The Science of Everything Podcast - Episode 125: Earthquakes

Episode Date: January 31, 2022

A discussion of earthquakes, covering the elastic rebound theory of earthquake generation, the phenomena of deep and intraplate earthquakes, types of seismic waves, methods of earthquake detection, different ways in which earthquakes cause damage, and attempts at earthquake prediction. Recommended pre-listening is Episode 111: Plate Tectonics. If you enjoyed the podcast please consider supporting the show by making a PayPal donation or becoming a Patreon supporter. https://www.patreon.com/jamesfodor https://www.paypal.me/ScienceofEverything

listening to The Science of Everything podcast, episode 125, earthquakes. I'm your host of James Fodor. So in this episode, we're going to talk about, well, you guessed it, earthquakes. And this is a kind of a follow-up to the previous episode of volcanoes, although that's not a prerequisite, because they're related, but largely distinct phenomenon. And in this episode, we're going to look at, obviously, the causes of earthquakes and sort of key ideas surrounding that and words to describe them. We're going to look at some of the different types of earthquakes and how they differ in terms of cause and effects. We're going to talk about seismic waves, how earthquakes are measured, and there's size magnitude that are described. We're going to talk about the different

forms of damage that earthquakes cause, and a little bit at the end about prediction of earthquakes. Recommended pre-listing is episode 111 on plate tectonics, and that's a fairly strong recommendation because I will talk a bit about plate tectonics and use terminology now and then to explain faults and how earthquakes occur, so some background there would be quite helpful. Right, let's jump in and start talking about earthquakes. So, what is an earthquake? An earthquake is defined as the perceptible shaking of the surface of the earth, which can be violent enough to destroy major buildings and potentially kill large numbers of people.

Now, note that that is defined as something that is perceptible, that is you can sort of detect it, using the senses you can tell that it shakes. Most earthquakes are actually not perceptible. they are so small that we can only detect them using the most sophisticated instruments. But obviously there's a much more of an interest in the perceptible, so significant larger earthquakes. Also, there can be a distinction apart from whether there has to be perceptible or not, in terms of whether an earthquake is defined as a seismic event, that is something that's caused by the operation of tectonic plates,

or whether it can be caused by whether an earthquake is any type of shaking of the ground on a large scale, regardless of the cause. And so this latter, broader definition, would include earthquake caused by phenomena such as landslides or even nuclear bombs. So there are slightly different definitions that you will see, depending on whether it has to be perceptible or not, depending on whether it has to be seismic or not. Most earthquakes are seismic, and so those are the ones we'll be focusing on here. I'll mention other types of earthquakes a little bit. Also, obviously, the focus will be on larger earthquakes rather than smaller ones, but there are similar phenomena in either case.

So before we get stuck into talking about the causes of earthquakes, there's a few sort of basic terms that I want to define, which many of you probably heard some of these, but perhaps don't know exactly what their meaning is. So let's start by talking about a fault. So a fault is a break in rocks or break in a sort of a large segment of rocks along which movement occurs. So they're typically the site of earthquake generation. So earthquakes occur at faults or fault lines, you may hear them described as. Now usually there's not just like one single crack that extends many kilometers, but a series of cracks in a general region that are related but, you know, sort of distinct from each other. And this is called a fault zone, this general region of many distinct cracks in rocks. Now, faults are often below the surface of the ground in the continental crust in some area. The continental crust can go to several dozen kilometers below the surface of the earth, so these can be quite deep. Although in the scheme of things, of course, the size of the crust is, fairly small, but deep, you know, by human standards. Anyway, so faults are often below the surface of the earth. When they do reach all the way up to the surface, then they are described as a fault trace or a false scar. So a fault trace is basically like a crack in the ground. A scar is more like one section of land that's pushed up relative to another leading to a kind of a cliff. So we talk about earthquakes as typically occurring at faults. When we talk about a crack in the rock, probably the word break is better because at least I tend to think of crack as sort of a two-dimensional.

thing. And that's more of a fault trace that you see on the ground is kind of two-dimensional. But the fault itself will be three-dimensional. So it's a large sort of discontinuity between two segments of rock. And again, when we talk about earthquake faults, typically we're talking about fault zones that extend over potentially even hundreds of kilometers, although there will be one particular site where an earthquake is generated, which we'll get to in a moment. And so this fault zone will consist of many individual faults that are kind of linked together or separated by short distances, but are connected in the sense that, or related in the sense that they are both, basically at the interface of two large segments of rock. Now, faults can exist

between, at the boundaries between tectonic plates, but there are also faults within tectonic plates. Faults can be large or they can, or they can be quite small. In this case, we're going to focus on the larger faults, especially between tectonic plates as that's where most earthquakes occur. So moving on from faults, stress refers to internal forces between neighbouring particles in a continuous materials, particularly a solid material. Strain refers to how much that material is deformed. So stress causes strain. Applied forces, which build up as a stress, cause over time the material to become deformed or strained. It's built up stress within rocks over time, which leads to defamation and then ultimately to earthquakes and we'll talk more about that in a moment.

Now the focus, or also called hypocentre, but I'll use the term focus, is the place where rock breakage actually occurs. So it's actually like a physical site, like a point, not like a geometric point, it's not an exact point, but it will be a fairly small region. Although there may be multiple foci of different earthquakes that occur in a region, in a fault zone, right, which may occur within a relatively clustered period of time. because if the whole region shifts, then you're likely going to have subsequent earthquakes occurring as the particular locations of specific faults in different parts of the whole fault zone adjusts and you get smaller earthquakes.

But the focus is the part where rock breakage actually occurs. You probably heard the term epicenter. The epicenter is the point on the earth's surface directly above the focus. We often think about an earthquake as originating at the epicenter, and indeed epicenter is also use more generally to refer to kind of the center where something originates from. In terms of an earthquake, though, earthquakes don't originate from the epicenter. They originate from the focus, which is essentially always underground. And it can be only a little bit underground or it can be many, many kilometers, even hundreds of kilometers underground. So you have to kind of draw a line up

to the surface of the earth, and that is the epicenter. That's obviously relevant from the point of view of understanding the spread of seismic waves, which we'll get to later. And the epicenter is where you would expect the vibrations to be most intense. But it's not where the earthquake actually originates. I mentioned seismic waves. So seismic waves are waves of energy that travel through the earth, either internally or on the surface of the earth, there's different types of them, which result from typically earthquakes, but can also be, as I mentioned before, explosions, volcanoes, or other sources as well. Again, we'll talk more about those different types later. Seismic waves are the mechanism by which energy is, much of the energy is radiated out from the focus.

A final term that you probably heard is aftershock. So an aftershock is a smaller earthquake that follows after a larger earthquake in the same general area. It's generally not going to be at the exact same focus. Indeed, I don't know that that's actually possible. But what you're going to have is perhaps an initial earthquake with one focus at a particular fault where the rocks shift in location to each other. But then that moves the, that causes a motion in a wide area across the whole fault zone

where two plates or regions of cross move relative to each other, which then sort of sets off a cascade of effects in local faults in other parts of the fault zone, which then will build up stresses and then trigger smaller earthquakes in the days, weeks, after the main earthquake, and these are aftershocks. So they're not necessarily directly caused by the initial earthquake, they're kind of indirectly caused in the sense that it's shifted things, and now things have to settle down, and stresses have to be let off and so forth, or released, and you get a series of aftershocks.

The distribution of aftershocks follows known fairly regular laws, and so they typically become smaller and less frequent in inverse proportions of the time after the main shock. And so there's some fairly accurate, at least aggregate ways of describing and predicting those, although not the individual timing of particular ones. So the picture we've got here by kind of outlining these key terms, which we'll be mentioning throughout the episode,

which is why I wanted to find them at the outset. The picture that we've got is that an earthquake occurs when you have two big slabs of rock next to each other with a break in between. The brakes called a fault. Because the rocks move relative to each other, ultimately due to plague tectonics, again see episode 111, stresses build up over time, that forces within the rock itself, ultimately at a microscopic level in the crystalline structure of the minerals in different rock materials. It gradually builds up over time. As the stress builds up, eventually it reaches a critical point where it snaps. And the stress is released when the rock breaks.

And where that occurs is called the focus. The release of this energy at the focus then propagates outwards in seismic waves, which travel away from the focus, and are often most intense at the epicenter, the point on the earth's surface just above the focus. But of course, the seismic waves will be felt a much wider area than just the epicenter. Following an initial earthquake, you will typically experience aftershocks as a result of movement of wide areas of crust leading to then more localized regions of stress building up and then being released in smaller earthquakes as kind of it gradually

settles down and reaches a new kind of equilibrium. So that's the general picture we have. There's much more to say about specifics there, but that sets the stage for some of the more of the detail we'll be discussing in the rest of the episode. So let's now talk about the causes of earthquakes, or at least many earthquakes, in a little bit more detail. Or the general theory that we have for earthquake formation, at least between different tectonic plates, which is called interplate earthquakes, is called Elastic Rebound Theory.

So again, this describes the generation of earthquakes between tectonic plates, which is where most earthquakes occur, but not all of them. We'll talk a bit about some of the other ones later. The general idea is that when you have tectonic plates pressed up against each other, either sort of being forced towards each other, which is compression, being pulled away from each other, which is tension, or sliding past each other where it's a shear, in any of those cases, it produces stress in the rock around those fault areas or the fault zone. So that generates forces. You know, rock is not smooth. It's rough. There's many, there's a complicated, intricate shape that the actual fault will have. It's not exactly going to slide smoothly. Also, there's just friction

between the rock surfaces at the fault. So all of these generate stress forces. Well, stress is a force. And so basically what happens is, you know, maybe due to the particular shape of the rocks and the way they're moving, they'll be out of slide a little. a bit for a while. And we're talking kind of millimeters and less when we're talking about the amount of movement here. Now, when there's a large earthquake that can move, rocks can move several centimeters, and ultimately that can lead to very substantial changes in the landscape, like mountains coming up, particularly in the middle of the ocean, when you have underground earthquakes. But in terms of the actual sizes of the movements, think millimeters, maybe a few centimeters

isn't like really large changes over time. But anyway, there's an initial period where they can move with respect to each other, but then, you know, it kind of gets caught. It gets caught on something. The rock gets caught on some piece of the other rock. There's friction and rocks are relatively rigid, so they get kind of stuck and forces build up, stresses build up. Eventually, you reach a point where the stress can't build up anymore.

It exceeds some threshold, which obviously depends on the local environment, the type of rock and pressure and so forth. and at that point you have a fault rupture. And the accumulated strain, which is the deformation of the rock, it can deform a certain amount, but rocks are relatively brittle, at least in the crust, and so they can only deform a little bit before they break. You can kind of imagine something like a ceramic plate, and if you apply pressure in it, I don't know, you shut a door on it or something, it will deform a little bit.

You may actually not be able to see how much it can deform because they're quite brittle, right? But if you could measure it with detailed instruments, you could see this. stresses build and you can see the strain, the deformation resulting on the ceramic plate as you shut it in the door. But eventually the stresses reach a point where it can't strain anymore, it can't move anymore without just snapping. And the whole thing breaks and you get what's called a fault rupture in the case of the earthquake, which generates an earthquake. When fault rupture occurs, or just after fault rupture occurs, you'll have what's called slippage. So that's when one plate moves relative to the other as a result of

of the release of those stresses. So that's what I said is on the order of a few millimeters. Tectonic plates can move up to a few centimeters a year, but in terms of slippage after a neutral earthquake, millimeters is more typical. So this is the basic idea. You have a fault between two tectonic plates, or the boundaries between them. Relative motion of the two plates gives rise to a buildup of stresses because there's friction and there's rigidity of the rocks and the bits get sort of caught on each other. Because rocks are relatively rigid, they're relatively brittle, they can't deform very much. Once they've deformed as much as they can, you reach a breaking point where it snaps effectively

or you get fault rupture. That leads to slippage, so one plate moves relative to the other, or they may even both move, but that relieves the strain. Now, that instance of when the fault ruptures, effectively it snaps and you get motion, that releases a huge amount of energy which radiates a way in the form of seismic way. Then there'll be, as I mentioned before, a series of aftershocks where the local different regions in the fault zone gradually sort of settle into place and there's smaller fault ruptures that occur in that sort of iterative process.

Now, that's sort of classical elastic rebound theory. This theory has been under a little bit more, I mean, it's still the standard theory of earthquakes, but it hasn't proven to be completely adequate to describe all aspects of earthquakes. In particular, there's increasing evidence that many seismic faults are not strong. that is the forces that build up, the stresses, are not as strong as has been thought. It seems that there is more gradual deformation, more gradual motion, so it's not just build up and then snap. There is more deformation, which happens over a slow period of time, and the stresses that have built up are smaller before snapping.

So these are called weak faults. So there's increasing evidence for these weak faults that don't properly fit into the elastic rebound theory. And as far as I can tell, there isn't a really good understanding of how, this happens or why it is that way. It seems to be related to the fact that the mineral composition of rocks in many actual faults is heterogeneous. There's many different types of minerals in complicated arrangements and typically what's been studied are homogeneous minerals. Obviously because we can't observe the actual fault itself because they're typically under many kilometers of rock, we rely on indirect measurements through seismography, which I'll talk about later, as well as laboratory experiments to understand how faults work.

And there's increasing evidence from these that many faults are actually weaker with smaller stresses and slow, more incremental slippage, and this may be related to heterogeneous composition of these minerals, as well as the effects of water and other sort of complicated factors there. So that's not really a very full answer there, but it appears that it's more complicated as often it is than the sort of simple elastic rebound theory would state. Nevertheless, I think that that's a useful starting point to think about how earthquakes work and how they occur,

with a gradual buildup of stress over time, leading to strain to the point where there's fracture, and then that releases the seismic energy in the form of seismic waves, which is essentially the earthquake. Now, it's estimated that there are about half a million earthquakes every year. Of course, it depends on the threshold you give for that, as I mentioned at the outset. Only about 100,000 of those can actually be felt that is they're perceptible. The rest can be detected by best current instruments. Of course, as their instruments get better, then we'll detect more earthquakes, but most of them are so small, we can't feel them.

Minor earthquakes occur almost continually. Larger earthquakes are rare, and there's a logarithmic relationship. So essentially, as the energy of an earthquake increases by a factor of 10, then it becomes, then you have roughly a tenth of many of those types of earthquakes every year. It's not exactly that, but that gives you the idea, right? So it's logarithmically less likely that you have progressively larger earthquakes. I'll talk a little bit more later about how we measure and describe the size and magnitude of earthquakes. For the moment, though, what I want to talk about are the sort of mechanisms of earthquakes.

And there, with respect to that, it's important to know that most earthquakes, including most of the large ones, occur in essentially one extended region of the earth. And this is called the Pacific Rim of Fire. mentioned this in probably the previous episode and definitely 1-11 when I did talk about plate tectonics. This is sort of a U-shaped region or line around the Pacific Ocean, which you can imagine it's starting, it starts a kind of western coast of South America and extends up through the Andes along Central America, the west coast of the United States and Mexico, around through Alaska, the Aleutians, and down, then around the other side of the Pacific through Japan and the Philippines, as well as sort of Taiwan, it kind of splits a little bit, and then down through

Indonesia and across then east a bit down through New Zealand. So it encompasses almost the entire Pacific Ocean apart from the southern parts. And this region is the location of so many earthquakes because it is the site where the Pacific plate, which is moving a fair bit, comes into contact with the surrounding plates in the Americas and Australia and so forth. And so that's where most earthquakes occur. Other major sites of earthquakes are also at the intersection between tectonic plates, so particularly the boundary between the African plate and the Arabian plate, and the African plate and the Eurasian plate, which is sort of along the Mediterranean, and as well the boundary between the Indian plate and the Eurasian plate, which is along the Himalayas.

So those are also larger sites of earthquakes. So that's why you have earthquakes, for example, in Greece, which is a region that gets a fair few earthquakes. You tend not to see very many earthquakes around the center of tectonic plates. So, for example, that's why Africa doesn't get too many earthquakes. Now, there are some, particularly around the area of the East Africa Rift, where the continent's actually the Rift Valley, which is kind of coming apart. So there's some tectonic activity there. But relatively speaking, there's not so many in, say, Africa or in Central Asia north of like the Himalayas. There are some earthquakes that can occur even just, like, way out in the middle of a plate. We'll talk about those later. But they tend to occur at the

boundaries between plates. There are also a number of earthquakes that occur along the middle of the ocean, along the middle of the oceans at the bottom of the oceans because of the divergent plate boundaries that exist along the mid-oceanic ridges. And I talked about them again in the Tetonic Plates episode. So that's where you have divergence of the oceanic crust on either side and therefore earthquakes tend to occur there. But the most intense ones tend to occur at convergent plate boundaries, which occur mostly either kind of along online stretching along the Mediterranean and through part of the Middle East, along the Himalayas, and the Pacific Rimafire. Those are the two kind of big areas where you see most of the convergent

plate boundaries and therefore the biggest earthquakes. One earthquake that I want to mention, I'm not going to go through and list a large number of earthquakes. I'll talk a bit more about the 2004 Indian Ocean earthquake later. But one that I want to mention here is the night is the 1755 Lisbon earthquake, which almost completely destroyed Lisbon and surrounding areas, and killed about 40,000 to 50,000 people. That's one of the larger recorded earthquakes. There have been significantly larger ones in terms of loss of life. But one reason that I want to mention this is this event was widely discussed at the time, particularly as part of sort of the Enlightenment and inspired developments in both science and philosophy. Particularly, it was

the, it was the first earthquake to really be studied scientifically in the modern period, at least the first large one, and it kind of led to the development of modern seismology, earthquake engineering and other relevant fields. So it's kind of worth knowing about the 1755 Lisbon earthquake. All right, so that's a little bit about earthquakes and kind of where they occur. Let me now talk a bit more about the different types of earthquakes. So so far, what I've been talking about are tectonic earthquakes that occur at interplate boundaries. And those are the majority of earthquakes. These are the earthquakes that occur when you have two different tectonic plates moving relative to each other and sort of scraping along the edges or pushing

into each other and as I said give rise to the majority of earthquakes. So a few more facts about those before we then talk about some other types of earthquakes. I mentioned faults which is again a break in the rock between two regions of rock in this case particularly between two tectonic plates though again faults don't have to be separating two tectonic plates but that's particularly where a lot of the earthquakes occur. There are three main types of faults. Now, I probably mentioned these before, and I'll probably mention them again

when we talk a bit more about landform and effects of that in terms of rock motion and things. But here I'll just go over this again. So the three main types of faults, particularly between interplates, are normal faults, reverse faults, and strike slip faults. So strike slip fault is, at least in my mind,

what I typically think about one I imagine an earthquake fault. And that's when you have a fault where you have two plates or regions of rock that are moving kind of horizontally relative to each other. So like one goes forward and one goes backwards, so to speak. And the motion is entirely horizontal. That's called a slip strike fault and tends to occur when you have a transform plate boundaries. So that's when plates are moving, like one's moving north and the other's moving south. So southern California, that region is mostly like a transformed boundary with this strike slip folding. normal and reverse faults are different. In both of those cases, there is also vertical motion,

often in addition to horizontal motion. And they're also called slip-dip-dip-faults or dip-slip faults, which I think is a rather odd name, but I guess they sort of slips and dips up. Anyway, reverse faults are those where you have a rock being pushed or a region of rock being pushed upwards relative to another stationary rock, whereas a normal fault is when the moving rock kind of moves downwards. So it's kind of the same effect, like one moves often diagonally up relative to the other and the other is moves sort of downwards. But the issue is which rock is motionary, or again, like a collumeration of rock, it's not a single rock, which plate, if we think of in terms of plate boundaries, which plate is moving up and which plate is stationary, or mostly stationary.

And so in the case of a reverse fault, the moving plate or region of rock is the one that's going upwards. Whereas in a normal fault, the moving. plate goes downwards. And so this leads to different landforms. In the case of a reverse fault, you get an overhanging wall because the moving plate or region of rock tends to push upwards and extend over and start of sticking outwards relative to the existing surface of the earth. Whereas in the case of the normal fault, the moving region of rock or plate pushes downwards, and so you tend to get a wall rock that bends inwards instead of extends outwards. That's hard to explain without seeing a diagram. And it doesn't matter too much for our

I just wanted to explain the difference between them. But whether it's normal fault or reverse fault, in both of those cases, you have vertical motion of large areas of rock or entire tectonic plates. So as I said before, strike slip faults tend to occur when you have transformed plate boundaries, so plates moving horizontally relative to the other. Normal faults tend to occur when the crust is being extended, so pulled apart when you have tensional forces. Typically, it are divergent plate boundaries, such as the East Africa and Rebels.

Earthquakes associated with normal faults tend to be smaller, magnitude 7 or less, and tend to be shallower. Whereas at reverse faults, you tend to get those at regions of compression, so that's when different plates are being pushed or coming towards each other. Those are at convergent plate boundaries, and those tend to have high magnitudes, up to magnitude 8. Again, I'll talk more about magnitudes later, but magnitude the higher the the larger the earthquake is. Convergent plate boundaries also tend to have deeper earthquakes. earthquakes, often owing to subduction. Subduction is when you have one plate that goes under another. Often it's a oceanic plate, which is thinner and denser going under a thicker and less dense continental plate. And because it subdux and then kind of dives downwards into the

mantle, it gets quite deep and, but there's still friction between the plates, right? So you get earthquakes occurring at much deeper in the earth. So that tends to occur when plates are crashing into each other, not being pulled apart from each other. So that's why reverse vaults, where it's mostly compressional forces coming towards each other, you tend to have the deeper earthquakes, whereas tensional forces being pulled apart, you tend to have the shallower earthquakes. So those are different types of faults that you have, and these are all tectonic earthquakes, all caused by motion of tectonic plates, but motion in kind of different directions or relative to one another. Now, there are other types of earthquakes, and the first we'll talk about

are deep earthquakes. Now, these are still tectonic earthquakes, but the difference is that most earthquakes occur at quite shallow depths. I kind of mentioned this before. When we say shallow, it's not really shallow by any sort of normal human conception. A kilometer underground is a very, very deep by normal human measures. And in fact, I think the deepest we've ever dug is only about 10 kilometers below ground. So most earthquakes occur between 70 kilometers and the surface. So 70 kilometers is still quite deep in terms of human conceptions, and it's far deeper than we've ever dug or drilled.

However, it's quite shallow in terms of the scope of the earth, whose diameter is, I don't know how many thousand kilometers. 70 kilometers is about as far as you can get. It is about as deep as the continental crust gets at its deepest places. So shallow earthquakes is essentially down to the bottom of the crust. Intermediate depth earthquakes are between 70 kilometers and 300 kilometers down. And then deep focus earthquakes occur at much larger depths, like 3 to 700 kilometers or so. And that typically occurs, as I mentioned, at subduction zones where you have the plates diving quite deep into the mantle.

Now, deep earthquakes, so particularly those 300 kilometers or more, are particularly poorly understood. And the reason for this is, remember, we've got the elastic rebound theory, which is basically you put a stress on the rocks that leads to a strain, but because rocks are brittle, they can't strain too much. And so when the force gets too much, they crack or they snap, you have a fault rupture, and that leads to a slippage or motion of the plates relative to each other, and that releases the seismic waves and releases the stress. That whole story shouldn't really work for deep earthquakes, like below a few hundred kilometers. Because at those depths and pressures and temperatures, rocks should be a lot more plastic. They shouldn't be brittle. They shouldn't snap like that.

They should deform more gracefully. And so you shouldn't have these buildup of stresses. So several mechanisms have been proposed for how you can get earthquakes at these depths. One of them is related to, oh, sorry, I should say, although several mechanisms have been proposed, all of them are still somewhat speculative. And so the exact process is still fairly poorly understood, especially because it's so hard to study these because they so far below the Earth's surface. I'll just talk briefly about a few of the mechanisms. This is obviously not my area of expertise, so I don't want to make any particular claims about which is more plausible, but I thought it was interesting just to give you an idea of how we think these things might

happen. So the first is solid state transition. So this is an older idea, and the idea is that, you know, once you get to certain depths and certain pressures, minerals, certain types of minerals, essentially implode because they undergo a phase transition from a less dense material to a denser material. And so they take up less space. And so there's kind of an implosion which then triggers relative motion of different, along some sort of fault, between different regions of the of the plate or between different plates, and then that triggers an earthquake. Now, it seems that theory is largely being discredited, basically because as far as I can tell, it doesn't really work like that. You don't have this sort of rapid phase transition in a way that could, and at the depths and pressures

and everything that could trigger the earthquakes and the ways it's thought. But modern theories kind of build on this idea. So another theory is called dehydration in brittlement. And the idea here is that reactions of certain minerals in certain phases and pressures and so forth that have higher water content could lead to effectively an increase in the pressure of that water being extruded from the pores between, you know, like the pores are the gaps between the, the gaps between the microstructures or the minerals effectively where water can be stored. And it's thought that reactions of this water that's extruded that's forced out with the high pressures might change the properties of certain types of minerals, causing them to be more brittle than they otherwise would be.

That's what's called dehydration in brittlement. Basically, you remove the water in a way that causes it become more brittle and thereby could lead to the kind of buildup of stress and strain and therefore fault rupture that is more consistent with the elastic rebound theory. The final type of theory that I came across is called transformation faulting. And this is related to a phase transition that we talked about before, to a high-density material. But it's not just because of pressures, it's a result of sheer stress, so a particular type of force that's applied to the material. There's been recent experimental results that led support to this. Again, as I said, many of this has to be conducted in laboratories because we can't really observe

this sort of stuff happening in the wild. So whatever the truth of the matter is, it seems to be the cause of deep earthquakes is probably related to detailed interactions of pressure, temperature, and the mineral structure, possibly also with water being involved there and exactly how the water exists inside or is extruded out of the crystal. Some combination of these factors leads to the rocks being more brittle or behaving in a more brittle manner than they otherwise would, which can give rise to the built-up. stress and therefore earthquakes. The exact cause of deep earthquakes is still fairly poorly understood. Most earthquakes are, as I said, shallower, but there are still earthquakes that can occur at quite large depths. Now let's move on and talk about in some sense even more and even more puzzling phenomena, which is non-boundary tectonic earthquakes. So again, just to keep us on track, remember I said that most earthquakes are tectonic in nature, so motion of tectonic plagues,

occur at fairly shallow deaths, like less than 70 kilometres, and occur at the boundaries between tectonic plates. So these would be shallow tectonic earthquakes at plate boundaries. We've talked about deep earthquakes as a violation of the shallow part, and that's puzzling because you expect deep rocks to be more plastic and not build up as much of a stress. Another kind of exception is when earthquakes occur away from plate boundaries. So they're still tectonic earthquakes because they're still driven by tectonic forces, but not at plate boundaries now.

It's inside a plate. This is called an intra-plate earthquake, which occurs in the interior of a tectonic plate. One prominent region where these occur is the southeastern United States, centered around parts of Tennessee. And this gave rise to, or this area experienced a very large series of earthquakes in 1880, sorry, in 1811 to 1812 called the New Madrid earthquakes. and has experienced smaller earthquakes since then. Now this is not at the boundary of any tectonic plate, so this would count as intra-plate earthquakes. As I said, the cause of such earthquakes is largely unknown

because you can't explain it in terms of tectonic plates moving relative to each other because it's all one plate. As I said, though, there are still faults that exist within the interior of tectonic plates. The issue is that they shouldn't be especially active because there's no major forces that are operating to move one part of it relative to the other. Like you can still have gaps, right? But one side on one side of the gap and the rock on the other side of the gap shouldn't be moving relative to each other either sideways or up and down. So that's, you wouldn't normally expect earthquakes there. And typically you don't get earthquakes, right? The intraplate earthquakes are relatively rare compared to interplate ones, but they do happen. One hypothesis is that ancient faults that are still there may be activated either by some sort of interaction with water or perhaps induced by interplate activities like transmitted across the plate. One factor here. here that may be relevant is that energy tends to propagate much more effectively through consolidated compared to unconsolidated media, meaning that if you have lots of sort of contiguous rock,

then seismic waves will propagate more effectively through that than, say, at the boundaries between plates, or if you have lots of sediments. So certain types of rock may help to propagate energy from disturbances that occur at interplate boundaries and then perhaps have secondary effects at fault internally, although that's speculative. We don't really know what the cause of these intra-plate earthquakes is. Again, there are faults that exist within continental plates. It's not clear why they're active in this way, but they can be, and that can lead to earthquakes even in quite unexpected places. Now, the final type of earthquake that we're going to talk about are non-tectonic earthquakes. So these are not even caused by tectonic forces. As I said earlier,

you may not even define these as earthquakes, depending on whether you want to define earthquake as motion of the ground caused by a seismic event or just any kind of large-scale motion of the ground. non-technic earthquakes can be caused by any major release of energy that occurs either below the earth surface or around it or even above it. So volcanoes can cause earthquakes. There's a called volcanic earthquakes. So essentially as magma is moving and gases are being released and it's being extruded to the surface, that generates vibrations in the rocks which are seismic waves, which then propagate out. So the motion of magma itself can cause earthquakes. as well as whenever there's a volcanic eruption, fairly obviously that can lead to vibrations of rocks

which can cause earthquakes. There are also collapse earthquakes, which is caused by the collapse of some sort of, often sediment somewhere. So this could include collapses of mines, landslides or underwater landslides, anything of that sort can cause an earthquake. Finally, explosion earthquakes.

Now these are caused by, well, explosions, as you might have guessed, such as asteroid impacts, which probably have accounted for the life, are just earthquakes the planet's ever seen, although we haven't had any of those recently, the very large ones. Nuclear explosions can also cause earthquakes. Nuclear explosions can certainly be detected using seismographic equipment, which we'll talk about later, but some of them actually caused, you know, like moderately sized earthquakes,

not just very small ones, but they can actually cause like, you know, kind of real earthquakes, if you like, because the explosion is so large. So really anything that releases enough energy to vibrate the Earth's surface and rock underneath can cause a non-technic earthquake. All right, so we talked about some of the major types of earthquakes and some of the mysteries of deep and non-boundary earthquakes. Let's now talk a bit more about seismic waves, which I've mentioned but haven't gone into a lot of detail into. So seismic waves are waves of energy that are released by earthquakes. In some sense, the seismic wave is the earthquake, I suppose, if you define an earthquake as motion the earth's surface.

Well, that is seismic wave, or at least a type of seismic wave. So there are four main types of seismic waves in two categories. nice and neat two by two. There are body waves which travel through the earth's interior and surface waves which travel on the surface of the earth. So let's talk about body waves first. First of all, we never directly observe body waves because we don't see under the earth's surface,

but we can certainly measure their effects. The two main types of body waves are called P waves and S waves. The way to remember this is that it's named after the order that they're detected in. So P waves are primary, they come first. S waves are secondary, they come second. So, as kind of makes sense, P waves are fastest. That's why they are detected first. S waves are slower.

Now, P waves, or the fast ones, they can travel through solids and liquids. Effectively, you can think of them as kind of longitudinal waves. They consist of compressions and rarifactions. So they're kind of like sound waves in air, except they occur in rocks. But they can also occur in liquids. Now, this is important because it means that P waves can travel through the outer core of the earth, which is liquid. S-waves, by contrast, cannot travel through the outer core of the earth, and therefore cannot travel through the core at all, because although there's a solid inner

core, they can't get to it, right? Because the outer core surrounding it, they can't travel into. So, P-waves from an earthquake will be detected, as long as the earthquake's large enough, effectively over the entire surface of the planet, if you, you know, wait for them to reach that area. There is a region called the P-wave shadow zone, and this originates from, well, is that actually two of them, kind of one on each side of the earth. And this originates from the fact that when waves move from one medium to another, in this case we're talking about seismic waves, but this applies to any wave. When they move from one region to another, they change in speed,

depending on the properties of the medium. And this gives rise to a phenomenon called refraction, which is a change in the direction as a wave of a wave as it changes medium due to a change in speed in that medium. And so basically what happens is P waves refract, they pass from the solid mantle into the liquid outer core. They bend. And they bend again when they get back on the other side of the outer core and travel across the other side of the Earth's out to the Earth's surface on the other side of the planet.

But because of the bending, there's a region where no waves are detected. Again, it's easiest to see on a diagram, but you can kind of imagine they bend in a way that there's a region where they're not detected. And this is called the P wave shadow zone. There's one kind of on each side of the planet. And then there's a region in between them where you do see the P waves again. S waves on the other hand are not detected over the entire planet. They're only detected in regions where they can kind of propagate directly to, so across the mantle.

Remember, the mantle surrounds the liquid outer core, which then surrounds the solid inner core. Any trajectory of S waves that would require going through the outer core is cut off because they can't travel through water. So S waves are only detected in kind of nearby regions of the Earth, and that could still be a very large part of the Earth, but not the opposite side of the Earth, whereas P waves are detected everywhere except in the shadow zones. That's because of P waves can travel through solid and liquid. S waves cannot travel through liquid. Now, you might ask, why can't S waves travel through liquid? Well, the reason is because they are transverse waves. So they involve up and down motion, and therefore can only travel in solids.

Basically because if you have a liquid near the center of the earth, well, the outer core, you're not going to get transverse motion there. There's too much pressure. You can still have longitudinal motion. Compressions are rare fractions, but you're not going to have kind of up and down motion there. but you can have that in the Earth's mantle. That finishes body waves. Again, you've got P waves, which are longitudinal,

travel through everything, S waves, transverse, don't travel through liquid, and are also slower. Now we've got surface waves. That's the second category. These are classified into two different types of waves, which in this case are named after the scientists who describe them. So love waves and relay waves. Love waves are basically the same as S waves,

except they're trapped on the surface of the Earth. They're probably the most damaging type of waves. Love waves are, they're on surface of the earth, obviously, but they're effectively side to side motion across the ground. And that's why they can be particularly damaging, because they'll basically shake a structure apart as it sort of passes across the structure. One part will move one way, the other part move the other way,

and it will tend to sort of tear it apart. So that's why they're so damaging. Relay waves are kind of like up and down waves, and they're more like your typical beach waves, you know, that come up and go down and kind of bob up and down, whereas love waves are only side to side. Relay waves have a more circular rolling motion, and they also can be quite damaging,

but perhaps a little less damaging than love waves, because if the building is well built, it can survive being sort of pushed up and down a little bit, but being sort of ripped kind of from one pool one way, another part pull the other way, is a bit trickier to deal with. Now, both love and relay waves, that is surface waves, travel slower than body waves.

So you can measure, kind of in order, the type of waves that you're going to detect when you're detecting seismic waves. So the first will be the P waves, then the S waves, and then the surface waves. Let's now talk about, in transitioning to that, how we measure earthquakes, and particularly seismography, which is the science of measuring earthquakes and seismic waves generally. A seismometer is a sensor that detects the motion of the earth arising from seismic waves. A seismometer is essentially a weight attached to a spring or something that's equivalent to that. The spring is anchored to a frame which is attached, you know, basically to the earth,

to the ground. So when the earth shakes, the frame will shake and that will move the spring, but the inertia of the weight will keep it in place. And therefore it will, the weight will move relative to the frame. Or I suppose, more technically, the frame is moving relative to the weight, because actually the frame is the one being accelerated. But anyway, and that relative motion can then be used to measure the magnitude of an earthquake. You can put seismometers pretty much wherever at the earth's surface, inside boreholes, underwater, wherever. Usually a seismometer will be combined with some sort of measurement that can record that motion rather than simply just producing the motion. And that package is called a seismograph. The record that's made by a seismograph is called a

seismogram. So the seismograph produces a seismogram. And, you know, everyone has seen these. This is the scribbles on the piece of paper. Traditionally, there was a rotating drum. which was placed next to the weight and some sort of pen or pencil attached to the weight, so that when the weight moved relative to the frame, the drum being attached to the frame, it also moved relative to the drum, and then recorded sequence of motions on the drum as it rotated. And then you get a piece of paper, which recorded sequentially over time, the motion of the earth. So when there's no motion, you get a straight line, and then when the P wave start to arrive, then you get a bigger squiggle, and then an S wave comes, and you get it even bigger,

squiggle and then the surface waves are the biggest squiggles of all as the earth vibrates more intensely. Now these days as far as I can tell those sort of analog straight to paper, seismographs are really only used for tourists or public display. These days, you know, it's all recorded digitally of course to a computer. So how can you use seismography to determine the location of an earthquake? And when we say the location, here we're focusing on the epicenter, so the location on the Earth's surface that's closest to the focus. So to do this, you need three seismometers at different locations on the Earth's surface, ideally sort of fairly wildly separated to be a bit more accurate. One seismometer can tell you the

distance of an earthquake, but not its location. So you need three to tell you by triangulation, the location. But how does that work? So you got a thosom... I was going to say a thermometer. You've got a seismometer, right? The way that you can measure the distance of an earthquake is to measure the time taken between when you first detect a P wave and when you detect the S-wave. And when you detect the S waves. Remember, the P waves are faster, so they'll always detect it first. However, the time between the P and the S wave allows you to determine, given that we know the velocity that P waves travel and the velocity that S waves travel, the time between them, basically you just multiply out or divide out by the velocity to get distance that's been traveled by the waves since they were produced.

So that's all you have to do. You measure the time difference between the onset of P waves and the onset of S-waves, use the known speed of those waves, and then you've got the distance to the epicenter. Do that in three different locations, find the point of where those three circles overlap, and that gives you the epicenter, or at least that's an estimate of the epicenter. There will be subtle local variations, of course, due to seismic, due to geological conditions, but that's a pretty good illustration of the basic idea. If you want to determine the location of the focus, then that's going to be a location interior to the earth, usually fairly fair. fairly close to the surface in terms of the overall scale of things, but still not on the surface

of the earth. So you need four seismometers in that case. You use spheres. So there'll be some sort of point where all four spheres intersect, and that will be the focus of the earthquake. And that's effectively how it's done. Now, in addition to measuring the location of an earthquake, we also want to measure the size of the earthquake. And by size, we mean effectively the magnitude. So that relates to the amount of energy that's released. Additionally, though, there have been different ways of doing this. And you do hear different sort of measurements given in the press particularly, so you need to be careful about this.

Now the oldest scale is called the Murkali scale, and it's actually not a magnitude scale, it's an intensity scale. The intensity of an earthquake is different to its magnitude or size. And the difference is that the intensity of an earthquake is basically how much damage it does, the physical observable effects of the earthquake on people and the environment. obviously depends on where the earthquake occurs, right? When an earthquake occurs in Antarctica or in the middle of the ocean, as long as there's no tsunami, then the effects on humans are going to be relatively small. Whereas if it occurs in a populated area, well, that's going to be a lot of difference,

right, as well as the local geological conditions and many other factors, right? So the McCalli scale is useful for sort of measuring the human impact of an earthquake, but not very useful for sort of geological or sort of seismic studies because it doesn't tell you about the sort of physics behind it. So the McKalee-scale is measured empirically. Everything from the lowest level, which is not felt at all, up to barely felt right through to total destruction objects thrown into the air, river courses, and topology altered. So it's typically, McCulley scale, typically given on a one to 12 scale, although I think it's been modified a few times. Anyway, I'm not going to go through all of the levels on the scale there just to give you an illustration of how it works. And as I said, that is to

termed empirically and sort of qualitatively. More useful from sort of a scientific purposes are magnitude scales which actually measure the amount of energy released by the earthquake. However, there are different ways of doing this, of course. Now, everyone's heard of the Richter scale or the Richter magnitude scale. This was developed in the 1930s to describe earthquakes in Southern California and has been modified to apply to other locations as well. But because of this, it really is only a local description.

That is, the way that it works is that it's a logarithmic scale that defines the magnitude as related to the logarithm of the amplitude of the seismic waves that are measured to some baseline. So the ratio of the actual measured seismic waves to some arbitrary baseline, right? So basically, what it does is the Richter scale converts measured ground motion to energy using a conversion scale. And perhaps sort of intuitively, you can't do that just sort of without further information, right? you have to have some conversion factors to go from how much the ground moved to how much energy was there. And that's going to depend on the composition of the rock and the local geological outline and a whole bunch of factors, right? That's why the original Richter scale was only really applicable to Southern California. And there have been different versions that have been developed, like there's one version that was used in Japan, for example.

But it has a lot of shortcomings, lack of generality being one of them. And it's not really used by modern seismologists. but for some strange reason that I've not been able to figure out, the press continually refers to or reports earthquake magnitudes as so-and-so on the Richter scale, even though usually what they're reporting are moment magnitude scale. So that is the name for the preferred measurement of earthquakes by most scientists today. So the moment magnitude also measures the energy released by an earthquake, but it does so in a more accurate way that's based on physical first principles.

It's based on the rigidity of the earth multiplied by an amount of slip at the fault and the size of area that slipped. So it's more first principles focused less on actual movement of the ground, which is not as directly relevant to the amount of energy released by the earthquake. Obviously, it relates to it, right, but that's not as good. So effectively, what you can see is a progression of moving from more superficial characteristics to more fundamental characteristics in terms of characterizing an earthquake. So McCalleliskale is purely superficial empirical human effects. Richter scale is more related to the fundamentals in that it's related to the amount of Earth movement, like it's measured on a seismogram, but that's still fairly kind of indirect. Moment magnitude takes it to a much more fundamental level where it's about actually the,

you know, rigidity of the earth and the amount of slip at the fault itself and so forth. But as I said, the press for some reason still reports estimates of the magnitude of earthquakes as being on the Richter scale, even though as far as I can tell, the scientists that they get those numbers from are reporting it in moment, magnitude. So what you'll have is scientists who will estimate a moment magnitude, usually estimate, because when the earthquake initially occurs, they don't know, they just estimate it. Then they tell that to the press and the press report it as, you know, a six or a seven or whatever on the Richter scale. You know, a five is a strong earthquake, but not crazy. A six is a pretty bad earthquake. Seven is

like a really, really dangerous earthquake that causes buildings to fall over and so forth. Eight is catastrophic, like everything is destroyed. So that gives you a kind of an idea. and four and below is barely detected at all. The very strongest earthquakes that have ever been detected are around a nine. Now, another point to make about detection of earthquakes is that measurement of seismic waves allows us to make inferences and study the Earth's interior. This kind of relates to what I was saying before, because P waves can travel through the outer core, whereas S waves can't.

And so the exact location where waves are measured, where the shadows are, how long they take to reach different locations and so forth, can tell us about the different layers. layers of the Earth. That's one of the main ways we know about, say, the liquid out of core, for example. Now, I want to talk more about this at some point in the future when I do an episode on geophysics, but I just thought I'd mention that briefly here, that there's an interesting connection between earthquakes and studying the Earth's interior. All right, let us now move on and talk about earthquake damage, so the different ways that earthquakes cause damage. The most obvious method is, of course,

shaking of the ground, which causes objects and buildings to fall over. Most people in earthquakes, as I understand it historically have been killed well either is a result of tsunamis but if you put that in a separate category tsunamis are often caused by earthquakes but aren't the same thing excluding those most people historically will be killed either by fires that are caused by earthquakes which I'll talk about in a moment

or they're caused by buildings falling on top of them generally poorly built structures that collapse which is why if you can it's generally good to be kind of out in the middle of nowhere when an earthquake happens you might fall over but at least nothing will fall on top of you one of the kind of characteristic things that often happens in larger earthquakes is overpass is falling over. There's so many pictures that you'll see pretty much in any book if you pick up about this, particularly in Japan, which obviously gets a lot of earthquakes, and because there's not a lot of space in Japan,

they typically build overpass, you know, overpass, you know, highways on top of other things or train tracks and so forth. And it seems they're always falling over. I mean, as I understand, they have quite strict regulations in Japan, and their buildings are actually quite well built, but there's still lots of images to see if overpasses falling over. Another factor that can be relevant in terms of the damage caused by shaking is that aftershocks cause further damage and even topple buildings that were initially damaged. So you'll have the initial structural damage caused by the main shock and then aftershocks can kind of do the coup de grace straw that broke the camel's back if you like. However, shaking of the ground is not the only method of damage or cause of damage insicated by earthquakes. You've also got something called ground rupture, which I kind of find the most interesting, at least visually interesting of the forms of damage caused by earthquakes.

And this is visible breaking and displacement of the earth surface that you can like see at the earth surface. It will look at along the fault trace. Remember, that's the physical manifestation of the fault at the earth surface. Faults are often not extend up to the earth surface, but when they do, you have a fault trace. Ground rupture is visible manifestation of the fault trace. So you may not be able to see the fault trace if it's just like, if it's already a craggy, rocky area, it may not be as obvious. But some of the cases where it is most obvious is when you have like a fence. The fence extends along both sides of a tectonic plate, or at least across a fault, right?

And then there is a horizontal motion, slip strike fault, such that the fence which was once in a line is now, there's a gap of one or two meters between the two different types of the fence. Ground rupture is a major risk for large engineering projects such as dams, bridges, nuclear power stations, as these all require very careful sighting and planning about where to put them. Something you may have heard of is a phenomenon called liquefaction. Soil liquefaction occurs when a saturated or partly saturated soil, so that is saturated with water, loses strength and stiffness and starts to act like a liquid. This seems to happen because basically when soil is largely saturated, when you have loose soil, like sandy soil that's mostly saturated with water and then compressed,

and particularly if that compression is sort of repeated, the water pressure builds up to an extent where it kind of pushes the grains of sand, sorry, the grains of soil apart, and so they're no longer in full contact with each other. So normally the grains of sand or soil would kind of settle, even if there's a lot of moisture in the soil, they'd be in contact with each other, which helps to give it rigidity. But if there's enough force that kind of is continually pummeling it, so to speak, or it's under enough pressure and there's enough liquid there. Basically, the liquid can be forced out, and therefore the grains of soil don't come into direct contact with each other anymore, and therefore they behave like a liquid. this can even happen when the soil appears very solid beforehand. It's not like it looked like a puddle of mud before and then it starts and then it starts

localifying during the earthquake. No, it looks solid like it looks fine. That can be because the soil was saturated with water but that's below, you know, the water table is below the surface of the earth, so you don't see that. That becomes manifest during the earthquake when the water is sort of pushed up due to pressures it pushes up towards the earth surface as well as the foundations of buildings go below the surface of the earth, right? So the point is that this can happen even when it seems that, you know, it's solid as a rock on the, well, maybe not a rock, but, you know, solid as good compact soil on the earth's surface. But if there's an earthquake, the soil can lose all structure and behave like a liquid. And so you can have these strange phenomena where buildings that are built to withstand an earthquake. So they're solid. Like the building has stayed intact. But it falls over, like it sinks over on one side. Again, look up these images. They're very, very bizarre. It's, it looks like, you know, a child's knocked there, their building blocks over or something.

not like full horizontal, but they kind of like have sunk into the soil and are at a crazy angle. It's very strange. I mentioned fires. Fire is very common with earthquakes. In the past, this would be one of the major causes of loss of life because lanterns and candles would be knocked over and because most things are built of wood, everything would burn and people would die. Another factor is that water mains are also typically ruptured, thereby making it difficult for firefighters to get the pressure to fight fires. Also tends to block roads, make it hard to get to the fires. Mast mains can also be broken, causing more fires. And you may have heard of the 1906 San Francisco earthquake, which caused great loss of life. That was, most of those deaths were caused by fire

compared to the actual earthquake itself. We then come to the final and in some sense largest cause of death as a result of earthquakes, which are tsunamis. These are also called tidal waves. I think that maybe that usage is decreasing. I don't know. I feel like I hear tsunami more than I used to, but maybe that's just me. At any rate, this is a bad name because they don't have anything to do with the tides. Tsunamis are not caused by tidal forces. They're caused mostly by earthquakes out at sea, or underwater earthquakes. It's still not fully understood why some earthquakes cause tsunamis and others don't. It does appear to require vertical motion of the plates relative to each other, so slip strike motion like a transform fault won't typically cause tsunamis.

What you need for a tsunami is up and down plate motion, which then physically, like vertically displaces large volumes of water, which then produces very, very large, long wavelength waves, travel across the surface of the ocean and when they reach shoreline typically decrease in wavelengths so they get like higher frequency but also increase in amplitude and the wave height of like tsunamis can vary a lot you don't typically get big cresting waves like sometimes you see in the movies of these crazy like monster cresting waves that doesn't usually happen but you can typically it just looks like an in rushing torrent of water i think a lot of people have seen the striking videos from the 2004 tsunami in the indianian

an ocean to get an idea of what that looks like. It just looks like a big torrent of water that's been let loose. And these can be anywhere from like a few meters to 10, 20 meters high. Typically what happens is, you know, because it's a wave, so it's a series of peaks and troughs, either of those can reach the land first. So sometimes the peak will reach first and you'll have an inswell of water. Sometimes the trough reaches first. So that means the sea will recede in the drawback phase for a few minutes uncovering large areas of a region that's typically underwater. Unfortunately, people will often go out and explore. In some cases, many shipwrecks have been revealed, and people will go and explore and try

to pick up fish and other stuff. However, within a few minutes after that, then the ridge is going to come in, the peak, and then that covers not only the areas that were just revealed, but also large areas of the land. That process of wave comes in, wave goes out can occur multiple times. So what you'll have is that, you know, people will be taken up or affected or washed off by the flight as it comes in, and then the victims and the debris will be swept out back into the ocean again. So some people may survive on top of the debris, for example.

You can see examples of that in the footage from 2004, but they're at risk of being sucked out back to sea again when the wave recedes. I mentioned the 2004 Indian Ocean tsunami a few times. So this is one of the deadliest natural disasters in human history. It killed about 230,000 people. There are a few earthquakes in Chinese history, a few in the early 20th century and a few in medieval times, which probably killed more people, say probably because we don't have accurate accounts, but probably not a lot more. Like maybe they killed 250, 300,000.

There's one that has, in the Middle Ages, highly uncertain numbers. The earthquake was also the third largest ever recorded in terms of energy, and the largest in the 21st century. It also had the longest duration of faulting, so like actual motion of plates, which is between 8 and 10 minutes, which is pretty crazy. Before we finish out, I just wanted to say a couple of words on earthquake prediction. Despite efforts of scientists and, as well as quacks, over many decades, earthquake prediction has never really happened.

There have been some cases of claims being made to predict earthquakes, but often there's no official prediction. It's kind of like people saying, or maybe there's going to be an earthquake and then there was. But of course, then you've got to think about all the times when people didn't predict an earthquake and then there was, or all the times where they predicted an earthquake and then there wasn't one. So you've got to adjust for that, right? you can't just take one case of a successful prediction and say, oh, they predicted it. You've got to take a series of cases and evaluate it statistically.

Because especially if you're talking about high earthquake-prone regions, well, eventually there's going to be an earthquake, right? So there just really aren't any examples of successful earthquake predictions, at least not of any large earthquakes. When we're talking about earthquake prediction, we have to distinguish this from earthquake forecasting, which might sound like they're the same thing, but they're used to refer to different things. Earthquake forecasting is a branch of seismology, which describes probabilistic assessment of general earthquake trends and levels of hazard. This can be done accurately. You can say the probabilities and amounts and frequencies and magnitudes of earthquakes in different regions. Obviously, you can say there's more earthquakes in California and Japan than there are in Mali, for example, which is in the middle of the Saren Desert. So yes, we can make those assessments. But what you can't do,

or what we can't do, is predict particularly like there's going to be an earthquake on this date. And the thing is that you need to have a fairly narrow prediction in order of an earthquake for that to be useful. It doesn't help if you say there's going to be earthquake this month? Well, what are you going to do about that? You can't evacuate for the whole month. It basically needs to be one day or maybe like a weekend or something. And even then, it's not entirely clear what value the prediction is, even if you can get it to the day, because large cities can't really be just evacuated wholesale, and that would cause a lot of chaos, and people would die in traffic jams and crime and, you know, hardships and, you know, people being taken away from

medical care and all sorts of other things. So you have to balance that against the benefit of trying to evacuate people or something. So it's not clear that earthquake prediction is possible or even really desirable in most cases. Obviously there'd be cases where it would be helpful. For example, if we knew that the 2000-4 tsunami was going to happen, people could have moved away from the coastline. But overall, it's probably not worth all the effort that's been put into it. In some spheres, huge amounts of money have been spent trying to develop sophisticated statistical models and looking at relationships between aftershocks or time periods between earthquakes or trying to know, take measurements based on atmospheric or other seismic measures. All sorts of complicated methods

have been developed, which I won't get into here, and just none of them have been very successful. Again, not when evaluated rigorously, you know, taking into account all the factors that I mentioned before. The naive idea that the next big earthquake will occur in some area that hasn't had an earthquake for a while is sort of true. It can be useful, as I said, in terms of broad overall trends, but that's not helpful in terms of actual prediction. You know, you'll say, well, this area is probably due for a big earthquake in the next 10 years or something, but that's not helpful for actually predicting when specifically it will happen. And so those methods don't really work well. Another thing that you may have heard of is the idea of animal behavior as being useful for earthquake prediction.

So there are many reports of animals acting strangely before earthquakes, such as running around, barking madly, or behaving in other strange ways. It is actually quite plausible that they are detecting something real, at least some of the time. Some of this would just be like confirmation bias, right? But some of this is plausibly real. For example, they may be detecting four shocks that humans can't detect because animals can detect different subs of vibrations than we can. They may also be detecting changes to groundwater or other subtle phenomena that may be indicative of quakes. So it's possible that animal behavior is connected to earthquakes.

However, these four shocks and other things that animals may be detecting are already detected to much greater precision and with much greater objectivity by seismometers. So it's not like the animals are giving us any additional information, or at least there's no evidence or reason to think that animals are providing information that we don't already have. So they may be detecting something real, but that doesn't mean it's useful because we already know that. The issue is we don't know how to incorporate all of these potential factors, like foreshocks or maybe changes to the groundwater or there's possible and even atmospheric factors which may or may not be related to earthquakes. It's not clear how all these factors can be used to actually predict earthquakes. and so animals just aren't telling us anything that adds to that. Now, probably the most useful thing that can be done about earthquakes is to enact policies in areas that are prone to earthquakes,

proportional to the risk and proportional to the cost, to reduce the damage. And that's by far the most successful way of dealing with earthquakes. I think Japan has done fairly well at this, although perhaps 2011 notwithstanding. So that's probably the most useful policy aspect of earthquake forecasting, not prediction, but forecasting, to inform public policy about building codes, insurance, general public awareness and preparedness, and public policy relating to response about what happens when there is

an earthquake. And that can absolutely save a lot of lives if buildings are constructed in the right way, if there's appropriate plans in place and general public understanding of what to do. All of these things can be very helpful, much more than trying to predict the specific time when an earthquake will occur, especially because earthquakes generally only last a few minutes and trying to predict when those few minutes will be in the span of decades, it's just probably not possible. Some scientists think that it's in principle impossible to predict earthquakes like that. Others think that maybe it's possible, but it's certainly too difficult to do at the moment. So that brings us to an end of the episode, hopefully not on a too negative note.

We've covered the causes of earthquakes, different types of earthquakes, seismic waves, how earthquakes are measured, as well as some facts about earthquake damage, and a little bit on prediction of earthquakes. So if you found this episode interesting. If so, you might consider making a, positive review of the podcast on iTunes or Spotify or whatever aggregator you use. You can also get in touch with me at FODs12.g email.com. That's FODDS1.2 at gmail.com if you'd like to make any recommendations for future episodes

or just say hi and leave a comment about the show or ask any questions. I'm always happy to hear from listeners. And thanks to all those, particularly recently. I think I've got a bit of an uptick in email correspondence. It was always nice. So thanks for that. I respond to everyone. If you would like to support the show financially, you can make a one-off donation via PayPal. You can just use my email for that, or you can become a Patreon supporter at my Patreon, which you can Google that up. I appreciate all of my donors who help with the, you know, small, roughly monthly donations to keep the show going and help me devote a bit more time to it.

That's much appreciated. In terms of what's coming up, so there's a few episodes in the works. One I want to do on sleep and another one that I'm thinking of doing on depression or depressive disorder, which might be a bit of a downer, but it's an important topic. Also, a bit of a longer-term project. I am wanting to do an episode, a couple of episodes probably on muscle, the muscular system. We just did the respiratory and circulatory system, so I'd now like to do one on the muscular system. That will probably be at least two episodes. And then longer term after that, I want to do

a series on nutrition and digestion. I'd like to combine a few elements that are often discussed separately. So I want to combine kind of the anatomy and physiology of the digestive system itself, alongside with nutritional advice and information and dietary studies and things like that. there's a lot of misinformation there, plus the bichemistry, the actual metabolism of, you know, different types of nutrients and so forth, because they're all closely related, but often discussed separately. So I'd like to do probably a three to four episode series on that. That's a longer term project. That will take some number of months. But anyway, there's a few things to get you excited. So once again, thanks very much for listening. Take care of everyone, and I'll talk to you next time.