searle: The geography of Central Asia is a little bit complicated, but as you said, the Himalaya is approximately two and a half thousand kilometers long, running along the northern borders of Pakistan, India, Bhutan, Nepal and Tibet. We're going to talk about the mountain ranges to the north of the Himalaya. This group includes mountain ranges like the Pamir, which is almost entirely in Tajikistan; the Hindu Kush, which is the mountain range along the Pakistan-Afghanistan border; and the Karakoram, which is in northernmost Pakistan and Ladakh. And then there's a bunch of mountain ranges around Tibet. The Gangdese range is about 2,000 kilometers long, stretching from Kohistan in the west all the way along the southern margin of the Asian plate in south Tibet. The Kunlun runs along the northern margin of the Tibetan Plateau in China, and the mountain ranges to the north of this are all in Outer Mongolia. The Tian Shan runs from the Pamirs in the southwest for about 3,000 kilometers along the borders of Xinjiang province in China with Kazakhstan, Kyrgyzstan and Uzbekistan.

Oliver Strimpel: By world standards, these are massive ranges. The Karakoram boasts four of the world's 8,000-meter peaks, and the Hindu Kush, the Pamir, the Kunlun Shan and the Tian Shan each have many peaks above 7,000 meters. No mountain ranges outside this region have such high mountains. Yet we seldom hear much about these ranges. Do you think that is because they're overshadowed by the Himalaya?

Mike Searle: Well, quite possibly, yes. The Himalaya is actually readily accessible from the plains of India and Pakistan and is inhabited, so there are access roads, lots of villages up there and lots of trekking routes. Access in the Karakoram is almost entirely by trekking, or else mountain climbing at higher elevations. The Pamir, Tian Shan and Kunlun are all extremely remote, but they're high, so they're on the plateau or around the plateau, and they do have road access into their margins.

Oliver Strimpel: Before we delve into the geology of some of these individual mountain ranges, can you give us a high-level overview of the origin of these mountains?

Mike Searle: Yes. Overall, these mountain ranges are all part of the present-day Eurasian plate, in contrast to the Himalaya, which, remember, is all Indian plate. The suture zone dividing the Himalaya from southern Tibet, which runs right the way through Ladakh and south Tibet, is the boundary, the collision zone, between India and Central Asia. These mountain ranges to the north of the Himalaya are, in a nutshell, all related to the progressive accretion, or collision, of continental blocks onto the stable Precambrian shield areas of Siberia and Mongolia. These continental terranes caused collisions dating way back into Paleozoic times, so that's more than 300 to 400 million years ago. And the accretion of these terranes becomes younger as you go towards the south. The oldest suture zones, like the Mongol-Altai suture zones in Mongolia, are Paleozoic. The Kunlun is Triassic, getting on for 300 to 250 million years ago. Central Tibet is a Jurassic-Lower Cretaceous collision zone along the Bangong suture. And the Indus suture is the youngest of all of them, which is the one that caused the Himalayan collision. So this accretion of suture zones onto the main Precambrian blocks of Siberia and Mongolia could be the beginnings of the formation of a giant craton, like we see in many of the old Precambrian belts, such as the Kaapvaal-Zimbabwe craton in southern Africa, where you have a bunch of Proterozoic terranes all accreted around a core of old Archean rocks, some of the oldest rocks in the world.

The southern margin of Eurasia was an active plate margin right up through the Mesozoic. As Tethys was forming to the south after the breakup of Gondwana, you had active subduction underneath southern Tibet, with the 2,000-kilometer-long granite batholiths of the Gangdese, and that was all associated with a northward-dipping subduction zone, which consumed the oceanic lithosphere of Tethys to the south. The Indian plate margin, the Himalaya, was a long-lasting passive continental margin, with over 200 million years of stable shallow marine sedimentation. But the southern margin of Asia was an active Andean-type margin, so it could actually have been relatively high, with thick crust, prior to the collision of India.

Oliver Strimpel: Wow, that's fascinating. I said in the introduction that you've logged more hours in the area than any other geologist. Have you been to all these mountain ranges yourself and studied their geology?

Mike Searle: Well, yes, I have. I've worked for over 15 years along the Karakoram ranges in north Pakistan, mostly in the Baltoro Glacier region. I published a big geological map of the central Karakoram in my book Geology and Tectonics of the Karakoram Mountains. At least six of these expeditions to the high peaks were mountaineering expeditions. I soon figured out that the only way you could get up into this terrain was to have the logistics of a major Himalayan mountaineering expedition, and I usually invited three or four of my climbing mates along so that we could actually get high. So we had expeditions to K2, Masherbrum, the Trango Towers, Biale and Hushe, all during the late '80s. And I've been to the Hindu Kush several times, along the Pakistan-Afghanistan border. It's always a bit dangerous crossing into Afghanistan. That's where the Panjshir Valley fighting was going on with the mujahideen in those days. And I've also worked around the Tibetan Plateau for six or seven field seasons.

Oliver Strimpel: We're going to talk in a moment about three of the ranges that you've studied in more detail than the others. But first, I wonder if you could just briefly put into context the origin and nature of the ranges that we won't be talking about in detail, which are the Tian Shan, the Kunlun Shan and the Gangdese.

Mike Searle: Yes, sure. Well, the Tian Shan and the Kunlun are older mountain ranges. They were formed during the middle or late Paleozoic, so we're looking at somewhere around 300 to 400 million years ago. But their recent uplift has been attributed to the far-field effects of the India-Asia collision within the last 40 to 30 million years. Like all of the mountain ranges of Asia, the Gangdese range is different. This is a huge Andean-type range that runs right the way across southern Tibet, and it's almost entirely made up of subduction-related I-type granites: hornblende-biotite-bearing granites and granodiorites, and andesitic volcanics as well, which span a long time, over 150 million years. These were formed above the northward-dipping subduction zone that was consuming Tethys. So it was a geological setting very much like the Andes today.

Oliver Strimpel: Could you explain what I-type granites are, and andesitic volcanics?

Mike Searle: Yes. I-type granites are ones that are derived from melting an igneous protolith, so they have a very distinct mineralogy and geochemistry from granites that are formed by continental crustal thickening, like in the Himalaya. The andesitic volcanics are very much like the volcanoes you see today in Washington State and Oregon. Mount St. Helens is a classic andesitic volcano, and that is sitting above the Sierra Nevada batholith, which is very similar geochemically to the Gangdese batholith. So we can say with a fair degree of certainty what the geological setting of the Gangdese range was, and it would have been something very similar to what we see in the Andes, or even the stretch from California up to British Columbia.

Oliver Strimpel: OK, let's talk about the mountain ranges that you have studied in some detail, starting with the Karakoram. What are its salient geological features, and how do they compare to the Himalaya?

Mike Searle: The Karakoram is relatively complicated compared to the Himalaya. The Karakoram has three major terranes. The northern Karakoram is composed dominantly of Paleozoic to early Mesozoic sedimentary rocks, so these are rocks that span from about 350 million years up to around 200 million years. They are intruded by some older dioritic plutons. For example, the most spectacular one you see is on the west face of Gasherbrum IV, this spectacular mountain that you see as you trek up the Baltoro Glacier. The central Karakoram is all composed of this amazing backbone of the granite batholith. And the southern Karakoram is all a high-grade metamorphic terrane. These have classic Barrovian facies: kyanite and sillimanite gneisses. Kyanite, sillimanite and andalusite are all aluminum silicates, and they're all the products of crustal thickening and regional metamorphism.

Oliver Strimpel: Why is it called Barrovian?

Mike Searle: That's named after George Barrow, who was one of the first geologists who traipsed over the Scottish Highlands, and he first came up with the Barrovian sequence of metamorphism. This was over 120 years ago. At least, he was the first person to really say that there was this distinct mineralogy: as you increase pressure and temperature, you get a distinct sequence of rocks. As you go from low-pressure-temperature greenschist facies to higher-pressure-temperature amphibolite facies, you go through a series of metamorphic isograds. An isograd is where the first incoming mineral forms in that terrane. So you go from chlorite to biotite to garnet, that's all greenschist facies, and then staurolite, kyanite, sillimanite, which is the amphibolite facies. And at the top of the amphibolite facies, when you're getting up to temperatures in excess of about 700°C, you start melting the rock to produce a migmatite, and that's the first incoming melt that forms in that sequence, when temperatures become high enough to melt the rock.

In the Karakoram, the most unique thing is this incredible central zone, the whole backbone of the Karakoram. This is an enormous granite batholith, the Baltoro granite, which is composed of S-type granites, granites derived from melting sedimentary rocks. They have a very distinct mineralogy, totally different from the Gangdese. The Baltoro granites have garnet, tourmaline, muscovite and biotite, and they form some of the most spectacular peaks along the Baltoro Glacier. I'm sure if anyone knows anything about the Karakoram, they've heard of the Trango Towers, which are these two- to three-kilometer vertical walls of granite. Well, these are made of the Baltoro granite, and they have very high strontium-87/86 ratios, which is indicative of crustal melting. And the uranium-lead zircon ages span around 21 to 13 million years, so that's early Miocene. Bizarrely, they happen to be exactly the same age as the Himalayan granites on the other plate to the south, the High Himalayan leucogranites.

Oliver Strimpel: Can we then infer that they are in some sense the mirror image, on the Eurasian plate, of what was going on to the south on the Indian plate during the collision?

Mike Searle: Well, yes, temporally speaking, they were mirror images. It's sort of incredible that although the structure of the Karakoram and the Himalaya is very different, following the collision you've had the same 30 to 40 million years of crustal thickening and regional metamorphism producing these sillimanite- and kyanite-grade gneisses and migmatites. And you've also had the end product of that, which is the crustal melting to form the granites. But the big difference is that the Baltoro granites are absolutely enormous, and there's no doubt that it was a continuous batholith, whereas the Himalayan ones were all these sort of individual plutons and sill complexes, produced from a regional migmatite. So the structure is very different. They're both pure crustal melts, but the Baltoro requires an extra heat source to explain the massive volume of melt.

Oliver Strimpel: Do you think that heat must have come from the mantle?

Mike Searle: Yes, it almost certainly came from the mantle, and we can get an idea of the involvement of the mantle from the very scarce lamprophyric dikes that intrude around the Baltoro. Lamprophyres are these unusual mantle-derived melts; they're basaltic-type rocks. They have ages in the Karakoram that overlap exactly with the granite, so they're all early Miocene, suggesting that there was heat coming out of the upper mantle at the same time as the melting of the lower crust. Also, in the Pamirs, which is the northern part of the Karakoram extending off into Tajikistan, there is a spectacular series of ultrapotassic volcanics and syenite intrusions, which are very similar geochemically and temporally to the lamprophyres that we have in the Baltoro, and they all have ages of around 11 million years in the Pamirs. These mantle-derived intrusions are part of a huge range of similar rocks intruded throughout the plateau of Tibet.

Oliver Strimpel: How do we know that the lamprophyric dikes and the ultrapotassic syenite intrusions originate from the mantle? Is it because of their chemical makeup?

Mike Searle: Yes, they're geochemically very distinct. They're highly alkaline, they're silica-undersaturated, so they're at the opposite end of the scale to granite, and they're ultrapotassic, so they're indicative of a highly enriched upper mantle source. Syenites are like granites; they're coarse-grained rocks with alkali feldspar. They form in thick continental crust by low degrees of partial melting, and from the geochemistry, they have to have a mantle component.

Oliver Strimpel: Do we have an idea why that extra heat from the mantle was injected in the Karakoram and not elsewhere?

Mike Searle: Well, that's a very intriguing question, and we don't really have an answer to it. We know from the concentration of these highly alkaline lamprophyric rocks that they are an important component of many orogenic belts. They usually intrude at a very late stage, so whatever happens during the continental collision, the mantle has an anomalous, highly alkaline component right at the end of that orogenic process.

Oliver Strimpel: How does the Pamir relate to the Karakoram range, which lies to its south?

Mike Searle: Well, the Pamir generally shows older geology, more Paleozoic and lower Mesozoic rocks. There's nothing as young as we see in the Himalaya. But it also shows very large-scale metamorphism after the India-Asia collision, similar to what we see in the southern Karakoram and in the Himalaya. The pressure-temperature conditions are very similar in all three mountain ranges, and the uranium-lead monazite ages, which give you an indication of the timing of peak metamorphism, are again very similar to the Karakoram and the Himalaya: early Miocene. They occur in a series of what we call compressional core complexes along the southern Pamir, and they are directly related to the metamorphism in the southern Karakoram.

Oliver Strimpel: Isn't it a bit surprising that the metamorphism in the Pamir reflects similar pressure and temperature conditions to that of the Karakoram, even though the Pamir, being further north, is much further from the collision zone with India?

Mike Searle: Well, it's not that surprising, because the southern Pamirs are geologically part of the same continental plate as the Karakoram. There's a big suture that runs right through the central Pamirs. So the northern Pamirs are part of a different plate, but the southern Pamirs and the Karakoram were contiguous, on the same plate.

Oliver Strimpel: OK, let's talk about the Hindu Kush.

Mike Searle: Well, the Hindu Kush is really the western extension of the Karakoram. It runs from northern Pakistan, and then it swings south along the Afghan border. And although it's on a similar geological plate to the Karakoram, it has a very different geology. Most of the rocks in the Hindu Kush are, again, older sedimentary rocks, Paleozoic to early Mesozoic. They have very few collision-related granites. Whereas the Karakoram has enormous volumes of Baltoro granite, there's only really one intrusion, near Chitral in northwest Pakistan, which is a Miocene post-collisional granite. The rest of the granitic rocks up there are older, deformed and metamorphosed amphibolite rocks, which started life as I-type granites, very similar to the Gangdese, but they've undergone huge metamorphism and deformation after intrusion.

The Hindu Kush also, interestingly, has the world's deepest known continental seismic zone. So despite having this old geology and not a lot of indications of crustal melting after the collision, it has this continental seismic zone in which earthquakes extend down to about 250, even 300 kilometers. These earthquakes define a very narrow zone. It is the deepest continental seismic zone we know of on the planet. There's a second seismic zone to the north, called the Pamir seismic zone, which defines a shallower, south-dipping zone. So you have two converging continental subduction zones, the deep Hindu Kush and the shallower Pamir, and it's probably the convergence of those two subduction zones that lifts up the whole of the Pamir Plateau.

Oliver Strimpel: Wow, that's very interesting. Continental subduction is fairly rare, and here you have two of them going on, with opposite polarities, right in the same part of the world. How can continental crust subduct to these enormous depths, given that it's a lot lighter than what it's subducting into?

Mike Searle: Yes, well, good question. Most deep seismic zones are in the oceans. All around the Pacific Ring of Fire, you have very deep oceanic subduction zones, which is the subduction of oceanic lithosphere. But in the continents, it's very uncommon to get these deep earthquakes. Along the Himalaya and south Tibet, the deepest earthquakes occur at around 70, maybe 80 kilometers depth. So there are several factors you need to make this deep seismic zone. You need rapid subduction, you need old, cold material being subducted, and you need to keep the temperatures low enough to make the earthquakes. Earthquakes only occur at temperatures less than 500 to 600°C. So how on earth can you get low temperatures 250 kilometers deep in the mantle? The answer is that the only way you can do it is by this very narrow subduction of old, cold continental crust, and that is most likely going to be lower Indian crust subducting northwards.

We know that in the geological record there are over 15 or 16 ultrahigh-pressure terranes. They're all from the Cambrian onwards, and they occur on all the continents. So we do know that it's possible to subduct continental crustal rocks to these depths. The rocks formed down there are, first, coesite-bearing eclogites (coesite is a high-pressure polymorph of quartz), and others contain microdiamonds. We see diamond-bearing eclogites in places like the Western Gneiss Region in Norway, and we know they occur in subducted continental rocks, and that is exactly what's happening under the Hindu Kush today. The other thing you need to do, apart from getting them down there very quickly, is to exhume them very rapidly, to get them back to the surface. So you do need some sort of slab break-off. You need to release the anchor that has dragged them down to mantle depths, so the buoyancy forces take over and these rocks pop straight up to the surface.

Oliver Strimpel: If these ultra-deep earthquakes in the Hindu Kush stem from the subduction of lower Indian crust, wouldn't we also expect to see such deep earthquakes further east, under the Karakoram or Tibet?

Mike Searle: Well, that is a very good point, and the answer is, we don't see that. People have speculated on this for a while. In the central part of the Himalaya and Tibet, we don't see this deep structure. What we do see is the lower Indian crust, which is all the Archean Indian shield, underplating the Himalaya and underplating the Tibetan Plateau. This was a mechanism that Émile Argand, the great Swiss geologist, suggested over 100 years ago: that rather than a deep subduction zone, you have a much shallower underplating of lower Indian crust underneath southern Tibet. So there's a kind of bifurcation between the mode of subduction you have under the Hindu Kush, at the western end, which goes very deep, and the shallow underplating of the Indian plate further east, which just slides underneath.

Oliver Strimpel: Do we have any idea why we get these two different modes, and what causes the Indian plate to do one at one end and the other at the other?

Mike Searle: Well, there's something very special about the two syntaxes, which are the northwest corner and the northeast corner, where the Himalaya goes from east-west to north-south. All of the structures swing around through 90 degrees, down through the Chaman ranges of the Pakistan-Afghanistan border on the west and down through the Indo-Burman ranges on the east. And at both of those syntaxes there are 7,000- and 8,000-meter mountains, which have the youngest metamorphism known anywhere on the planet. There's the Nanga Parbat metamorphism. Nanga Parbat is this 8,000-meter peak in the northwestern part of Pakistan, which is actually the western end of the Himalaya, rather than being in the Karakoram. And in the northeast is a mountain called Namche Barwa, which also has extremely young metamorphism. In both of these areas, the rocks were formed less than one to two million years ago and were exhumed extremely rapidly. So you have this enormous exhumation of deep crustal metamorphic rocks in a very short time. And the reason, we think, is structural. The rocks in the main part of the Himalaya are extruding to the south by channel flow, but in the syntaxis regions, they have compressive stresses from everywhere, so there is nowhere else to go but straight upwards, and that's exactly what we see.

Oliver Strimpel: Although India collided with Asia about 50 million years ago, it continues its northward advance to this day, albeit at about half the pre-collision speed. How is this movement accommodated, and is that part of the story of the ongoing uplift of the mountain ranges we've been discussing?

Mike Searle: Well, yes. There are two theories of how the northward push of India was, and continues to be, absorbed in Asia. Firstly, there was the one first proposed by Émile Argand in 1924, where he suggested that India was underthrusting the whole of the Tibetan Plateau and jacking it up. The second model is that of continental extrusion, and this was first proposed by Peter Molnar and Paul Tapponnier back in the 1980s. It was based on Landsat photography, when that first became available. When you laid out all these Landsat photos of Tibet and Central Asia, the thing that really struck you immediately was these huge strike-slip faults, San Andreas-type faults, but going on for thousands of kilometers across Central Asia. So Molnar and Tapponnier proposed that the thickened crust of Tibet was being squeezed out to the east to make way for India pushing towards the north. At the time, they wanted over 1,000 kilometers of offset on these bounding strike-slip faults.

Oliver Strimpel: So what observations can help us distinguish between these two models?

Mike Searle: Well, the continental extrusion model is very easily testable, because these strike-slip faults are exposed at the surface, and all you need is a pinning point: a well-dated granite or sedimentary rock, or anything that you know was once continuous and is now offset by some amount of slip. When this extrusion model was first proposed, there were no geological maps; there were no constraints on the amount of slip on any of these faults. So we went into the Karakoram Fault back in the 1980s and carried out very detailed work along the Karakoram Fault in Ladakh, where it's beautifully exposed, with a whole series of granites. By getting very accurate uranium-lead zircon ages on each one of these granites, we were able to map the dextral offsets. And we found that the offset of these Miocene granites was no more than 120 kilometers on the Karakoram Fault, one of the most active strike-slip faults of Tibet. That happens to be exactly the same as the offset of the course of the ancient Indus River, which was captured by the fault and dragged for 120 kilometers before it cuts south back into northern Ladakh.

Oliver Strimpel: Well, that would seem to be pretty conclusive evidence. Do the proponents of extrusion tectonics accept these arguments?

Mike Searle: They have to, basically, because it's good, proper science. But the problem was on the other faults, the Kunlun and the Altyn Tagh, which are nothing like as clear-cut as the Karakoram. There were no definitive pinning points, and you're more reliant there on the extent of the ductile slip along those faults. So the take-home message of all of this is that continental extrusion was working as a process, but nowhere near to the extent that Peter Molnar and Paul Tapponnier originally proposed. Each one of these faults had more like somewhere between 150 and maybe 300 or 400 kilometers maximum.

Oliver Strimpel: So if we have this lower amount of extrusion going on, the rest of the northward movement of the Indian plate is accommodated by this flat-slab subduction, and by uplift and folding throughout the crust to the north.

Mike Searle: Yes, and all of the Himalaya indicates hundreds of kilometers of homogeneous crustal shortening. That's folding and thrusting that shortens the crust and thickens the crust, and then you get the metamorphism. But all of the rocks in the Himalaya are Proterozoic or younger; there's nothing older than about 1,000 million years. And we know that the lower crust of India is dominantly the Precambrian Indian shield. These are Archean rocks, which are much, much older. They're very old, high-grade, granulite-facies rocks, so they're dry and they're competent. It's the shallow subduction of the lower part of the Indian plate that is jacking up the Himalaya, and it's jacking up southern Tibet, if not most of Tibet itself. That active fault is the one on which all the big earthquakes along the Himalaya occur today. The Main Himalayan Thrust is where, for example, the Gorkha earthquake of 2015 set off, and the Great Assam earthquake. All of those are on the southern boundary of the Himalaya, which is where the Indian shield is subducting beneath the Himalaya and jacking it up.

Oliver Strimpel: Do other mountain ranges, such as the Alps or the Andes, have the kinds of satellite mountain ranges that we've been talking about today with respect to the Himalaya?

Mike Searle: Well, all mountain ranges are different. The Alps are part of the Tethyan collision zone. They're the same as the Himalaya temporally, but they show totally different structures. The Alps have much more high-pressure and ultrahigh-pressure rock than we see in the Himalaya. There's very little regional metamorphism in the Alps, whereas in the Himalaya it's huge; most of the Greater Himalayan sequence is all that. And the Alps have almost no crustal-melt granites associated with the young collision zones, so there's nothing like channel flow. The Andes are completely different from the Himalaya. The Andes show a geology very similar to the Gangdese: I-type granites related to the subduction of oceanic lithosphere beneath the continent, big explosive andesitic volcanoes like Mount St. Helens, and, typically, porphyry copper deposits. So yes, all mountain ranges are different, which makes geology all the more interesting.

Oliver Strimpel: Mike Searle, thank you very much.

Mike Searle: Thank you, Oliver. It's been a pleasure.

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