A refolded fold from Scotland

Standing on the shores of Loch nan Uamh I was feeling distracted. There was a lot to attend to. Behind me was a flat strip of grass, growing on a beach deposit now left high and dry by the crust’s slow straightening of its spine after the weight of a huge ice cap melted away. Perched above the fossil beach was a Victorian country house where incredibly brave people from across Europe had trained before being dropped into Nazi-occupied Europe to wreak righteous havoc.  To my left sits an Iron age hill fort and straight ahead a sublime view demanding I pay attention to its wild empty peninsulas, small forested islands and wide rolling sweep, all under a cloudless sky. The drift of my thoughts towards why this landscape is so wild – heartless landlords and ‘improvement’ leading to clearance and gaelic songs sadly sung on a ship to Canada – is interrupted by my family. Not unreasonably they want me to stop staring into space and attend to their games on the rocky beach.

Fatherly duties still leave time for me to assume the geologist’s pose and stare at the ground, assessing the cobbles at my feet. I snaffle a nice bit of granite first. Then a nice piece of pinkish folded quartzite catches my eye and slips into a pocket to be forgotten as the glorious day rolls on.

Let’s have a look at it.

You can see how it caught my eye. Just short of a billion years ago it was formed as different layers of sand, some nearly pure quartz, others more muddy. It’s since been heated and squashed within the earth. The quartz is still quartz but the mud is now shiny mica.

What were once flat layers 1  are now folded. The pattern we see is the intersection of a 3D fold with the complex rounded surface of the pebble. Call the image above as a view of the top of the pebble. Let’s rotate it round (it sits beautifully in the hand, like a heavy cricket ball) and look at one of the sides.
These sigmoidal shapes are classic folds. You can almost visualise the rock sitting in a clamp being squashed horizontally to turn flat lines into wavy.

Rotate the sample 180 degrees around  a vertical axis to see the other side of the sample and you see the same folds coming out of the other side of the rock.

Looks simple. But appearances can be deceptive. Let’s rotate the sample to look at its underside (the side my thumb is holding in the picture above).

At the top of this photo you can see some of the edge in the preceding picture. So the folds we’ve seen before have their crests and troughs running vertically through the image. You can see that, but not just that. The darker layer sitting in the middle of the shot runs from top to bottom, marking a trough in the folding, but it is forming a rough cross shape, not a single line as I would expect.

There are two sets of folding in this sample. I’ll try and annotate that for you.

In red I’ve drawn most of the troughs and peaks of the main folds. The ones we’ve seen now from three angles. The blue line shows what I reckon is an earlier fold, where the crest of the trough is itself folded by the red folding.

Here’s a view looking at the side again, the edge on the right side above. It’s at an angle of 90 degrees to the other side views. We are looking down on a trough shape made by the earlier blue folding, now bent around by the red folds.

Here are some oblique views in the same direction, that makes the saddle shape formed by the two sets of folding more obvious. This stuff is hard to see from photos. It’s easier to visualise the three-dimensional patterns when you’ve got the sample in your hand, but even then I didn’t notice (distracted as I was) the full complexity when I first picked it off the beach.

The rock sample is probably from the local Moine rocks, where refolded folds are common. There’s a bigger and more complicated example in the hills above where I collected this a tale of past oceans opening and closing, lost continents forming and splitting. But we’ll come to this in a future post.

Categories: Scotland, tectonics

Volcanoes and mass extinctions – tracking a killer

Look in a bookshop and see how many shelves are taken up with murder mysteries. There’s little that is as compelling as the idea of a dead body on the ground and a search to find the culprit. I’m going to try out the genre here today. I can promise you the deaths of entire species, a glamorous prime suspect with spectacular methods and an overlooked serial killer who has poisoned many different victims. I can’t promise you detectives who are troubled mavericks who break the rules, but there are geologists who sometimes feel like they are underdogs.

The dinosaurs were killed by a giant impact. There’s little debate in the public mind about that and the role of extraterrestrial impacts on earth’s history is now inarguable (I’ve written about it myself). Sometimes though it irritates me how much focus is put on speculation about extra-terrestrial causes for mass extinctions. The worst example I’ve seen is speculation that dark matter (that we don’t understand) has caused past extinctions. Glamorous ideas about Death From Space, which (with the exception of the Cretaceous-Palaeogene) event have little supporting geological evidence always seem to get attention.

This makes me grumpy because geologists have a perfectly good explanation already. A serial killer stalks earth’s history. It doesn’t kill by flaming impact (or however dark matter is meant to work) but by poisoning, choking the life out of countless plants and animals. Death From Above is spectacular but Death From Below, a murderous force rising slowly and unstoppably from the Earth’s core is much creepier.

Forensic evidence

Mass extinctions leave an unusual sort of murder scene. Instead of there being a single dead body, there is a sudden lack of them, as fossils of particular species disappear from the geological record. For a normal murder, you would study the body for any clues, any evidence of what killed it. Same with a mass extinction, only you look in the layers of rock round about where the fossils run out.

These need to be places with a continuous sedimentary record, where we have sediments from the age of the extinction. Often these are marine sediments, which can contain relatively large volumes of small fossils. Chemistry is the best form of forensic evidence as it gives us an insight into the state of the ocean over time. Carbon isotopes track the ebb and flow of the Carbon cycle and often the extinction horizon is associated with a sudden change in them. This means that rocks from the time of the extinction event can be found even in layers with few fossils.

The K-Pg event (RIP non-avian dinosaurs, plesiosaurs, ammonites), is famously associated with a layer rich in Iridium, an element rare on the earth’s surface but much more common in material in space. Similar connections are found between other extinctions and Zinc. The P-Tr event (RIP trilobites, nearly everything else) shows a spike of Zinc in marine sediments immediately before the extinction. A recent study (Liu et. al 2017) also shows how the isotopes of Zinc change over time. Zinc is an important nutrient for marine phytoplanktons, meaning their growth changes the isotopic ratio of Zinc in marine sediments. Using this they demonstrate not only that more Zinc is found, but that it came from volcanic or igneous material entering the ocean. This happened abruptly around 35 thousand years before the extinction event. Soon after, the ratio shifts back in a way consistent with phytoplankton activity returning to normal within 360 thousand years.

Liu figure 2a

Figure 2b from Liu et al. Showing Figure 2b from Liu et al. Showing changes in Zinc concentrations and isotopic ratios immediately before the Carbon isotope changes associated with the extinction event.

Other studies show anomalous peaks of Nickel abundance just before the P-Tr event in many sections across the world. Once again the source is inferred to be volcanic activity. Different sets of forensic evidence point to an obvious suspect – the Siberian Traps – an enormous area of volcanic rocks covering a huge area of Russia that was formed across the P-Tr boundary.

The Murder weapon

Volcanic eruptions are dangerous to be near. It’s obvious why life suddenly swamped by lava will not survive, but a mass extinction is a global phenomena. How can a volcanic area kill animals or plants on the other side of the world?

Jerram figure 1

Figure 1 from Jerram et al.

Figure 1 from Jerram et al. showing extent of Siberian Traps, highlighting sill intrusions, coal and explosion pipes.

One clue comes from odd structures found around the Siberian Traps, for example within the Tunguska Basin1. These structures are pipes called diatremes, formed by gaseous explosions.

Some diatremes are formed by gas ready mixed within the magma, but with these Siberian pipes the gas came from heating of the sedimentary rocks that were already there. Buried below the many lava flows, are flat sheets of rock called sills that pushed between existing sedimentary layers. These sills heat up the surrounding rocks, which in the Tunguska basin include much coal and evaporite rocks. This heating produced vast volumes of CO2 and CH4 that poured out of the pipes into the atmosphere.

Figure 2 from Polozov et al. Showing portions of basaltic pipes, exposed within mining works.

Figure 2 from Polozov et al. Showing portions of basaltic pipes, exposed within mining works.

These gases of course affect the climate. A huge outpouring of CO2 and methane, plus also nasty gases such as SO2 represent a pretty convincing murder weapon. A brand new paper demonstrates malformed parts of terrestrial plants about this time, attributed to pollution. Sudden ocean acidification and climate change followed by a collapse in planktonic growth leading to the the death of the dependent food webs is a uncontroversial story. It may be a story we are beginning to retell as gas forms mysterious holes in the ground in Siberia once more.

Killed and killed and will kill again

The Siberian Traps are just one of many Large Igneous Provinces (LIPs). Other ones are also associated with extinctions. The pleasingly named CAMP province, found in Atlantic facing areas of Africa, Europe and the Americas, overlaps in time with end-Triassic mass extinction event (RIP conodonts and various reptiles & amphibians).  An open-access paper from mid 2017 demonstrates a link between sills intruding into organic-rich sediments and the extinction event – exactly as proposed for the P-Tr event.

Figure 1 from Davies et al

Figure 1 from Davies et al

The bodies are piling up. The Late Ordovician extinction event (RIP 85% marine species, no big groups) has been linked to Mercury enrichment in marine sediments. The authors link this to a LIP, even though one of this age has not yet been found.

What do we know about this serial killer? LIPs are thought to be formed by huge plumes of rock, rising from the edges of odd features on the very floor of the mantle. Their ability to kill may depend on the nature of the crust they rise into. Both the P-Tr and Tr-J events see sills intruded into sediments. The Deccan Traps, active across the extinction of the dinosaurs (K-Pg) rose through basement rocks, so there were no sediments to heat, meaning they pumped out only volcanic gases. Maybe this is why the extinction event required an impact to finish the job.

Case for the prosecution

“So, ladies and gentlemen of the jury, I suggest to you that the accused is a serial killer. The mighty plesiosaur, the ever-busy scuttling trilobite, even the wriggly conodont, all were killed by the monster sitting before you. It killed, not by showy eruptions and square miles of lava, but by the silent injection of sheets of magma deep underground. This devilish act then poured huge quantities of poison into the air, bringing the very earth to its knees.”

We’re not quite ready for a trial. Some of the evidence is circumstantial and we certainly don’t have a full roster of victims. But LIPs should be high on the list of anyone’s list of suspects for the greatest murders the world has ever seen.

Categories: impacts, sediments

Ultrafast eclogitisation through overpressure

My blogging torpor has been ended by a super-interesting new paper that links together many of my favourite topics. It includes eclogites, metamorphic petrology, ultra-fast metamorphism, determining timescales via diffusion profiles, tectonic overpressure and even the Grampian-Taconic orogeny and opens up new avenues of research. What more could I ask for from a scientific paper?

The Rocks

The paper, “Ultrafast eclogite formation via melting-induced overpressure” by Chu et al. has just been published. It’s based on a very close study of a set of rocks from Connecticut in the USA. They are part of the Taconic orogeny, (also seen in Connemara, Ireland, my old field area) which was caused by an Arc colliding with a continent.

The particular location is within a tectonic slice and consists of felsic paragneisses (metamorphosed muds and sandstones) containing lenses of mafic and ultramafic rock (metamorphosed gabbro). The gneisses are migmatitic – they were partially melted. The whole area of high-grade rocks is bounded by mica schists, within 50m, suggesting the heating of the rocks was localised.

The mafic lenses are eclogitic – they contain an eclogite facies assemblage of minerals that formed under high pressure.  As is common in eclogites, some of the peak minerals have retrogressed (changed) back into fine mixtures of lower pressure minerals called symplectites. The phengite and omphacite are now symplectite pseudomorphs but garnet (Twitter’s 5th most popular mineral) remains itself and contains some invaluable information.

The chemical analysis

The paper describes estimates of the conditions of metamorphism, derived from various forms of analysis. The garnets are the key to this, as they show a wide range of different compositions. Since they grow outwards, the change of compositions from the core captures changing metamorphic conditions as the garnet grew. The image below is a map of the amount of particular chemical elements within the garnet. The different colours represent different amounts and since the chemistry of the garnet depends on the pressure and temperature while it grew, this gives us a lot of information.

Part of figure 4 showing sharp and complex chemical zoning in garnet

Part of figure 4 showing sharp and complex chemical zoning in garnet

As marked on the image above, parts A-C record heating at around 8 kbar (typical crustal depths) from 660 to 790°C, then D represents eclogitic conditions of 14 kbar and E a return back to 8 kbar.


Garnets that have spent a long time at high temperatures lose their chemical zoning due to diffusion. Since we understand the speed at which diffusion works, we can use chemical zoning in minerals to estimate the speed of metamorphic processes. Look at the images above and you’ll notice that the boundaries between the different domains are quite sharp. Using two sets of forward modelling (one of the diffusion profiles and another of the garnet growth) they conclude that the pressure increase (C-D above) took only 500 years. The pressure decrease (D-E above) also took only 500 years. These are not the sort of timescales geologists are used to dealing with. I’ve been to educational establishments that have lasted longer than 500 years – it’s no time at all.

These dates cannot be explained by normal models of eclogite formation, which involve subduction of material and eventual return to shallower depths. Rates of subduction are measured in cm per year, but to explain the rate of pressure increase this way would require rates of 30m per year.

How does pressure change so fast?

The usual assumption in metamorphic petrology is that the pressure recording by the minerals is lithostatic pressure – that caused by the weight of rocks above. This is useful as it allows you to track the rocks passage into and out of a subduction zone or mountain belt and so link metamorphic P-T estimates to tectonic models. A number of recent papers have started to challenge this assumption. Their approach varies, but all suggest that rocks can experience high pressures without being buried to great depth.

What’s so great about Chu et al. is that is provides direct evidence of this happening and offers an explanation as to how it happened.

Burying rocks at 30m a year is not credible, so if we accept their estimates for the speed and amount of pressure increase (it persuades me, but more knowledge people than me may disagree) we must look for ways of generating high pressure without piling up more rocks on top.

Chu et al.’s explanation takes us back to the unusual location of these rocks, within a narrow active shear zone surrounded by colder rocks. They see rocks as being rapidly heated due to strain heating along the shear zone. This one of the mechanisms invoked by those arguing that localised and rapid metamorphism is more common that we thought. The narrow zone of hot rocks is surrounded by colder more rigid walls. As the rocks start to melt, the sudden volume increase causes an ‘autoclave’ effect, a sudden local deviation above lithostatic pressure. This causes the rapid growth of eclogitic minerals with the mafic pods. Soon the melt escapes or crystallises, reducing the pressure back to normal.

Part of figure 9, showing the model of transient overpressure due to partial melting

Part of figure 9, showing the model of transient overpressure due to partial melting

What does it all mean?

This paper adds a fascinating petrological argument supporting the theoretical studies suggest that overpressure within metamorphic rocks is possible. No longer can the mere presence of high-pressure minerals necessarily indicate the rocks was buried to great depths.

As the authors say, it does not invalidate most of what we know about eclogites. The mechanism described does not apply to the many ultra-high pressure terranes like Western Norway where large volumes of eclogitic rocks exist – these surely did get stuffed deep into a subduction zone.

My first thought was of the Glenelg eclogites in Scotland. This is a small area of eclogite within a high strain zone which makes very little tectonic sense. Perhaps this is a candidate for being formed by local overpressure? As far as I can tell they aren’t associated with migmatites, which makes them not directly analogous. But the point is that there is a new possible interpretation for them, new avenues of research. That’s what makes this paper so interesting to me.

Please read it, and if you think it’s wrong, let me know.

Chu, Xu, et al. “Ultrafast eclogite formation via melting-induced overpressure.” Earth and Planetary Science Letters 479 (2017): 1-17.

Update: another possible candidate for this effect in Scottish rocks is within the Central Highlands Migmatite Complex. See brief description and my write-up of a discussion of the significance of these rocks.

Categories: eclogites, metamorphism, tectonics

The Himalayan mountains: flow and fracture

Earth science departments are home to three styles of working, each of which tries to answer similar questions, but from very different perspectives. First we have the field geologists. Armed with field gear and a hammer, they gather data from actual rocks in the form of photos, diagrams and above all maps. Next we have those who live in the lab studying rock samples using mysterious machines to gain startling insights. Finally those who seek to recreate the world in a computer, building models to understand the underlying equations that can explain how the earth works.

In some departments the very different cultures and mental models that these approaches require can lead to conflict. But the best science is done when they are combined, in a department, research group or even a single brain. A recent paper by lead-author Andrew Parsons weaves together different techniques to provide a coherent model of how rocks flow to build huge ranges of mountains.

Different ways to build a mountain range

The Himalayan mountain range is the best place to study how mountains form. Sixty million years ago the Tethys ocean sat between the Indian plate and the Asian. Soon the Tethys oceanic crust vanished down into the mantle beneath Asia and the Indian plate collided with Asia. At the site of that collision now sits the highest mountains on earth, plus the huge and high Tibetan Plateau.

Geologically the Himalaya are long (over 2000km), thin and consistent. From Kashmir and Ladakh on the Indo-Pakistan border, through all of Nepal and further East through the Buddhist kingdom of Bhutan the same sequence of rocks is seen. Starting from the undeformed Indian Plain and walking up, we pass through the sub-Himalayan zone, through the Lesser Himalayan Sequence (LHS), the Greater Himalayan Sequence (GHS) and finally the Tethyan Himalayan Sequence (THS).

These distinctive packages of rock are separated by major faults. Sudden movements on these faults cause the major earthquakes that regularly affect the people who live in this beautiful part of the world. Early models for how the mountains formed focused on these thrust faults. We saw mountains as thick piles of slabs of rock, separated by brittle faults.

Thrust faults stack up layers of rock. Image from Wikipedia

Thrust faults stack up layers of rock. Image from Wikipedia

In these models deformation is focused on the thrust surfaces and the rocks in between are relatively strong and rigid. This is a reasonably good description of the rocks of the LHS. However studies of the GHS rocks showed them to have been strongly deformed at high temperatures. Intrusions of granite are common and appeared to form while the GHS rocks were deforming and moving rapidly towards the surface. Also the fault between the GHS and the barely deformed sediments of the THS is a normal fault, one with an opposite sense of movement to the thrust faults.

View of Mount Everest, with

View of Mount Everest, with sediments of the Tetyhan Himalayan Sequence forming the summit and metamorphic rocks of the Greater Himalayan Sequence below

To explain these features, the concept of ‘channel flow‘ was borrowed from fluid dynamics. This describes how viscous fluids flow when trapped in a channel between two rigid surfaces (think of jam/jelly squeezed out of an overfilled sandwich). We know that hot rocks deep in the earth can flow (slowly!) even while remaining solid. The channel flow concept saw the GHS rocks as flowing out from underneath the Tibetan Plateau towards the surface, perhaps moving into the space created by rapid erosion of the high Himalayan mountains.

An alternative model for explaining the hot and deformed GHS rocks is called wedge extrusion, where deformation on thrust faults brought it to the surface, rather than channel flow.

These three models are often seen as being mutually exclusive. But what if all were true?

Going with the flow

Andrew Parsons work is based upon his PhD studies at the University of Leeds in England. He’s published a map of the Annapurna region of Nepal, covering all the main sequences of rock and so is at home in the fieldwork tradition of the Earth sciences. No doubt he has pictures of himself in field gear, looking sun-burnt and grinning in front of some amazing scenery. At the core of his prize-winning paper is hours and hours of lab-work. After wielding his hammer to collect over 100 samples in a transect across the Himalayas, he had them cut by a diamond saw into thin sections to have their secrets probed.

His studies used an optical microscope (to get a sense of this, check out #thinsectionThursday on Twitter) and also a scanning electron microscope. The goal was to identify precisely how the minerals in the sample were squashed.

Rocks are made of minerals and the way they flow is by deforming those minerals. We know a great deal about the many ways common rock forming minerals (like quartz, feldspar, calcite) deform. There are a range of different ways in which the minerals deform. Minerals are crystalline; the atoms within are aligned in consistent repeated patterns. Slow movement of atoms along particular slip-planes or the propagation of defects change the shape of the mineral and so deform the rock. The mineral structure is complicated and different planes of slip are favoured depending on the temperature. Careful study of subtle patterns in the minerals, plus measurement of the average orientation of the atomic structure in each grain tells a great deal.

A portion of figure 6, showing different microscope images of rock samples, with different features highlighted.

A portion of figure 6, showing different microscope images of rock samples, with different features highlighted.

For each sample, the authors were able to estimate the temperature at which the rocks were deformed. Also they got a sense of the style of deformation. Imagine a perfect sphere in a rock that is then deformed. Rocks can be flattened turning the sphere into a cow-pat or M&M shape or maybe stretched into a cigar shape. As well as this, the deforming sphere can also be rotated. A combination of field observations, thin section studies and the scanning electron microscope work together gives a view of how the rocks were squashed.

Bringing it all together

The core achievement of this award-winning paper is linking the mass of data to the predictions of the various tectonic models.  Channel flow is predicted in the GHS rocks while they were hot and rapidly flowing. The entire channel should flow and rotation only be seen at the edges. This is what the samples from the upper section of the GHS show, plus an indication that this style of deformation ended while they were still at 550°C and changed from being hot and soft to being more rigid.

At this point channel flow ceased and another mechanism, rigid wedge extrusion, came into play. Here the lower portion of the GHS was deforming at a lower temperature. As the model would predict, deformation involves a lot of rotation, consistent with the whole of this sequence of rocks acting as a broad shear zone, bringing the upper GHS rocks  closer to the surface.

The lowest temperature deformation is found in a thin zone of rocks within the lower GHS that were acting as a thrust fault, stacking up different slices of rock.

Figure 12

Figure 12

This diagram emphasises the key insight – that different ways of building up mountains can be active at the same time, within different parts of the mountain belt. The mountain building (orogenic) system is composite, made up of different parts. Studying individual mineral grains helps us understand the structure of a massive mountain range. This is because that is how it grew. Countless millions of mineral grains slowly shuffling their atomic lattices over millions of years: this is what builds mountains.

The system is still active, as India pushes into Asia. We know from geophysics that there are hot rocks under the Tibet – perhaps these are right now flowing to the surface as a channel. As the Himalayas are eroded away they will get nearer the surface and start cooling, causing different styles of deformation to come into play.

The terminology of ‘superstructure/infrastructure’ used in this diagram is taken directly from computer modelling work. Numerical models of how mountain belts might work have directly informed this work. Cold rocks near the surface can be a lot stronger than the hot rocks below. By combining computer modelling with field work and lab studies, geologists are getting a good understanding of how the Himalayan mountains formed and have evolved over time. Many places are built on the roots of ancient mountain ranges and high-quality integrated studies such as this help us understand rocks across the world.

Parsons, A. J., et al. “Thermo‐kinematic evolution of the Annapurna‐Dhaulagiri Himalaya, central Nepal: The Composite Orogenic System.” Geochemistry, Geophysics, Geosystems 17.4 (2016): 1511-1539.

Categories: Himalaya, metamorphism, mountains, tectonics