Plate tectonics and earthquake prediction with Dr Rebecca Bell

This recording is from the Department of Education and Outdoor Learning at the Royal Geographical Society with IBG. I'm Laura and in each podcast I'll be meeting a geographical expert to discuss their research of how geography has taken them. Have you ever wondered what it's like to research plate tectonics and seismic activity? In this podcast we speak to Dr Rebecca Bell from the Department of Earth Science and Engineering at Imperial College London. We talk about her research into subduction zones and earthquake predictions. You'll find out what earthquakes and London buses have in common. So UIT broadly is on tectonic evolution. Can you tell me a bit more about what this means and how you developed an interest in this topic? Of course, so since the 1960s we've known that the Earth's outer layer has been divided up into a series of tectonic plates. So these plates are all moving around and interacting with each other in different ways. In some places they collide together, these are called convergent boundaries. In some places they're moving apart and when that happens magma can rise to fill the gap and that magma then solidifies to become new oceanic crust. So that's what's happening currently between us and the US when moving apart and the Atlantic oceans forming. And in other places the tectonic plates are sliding past each other. And very famously that's happening at the San Andreas Fault along the western coast of the USA. My research particularly focuses on places where two tectonic plates are colliding. And that's interesting because it's what's happening, that's what's causing the Himalayas to rise. And when you have an oceanic plate colliding, either with a continental plate or another oceanic plate, the denser, the heavier oceanic plate subducts beneath it and this produces something called a subduction zone. And the divide between those two plates, it's basically a giant fracture in the earth that we call a fault line. They can be hundreds if not thousands of kilometres long. These are the biggest faults on earth and this is where we get the biggest earthquakes on earth. The 2011 Japan Magnitude 9 earthquake, this happened at one of these subduction zones. The 2004 Boxing Day earthquake in Tsunami, you might remember, killed over about 250,000 people in the Indian Ocean area, that was also one of these subduction zones. So it's not really quite a terrifying but fascinating area to study. And I guess I got into it at school, I liked math, physics and chemistry, but I didn't want to take any of them on their own as a degree subject because I wanted to keep doing all of them together. So one day I was looking through an university prospectus and came across this subject called earth sciences or geology. So it seemed to kind of combine all of my interests together, but also allow me to learn about the earth, learn about earthquakes, volcanoes, things that I was always interested in as a child, kind of watching documentaries on TV. So my interests kind of developed from there. After I did my degree in earth science I went on to do a PhD at the University of Southampton learning about faults in Greece. I then went to work in New Zealand for a couple of years and now work at Imperial College as a lecturer. So where has your research taken place then? Have you visited areas or where is the data that you're looking at taking place? So I work in a number of different what we call tectonic settings, so areas where the tectonic plates are doing slightly different things. A lot of my research started off in Greece. So if you look at a map of Greece you'll see that right in the middle there's kind of this little strip of sea, a little strip of water and that's happening because Greece is breaking itself apart. The south of Greece is moving away from the northern part of Greece and this is a small what we call a rift basin, a place where two kind of proto-tectonic plates almost are beginning to kind of move apart and maybe in the future an ocean will form there. Other places I work on New Zealand and this is an area where the Pacific Plate is subducting under the north island of New Zealand. It's one of these subduction zones where potentially really big earthquakes can happen. So you've previously described predicting earthquakes to being a bit like London buses. What did you mean by this? So I think it's a really interesting analogy. So everybody wants to know how to predict earthquakes. We can't do it at the moment. Maybe we'll never be able to do it and I think it's interesting to compare the problem to trying to predict when a particular bus is going to arrive at a bus stop. Ideally you'd have some kind of timetable on the bus or you'd have the internet, you'd have a countdown but if you didn't have any of that, one thing you could do if you knew there were about every 15 minutes is just sit at the bus stop and wait for when the bus arrived and hopefully within 15 minutes the next one would arrive too. So you'd get an idea of the frequency between the buses. So let's say they came every 14 minutes and 37 seconds. So you could then say okay well every 14 minutes 37 seconds the bus is going to arrive and that's a little bit like some fault lines. Some faults regularly rupture in earthquakes at a reasonable kind of frequency. They might occur every few hundred years or every few tens of years. So if we know when the last two earthquakes happened we can get an idea of what we call the recurrence interval, the time period between the earthquakes and we could use that perhaps to say okay well an earthquake on this fault happened 50 years ago, the one before that was another 50 years, maybe in 50 years time will expect another earthquake. So if you hear any reports like people saying oh the north Anatolian fault is overdue for an earthquake, people are using those recurrence intervals to kind of get an idea of that. That's fine but it's pretty unlikely that the bus would arrive every 14 minutes 37 seconds precisely on the dot. And if we want to be able to predict when an earthquake is going to occur to the day or to the hour we need that level of precision. And although we've got some idea of the recurrence interval between earthquakes they certainly don't occur as regularly as being able to tell when they're going to occur to the year let alone the month or the day or the hour. And that's because lots of things can affect how stress builds up on faults and exactly when the stress can overcome the friction to fail in an earthquake. So for example if the fault gets permeated by lots of water that can change how much stress we need to build up before we can have an earthquake. Another thing that could happen too is if we have lots of land removed from over a fault like if you have lots of landsliding that remove some of the weight on top of the fault so the amount of stress we need to build up before we can have an earthquake changes again. So there's lots of things can happen that can change that recurrence interval. So although roughly we might know where it is we don't know exactly. And the biggest issue is that the biggest earthquakes might only happen every few hundred years or few thousand years. So as we've only been really kind of recording earthquakes in detail for the last 100 years if you have a fault that only ruptures once every 10,000 years then we don't have a good enough record to know that recurrence interval. So if your research has focused on rare earthquakes and as you said these rare earthquakes might not have as much data can you tell me a little bit more about particular type of rare earthquakes? So up until about 30 years ago I think we thought that faults could fail in one of two ways. They could either lock up and store stress for a period of time before that stress can overcome the friction and the fault slips in an earthquake. We call that stick slip behavior. 30 years ago we also knew that some faults don't do that, some faults just creep gradually. So they don't build up any stress, they don't have earthquakes, they just creep. About 15 years ago we discovered a completely new type of seismic event called a slow slip event. So these slow slip events release as much stress as about release as much energy as a magnitude 6 or 7 earthquake but it's not released over a few seconds like a typical earthquake, it's gradually released over a number of weeks to months. So the same amount of slip as a big earthquake is happening but it's happening really really slowly. So it's happening so slowly that there are no earthquakes, nobody feels it, there's no ground shaking, there's not even any seismic waves for our seismometers to detect either. So the only way we know they happen is that GPS stations, so stations which have GPS like your mobile phone, your satnavs have, we know that they are moving by as much as if a magnitude 7 earthquake is happening. So these are quite interesting, we don't know yet whether they are meaning that a large future earthquake is less likely or whether one of these slow slip events could actually trigger a large destructive earthquake. So that's one interesting new type of earthquake. The other one is something called a tsunami earthquake and the first tsunami earthquakes were probably described in the literature about 30 years ago but since then there's only been about 10 of them reported in the scientific literature so they're quite rare. And what they involve is an earthquake which is relatively small but it produces an enormously large tsunami. So one example is in New Zealand which is some of the events that I've been studying, they happened in 1947, so quite a long time ago now, they had a Richter scale magnitude of about 5.8 something like that which is big for the UK but it's not particularly big for somewhere like New Zealand that has really big earthquakes. It didn't cause much ground shaking but it was followed by a tsunami over 10 metres high so that's really weird. That's such a small earthquake could produce such a large tsunami. So one of their characteristics is that they tend to occur at very shallow depths under the sea so even though it's a fairly small earthquake it's still disrupting the water column a lot and another characteristic is they're quite slow so they rupture small slowly than a typical earthquake but faster than these slow slip events I was just mentioning. So the fact they're so rare does make them a little bit more difficult to study. In the case of the 1947 earthquake we have to rely a lot on eyewitness observations from people that were living in New Zealand at the time. These days we've got tsunami boys to measure tsunami heights but back then we didn't have it. So we've had to do a lot of trawling through newspapers looking for newspaper cuttings from eyewitness reports at the time. We've got reports such as one person says the tsunami waters came up to her rose garden so we've gone to the rose garden measuring the elevation and that's how we're getting our data points. So we're leading quite a lot on the people that were around at the time. Just a month or so ago I was in New Zealand and we were talking to the Council in Gisborne which is where this event happened and even in the audience there were people saying oh my mother was alive then she's told me this story so we've just made a new kind of appeal for eyewitness reports of what happened which we can kind of use as extra data points to help us model these events. Thinking about these rare events then and going off kind of this vernacular knowledge how does this lack of data and how does this affect people's preparedness for the kind of impacts of these events. So in the case of New Zealand from those 1947 events we know that these tsunami earthquakes these rare earthquakes occur and we know that one of their characteristics is that they're very slow and the ground shaking is very subtle. So the key is that it's subtle ground shaking but over a very long time for more than a minute. In New Zealand up until a few years ago and also in most places around the world the tsunami warning signs say things like in case of strong ground shaking head inland or go up high. In New Zealand those signs are now being changed and the public are being kind of told a new kind of mantra almost for tsunami preparedness which is long strong gone. So the message now isn't just if you feel strong ground shaking but also if you feel ground shaking going on for a very long time that should also be another warning for you to self evacuate. The problems with tsunamis is many people still think that they'll wait for a tsunami siren or something like that but the best course of action is to self evacuate as soon as you feel something. So these tsunami earthquakes are a bit dangerous because they don't have much of a ground shaking warning. So we are working with the council to let people in these areas that experience tsunami earthquakes know that this is a danger that they could face and what they should do. So as the Earth CO2 increases then and we're experiencing lots of different changes to our climate what might this mean for how landscapes are formed and experienced and what changes might we see. So it's an inclusive question I am in no means an expert on this at all but CO2 is changing our climate which is making things like storms and monsoons in some areas more extreme and more unpredictable. A couple of factors in the landscape brought about by weather and storms could affect earthquakes and faulting. So one example is if you have lots of storms it can produce lots of landsliding which changes the landscape produces lots of mudslides that we have seen recently in South America so they are devastating events in themselves. But also in areas where you do have faults you do have earthquakes those landslide slides can change the weight of the rock above those faults which can then change the stress on the faults could potentially make an earthquake happen a little bit earlier than it was likely to. So CO2 in the atmosphere changing the climate will certainly change the landscape in terms of perhaps more landsliding in some areas more mudsliding and that can could then influence the time scale of these earthquakes. It won't cause earthquakes to happen in areas where they wouldn't already happen but it could give them the extra prod to perhaps occur a bit more earlier than they would normally. Can you tell me a bit more about the methods you are using and also how you are advancing those methods in your research? So one of the key methods I use is something called seismic reflection imaging which is a method where we can see what's under the earth's surface. So it's really useful to learn about earthquakes because we can use it to try and work out where the faults are, work out how big they are and in some cases if we can also see sedimentary units either side of a fault we can look at how much they've been offset and if we know the ages of those from wells or from kind of dating methods we can even work out the slip rates on the faults and that kind of data is really useful for kind of seismic hazard models and risk models. So this technique involves sending sound waves into the earth which can be made through a variety of methods. We can do this at sea where we use something called an air gun which has air compressed into it. When we're ready we kind of press a button which releases the air into the water which produces an air bubble which then produces a sound wave. If we're doing this kind of analysis on the earth's surface we can use dynamite for example, we can drill a well 50 meters deep, put dynamite in the bottom, detonate it which produces our sound wave or if we don't need sound waves that energetic we could perhaps even just hammer on the ground to produce sound waves if we don't want to image very deeply. Those sound waves then travel through the different rock layers and get reflected and we detect those reflections with what are basically microphones and by knowing the time it takes those sound waves to go down and be reflected back up we can start to build a picture of all of the different rock layers in the earth and then that can help us see if those rock layers are disrupted by fault. So it's a technology which has been mostly used by the oil industry but us academics like me we're kind of pushing it and trying to use it to image even more deeply than the oil industry are interested in. Using it to image to depths of about 10 to 15 kilometres and see what the faults look like in areas that are actively, tectonically moving. Thanks for listening.