Instead of. Members of the Oxford Physics Department lets students, staff and faculty I'd like to welcome you to this physics colloquium, which is held today and the Oxford University Museum of Natural History, which was established as you all know in 1960, to draw together scientific across the university. And you should know that this lecture theatre today is thought, and this event is now being streamed live just outside the store.
It's the world's first scientifically described dinosaur, the making of science, but ran by the world famous Oxford Dodo. The only soft tissue remains extinct and remarkable, but today's lecture on gravitational waves and prospects for motion messenger astronomy is being given by the distinguished physicist Barry Barish, now at Oxford, where they'd be proud of the education of students. They go on to make new knowledge and then go on to careers that change the world.
And so an exhibition that we established with the first of these lectures last September. I've asked one of our students, Ben Fernando, as a Ph.D. student and physics and assumptions here working on seismic wave propagation and is also a member of the science team with the Russians Insight spacecraft, which recently landed on Mars to actually conduct the introduction for Barry.
And you'll notice when Ben comes up onto the stage, it is wearing a uniform because he is a member of the Navy gear that's attached to this university and actually you spend some of his spare time out at sea. I think about how should work this evening.
He's wearing a uniform because it's going to be conducting outreach to the local secret unit about universities such as Science at Sea, and he's wearing the uniform and making a group of people who, generally speaking, I've never gone to universities in their lives or come from different backgrounds. And this is a very important part of what we do in Oxford is reach out to the world. And so please welcome Ben, who will introduce part. Thank you, everyone, for coming.
We're delighted to welcome here today, a titan of modern science professor Barry Barish from the California Institute of Technology and the University of California at Riverside. Barry is originally from Omaha, Nebraska, and made his first forays into the fields of physics in the 1950s, when he gained a B.A. from the University of California at Berkeley and later a Ph.D. from the same institution.
From then, he went on to the staff at Caltech, first as a postdoc, working on experiments including those at Fermilab. Later in his career, Barry was involved in the macro experiment at Grand Sasso in Italy. And following that, the gem detector for the superconducting super collider proposal. In the mid-1990s, Barry became involved in like the Laser Interferometer Gravitational-Wave Observatory and was instrumental in setting up and directing the project.
In the last 20, 25 or so years, Legault came to a head. As many of you will know in 2015, with the first detection of merging gravitational waves, a discovery that was nearly a century in the making. In 2017, Barry was awarded a share of the Nobel prise in physics for his seminal contributions to the work of Lego, together with Kip Thorne and Rainer Wise.
We're delighted to welcome him here today to tell us a little bit about his life as a gravitational wave physicist and the prospects for science moving forward in this exciting new era of gravitational wave observation. So thank you. Very. Thank you, Ben. I've come to Oxford, you know, every few years through my career, so I'm happy to be back. I saw new physics building today that wasn't here last time I was here.
So hello to my friends and colleagues and the rest of you who came to hear the lecture. I'm going to talk today about gravitational waves themselves, what it takes to measure them and see them. But in particular, I'm going to concentrate somewhat on what I think this opens up in terms of part of what the future will be for gravitational waves. The part that I won't do is the deep physics. The gravitational waves present us with probably the best way to test general relativity.
I won't talk much about that. What I want to talk about is what I think is a new direction that you'll see in the next decades opened up partially by gravitational waves, which is being called multi messenger astronomy. So I'm going to talk about gravitational waves and then try to give you a picture of where we think this is all going. And in a sense, we're it's a little bit presumptuous, but we're in the same situation that maybe the science of astronomy was in at the time of Galileo.
Galileo was the first to take an instrument instead of our naked eyes and look at the sky. Famously, he looked at Jupiter and with a telescope and discovered that there were four moons of Jupiter. Now we know there are more, but there were four moons of Jupiter. It's been more than 400 years since then. And astronomy has evolved fantastically in that era.
We're at the very beginning. I don't know if it'll be nearly as rich, of course, as astronomy, but in a sense, we're at the very beginning of having come to the point where we're able to see signals from the universe that come from astronomical events that emit gravitational waves. And so you'll see how we do that and then what that might lead to in terms of astronomy today, mostly or what we call multi messenger astronomy.
First, astronomy a little bit in recent years, what we've seen happen in astronomy is an advance that's given a lot of richness to what we know about. The universe through astronomy by using different ways, using the electromagnetic spectrum to look at astronomical events, the same events, but with different probes. And I, for example, I show here the Crab Nebula looked at in different ways.
So we look at different parts of the electromagnetic spectra and look at the same phenomenon, and then we have a lot more information that we can use to do astronomy. In a sense, this is one of the big advances in astronomy that occurred in the last half of the 20th century and has led to many of the advances that we know about in astronomy as maybe a precursor to what I'm talking about today. We've seen other advances which. Have happened as we've moved into the 21st century.
We've seen a really fantastic demonstration of how to do science on a worldwide basis in astronomy by combining a set of devices that weren't designed to work together. These are infra-red telescopes in various parts of the world to make an interferometer that can look at the sky and work as an instrument that basically is the size of the Earth, instead of the size of one of these interferometers that relied on a lot of very sophisticated methods to bring the data together.
Computer computing around the world, bringing this all together, finding a target to look at. And we saw just early this spring these beautiful images of a black hole taken through these radio telescopes from around the world and at least in the US, it was a big colour picture on the front page of the New York Times.
So astronomy has been fantastic. The next frontier, I believe, is what we're calling multi messenger astronomy, which is really what I'm going to come back to after I talk some about gravitational waves, of course, an electromagnetic setting the universe through electromagnetic waves. We now have a whole set of sophisticated instruments that use the different wavelengths, some on the ground, some in space. We have a new generation of those that'll be coming along within the next decade or so.
In companion with that in companionship, with that, we now have the beginning of the ability to see a phenomenon that happen in the universe with gravitational waves, which I'll talk about and also with particles in this case, the particle that, well been coming through, and that is neutrinos. So in the future, I believe that we'll see a real richness in our ability to study phenomenon with a combination of sophisticated electromagnetic instruments that we're all used to.
And the next generations of those. The gravity, gravitational waves and neutrinos from the same phenomena. So so far, we're just in the infancy. Now I'm going to go back and talk about gravitational waves and then lead to what we have done and can do, I think with multi messenger astronomy.
We all learnt. Most of us, I think our gravity when we were in elementary school and our teacher taught us that when the apple falls out of the tree, the Earth pulls it down and the moon goes around the Earth, all due to the equations of Newton, which I show here. And that equation of of of Newton's is his theory of universal gravity, which is g times the product of the masses of two objects over the inverse square of the distance they are apart.
It took about a hundred years more to determine the strength G. But this basic description of gravity described. Everything that involved gravity and in nature for more than two hundred years, it described everything from the orbits of the planets, the dropping of the apple, the tides and so forth.
And basically, by the time Einstein came along and introduced a new theory of gravity, there were no major flaws, no major problems with excuse me, with Einstein's theory with with Newton's theory of gravity.
So this is the best theory we had, Einstein came along. I'm not going to talk, I just show it for symmetry about the detailed equations of Einstein's theory that he came along in the early 1800s and made a new theory of gravity based on bringing space and time together in four dimensions and the unified space time. First question is why? Why do we need a new theory of gravity when we had such a successful one by Newton for all this time? I think the answers primarily are two.
One is, there was at least one case of a observation that didn't exactly agree with with Einstein, with Newton's theory of gravity. That was the orbit of Mercury around the Sun. That wasn't a very big flaw, but that was the one discrepancy that existed in the early nineteen hundreds. But there were conceptual issues and Newton's theory, too. One is that it had what we call instantaneous action at a distance.
That is, when the Apple falls, you detect it immediately, and that's fine for the apple falling. But if the Sun were to disappear right now, we know that it takes eight minutes for the light to arrive here. And certainly the effect of gravity is going to take some time, we believe. So Newton's theory didn't have any time for the messenger or the signal to get to pass through space. Einstein Einstein's theory the signal from gravity travels at the same speed.
There's just one speed and the problem we call it the speed of light, but it's also the speed of gravitation and gravitational waves. The second problem that existed in Newton's theory, I find personally an embarrassing one, because when my teacher told me that when I jump up, the Earth pulls me down and I believe that maybe later went to learn this formula. I never asked why. And I don't know if any of you did or many of you did, but probably not very many.
And that question, which is a fundamental one, was not answered by Newton's theory. Why does your on Einstein answers that and Newton's case? People tried to answer it for the hundred and some years before Einstein came along, usually with explanations that were electromagnetic in nature and weren't really right.
Einstein basically does it by the fact that he in this four dimensional space, there's what we call curvature of space time, and it's that that affects the poles when you when the apple falls out of the tree or that you fall. So in this picture here, I show now the generation of gravitational waves.
Electromagnetic waves were discovered by Hertz in the late eighteen hundreds by taking charges, isolating them, making a dipole, going in the next room, detecting electromagnetic, the electromagnetic signal from electromagnetic waves, a signal and moving forward and backwards, and seeing that it had a wave like nature. Ideally, you'd like to do the same thing for gravitational waves. It's not a dipole. Instead, it's a quadrupole in the case, a source in the case of gravitational waves.
But likewise, you'd like to detect them move forward and backwards, just as Hertz did, and control all the variables that is, make a sauce. In this case, it would be a big bar bill that you'd rotate and detect gravitational waves because it's a quadruple. It turns out that that's not at all possible.
I could go through the numbers with you, but to compare all this data, no, if I tried to do that, inventing my own experiment, near-surface taking a barbell like shape, rotating at a high frequency, moving, putting a detector nearby. The key number is that the sensitivity I'd have to have in the same units I'll show you and a little bit are about 10 to the minus thirty eight.
I'll show you that we can barely detect gravitational waves when we have a source strong enough to make the Signal 10 to the minus twenty one 17 orders of magnitude better. So we couldn't do what a physicist wants to do, which is an experimental physicist that is to control all the variables to get it right. So you want to control the source, control the detectors, control the data. We can't do that. So we're forced to find a source that can do part of the problem for us.
And that's what drove us to look for gravitational waves from a source in space with the. With the negative part now, we don't control everything and may not understand that source very well. And of course, the positive part that we can find sources that will give a much bigger signal while we're fairly primitive in our ability to detect gravitational waves. So that's what we did. Luckily for us, the source that we found turns out to be fantastically interesting in itself.
So the deficit or the bad point as experimentalists, you don't like to have anything out of your own control in this case gave us a source that in itself was interesting and that we talk about probably more than we talk about the detection ability itself. OK. Gravitational waves. Themselves were first proposed by Einstein a year after he introduced general relativity in 1915.
He introduced general relativity, having put acceleration into the special theory of relativity that he introduced in 1945 and the theory of general relativity became the new theory of gravity or his theory of gravity. In 1916, he noticed that if he wrote the equations of general relativity in a particular way, I'm not doing it quite like he did. And only those who know a little bit more about general relativity can read what I did on the left side.
He basically noticed that if he set up the equations of general relativity in a particular way, the equations of the letters are different looked a lot like the equations of electricity and magnetism. He didn't derive the fact that there were gravitational waves. Instead, he just made the leap in Einstein that if the equations looked like the equations of electric, just magnetism, then there must be waves and gravity just as there is an electricity and then magnetism.
So he proposed that there are electromagnetic waves. The paper itself was a terrible paper. It had errors in it. He didn't derive anything, but he made this presumption. Two years later, in nineteen sixteen nineteen eighteen, he wrote a second paper put on that second paper fix. He fixed the errors. He never admitted they were wrong, but he fixed the errors.
But he did something else that was important for us, anyway, is that he demonstrated what kind of sauce would make gravitational waves that you need a quadrupole source. He only wrote one more paper on gravitational waves in his lifetime, and that was 20 years later. So in nineteen thirty six, he had immigrated to the US and was working at that time with Rosen.
There's famous work of Rosen and Einstein. At that point he was at Princeton University, and he revisited with Rosen the question of gravitational waves because he wanted to see if he could derive it out of the theory of general relativity itself, rather than just postulating it. He failed to do that. And he, in his calculations with Rosen, got a bunch of infinities and his calculation. He then decided he must be fully himself. And he had no idea that there really weren't gravitational waves.
There was an artefact. It turns out in general relativity for those who have ever studied it, it's easy to get an infinite is what we call coordinated singularities, trying to work in four dimensions. So it's a typical kind of problem. Anyway, he wrote a paper with Rosen. He submitted it to Physical Review Letters, which is the same place we submitted our paper. And in that paper, the paper was entitled On Gravitational Waves.
No, I'm sorry. It was entitled Do gravitational waves exist? A funny title for somebody who had proposed it 20 years before that paper. It's a long story. I'll just tell you the bottom line. That paper went through a peer review process and was sent to a reviewer named Howard Percy Robertson, who happened to be at Caltech my institution on sabbatical from Princeton at that moment.
And he saw how Einstein and Rosen had gotten these mistaken singularities and actually even showed how to get rid of them by using what he called cylindrical coordinates without giving the whole story that eventually went back to his physical review editor, who sent it to Einstein. Einstein then decided he wasn't going to publish in Physical Review. Rather, it wrote a rather negative letter. He never published again in Physical Review.
And instead, he published the article about six months later in the Franklin Journal, which is a journal that doesn't exist today, but is from the Franklin Institute in Philadelphia. And that paper changed the title to on gravitational waves and starts in the first sentence, describing gravitational waves and cylindrical coordinates. But that still didn't convince the theoretical community, and it's the last paper that Einstein ever wrote on gravitational waves.
It was about the 1950s when finally the theoretical community believe that there were gravitational waves and it happened in a meeting in Chapel Hill, North Carolina. And in that meeting, a actually a theoretical general relativity called Peroni, who I think was at this institution derived gravity gravitational waves from the fundamentals of general relativity.
And at the same meeting, my colleague Dick Feynman was there, and he made the presumption that if there are gravitational waves, they have to be able to transfer energy. And he made a little good Duncan experiment that could demonstrate how that could be. So I don't have time to go through that today. This is all an excuse why it's taken 100 years to realise I'm talking about theories. So after that, in about the 1960s, it became an experimental problem and the idea was first tuned scene.
Did you have to try to detect something from the universe? And I show here the generation of gravitational waves. You notice the top equation. If I set up general the message, the take home is if I set up general relativity in a particular way that I describe on the left, I get equations that, except for the letters, look like the wave equation. And so it's not surprising that I get basically a plane wave out of those equations and the plane wave I show in the next.
And that just below that and the little h that's there is the key parameter that we try to measure and gravitational waves. We call it the string. It's the strength of the gravitational wave, the amplitude. The gravitational waves travel at the speed of light, just like electromagnetic waves. You'll notice in this picture, though, that the two waves are not what you learnt in electricity and magnetism. They're at forty five degrees to each other instead of 90 degrees to each other.
And that's a direct result of the fact that gravity is spin, too. So being spin to instead of spin one, there's two components, but they're at forty five degrees to each other. We're on the verge in what we've measured. I won't show that today to be able to just to touch, to change or to pull out the two different components of gravitational waves, which turn out to be important for testing different ideas about general relativity might be like and whether there's variance of Einstein's theory.
But so far, we haven't really been able to do that, but I think it's very simple. OK. So gravitational waves themselves come out of something like the picture that I do here. That is two objects going around each other, you can imagine those that say black holes. Gravitational waves come out. They go at the speed of light. And what do they do first? They're not like electromagnetic waves going through space, electromagnetic waves have associated with them photons.
In the classical theory, as long as we stay with Einsteins classical theory, then there is no propagator art equivalent of the photon or something we might call the graviton that goes along with the gravitational waves. Instead, it's effectively just ripples in space and time itself, a little bit like having a pool of water still full of water. You throw a stone in the stone, sinks to the bottom, and there's ripples that don't have any part of the stone in them that travel along.
So you instigated ripples that travel in space and time, and we're looking at that. So there's a curvature of space time, the amplitude of the waves for the kind of sources that I talked about today is about one part in 10 to the twenty one that little each that was in my formula that turns out to be a really small number. I'll show you how small in a minute. And that's the number that we actually measure. It's proportional to something directly experimental.
That is the distortion of space or length in a particular direction. So Delta L change of lengths over length, which is what we call experimentally measure, is proportional directly to this little h, a number that's 10 to the minus twenty one for the sources that I'm talking about today. So imagine that I have a circle of three masses and now a gravitational wave comes through the board.
What does it do? It basically distorts it so that it's a little taller and that length is this Delta L over L or 10 to the minus twenty one. If it's a metre in size, so it's 10 to the minus 21 metres and depending on the wavelength or frequency of the gravitational wave, it then oscillate the other direction becomes shorter and fatter. So it's a little bit like going to an amusement park where you're in front of a mirror that makes you a little taller than the next one,
makes you short and fat and you go back and forth. And that's what happens when gravitational waves come through. Why is the effect so small? It basically comes down to the fact that if you think about space, which is being distorted as being, say, a material that the Young's modulus or space itself is just very stiff. So luckily for us, because we don't want space to change, too much space is very stiff.
These gravitational waves then can't do very much to it. So it's not that they're weak or don't carry any energy. It's if that space itself doesn't get affected very much that the effect is so small going through space. OK. So how do you measure a very small thing? We do it with interferometry. This is a picture of an interferometer. Light comes in from the left and as it comes in, it's split in two directions. If those two directions are equal in length, then the light goes down.
Shown in wave form, here comes back. And if you invert one with respect to the other, they'll just cancel. This is showing them cancelling and now hitting our detector, which would see nothing. But if one of the arms gets a little bit longer or shorter than the other one, then they won't completely cancel and we see some light. And that's the principle that we use. So that's called interferometry.
And we do lots and lots of tricks and have two to make this a very sensitive instrument, and I'll just give you a sense of the key ones. So this is a picture exaggerated of a frequency of a couple of times a second gravitational wave going through an interferometer. So it makes it tall and it makes it longer in the vertical direction first than the horizontal direction and goes back and forth the size of the interferometers l in this picture.
The accuracy that we have to have is to measure the difference in length in order to get this 10 to the minus twenty one of the light that's in the interferometer itself. And people who have ever used an interferometer know that you basically use an interferometer in a lab and you see these fringes, which come from the wavelength and you can measure something to some fraction of those fringes, maybe one in 10 or one in 100 if you're really good. We have to do it to one part in 10 to the 12.
So the first part of the challenge, I'm just simplifying the challenge for us and the two numbers for you. One is that we have to do interferometry to an accuracy of one part in 10 to the 12th. That's a small number of the wavelength of the light that we use in the interferometer itself. Interferometers that are used for very sophisticated purposes might be one part in a thousand or ten thousand or something. But this is orders of magnitude beyond that. And it was the principle.
Target goal that took us years and years to develop enough tricks and means to measure interferometry at the level of one part in 10 of the 12. It turned out, though, that the second challenge was even harder for us, even though maybe it seems more mundane.
So the two challenges that we have to succeed at to be able to see something at the level of this small 10 to the minus 20 one were first to be able to do interferometry at one part and 10 to the 12th of the wavelength of the light that we're using. The second is the fact that we're living on Earth and the Earth shakes like mad in the frequency band that we're in.
And so we have to actually do interferometry in a way that's different than in the laboratories here on campus, where the first thing you do is go to an optical table and put your mirrors down for them as stable as you can. We can't have anything to do with that because the Earth shaking is much too much. Instead, we have to basically isolate the interferometer even though we're living on the Earth from the Earth itself to one part in 10 to the 12th.
And that's a tremendously small number. I'll tell you how we do that. And finally achieving that was what was needed to to see gravitational waves. So if you want your elevator speech when you leave, I'll show you what was the final experimental or technical accomplishment that we made. It's pretty mundane. In order to make this measurement, so we thought we could do that, we didn't have all the technologies in place.
We got funded by the National Science Foundation and approved to do a project where we would develop two interferometers, two so that we could make sure within the speed of light that we saw the same signal are nearly the same signal in both interferometers. We proposed to the National Science Foundation in the US that we build them. We weren't allowed to pick a site, but of course, nobody can stop us from suggesting.
So we turned in a proposal where we made what we called sample sites, which was our choice, of course, one in Southern California near the Edwards Air Force Base and not too far from Caltech. And the other one in southern Maine, just not too far from MIT. It went into the political process, and you can see that it got rotated by about 45 degrees, and we ended up with two very friendly and supportive senators from the state of Washington and the state of Louisiana.
These two detectors are now three thousand kilometres apart. And so they should see a gravitational wave within the speed of light in both. So if we have if we have a. If we have a gravitational wave coming straight down, it'll be at the same time in the two if we have one coming in from this side. You know, if this detector 10 milliseconds before this detector and this one, 10 milliseconds before that detector.
So we expect that a gravitational wave will be within plus or minus 10 milliseconds of the same time. And that's a key feature of what we do. We then look for out of time coincidences. We're one of them is 20 milliseconds away from the other one or 10 seconds away. And we do that to see if there's any accidental. And that's how we determine our our ability to actually see a signal over accidental signals. So that's the technique generally the two devices.
Are shown here. They're identical. They don't look like it in this picture, but they are absolutely identical. One, however, is and they're in different geometry geographies. The one on the left is in Hanford, Washington. That's on the D.O.D. reservation, where they had reactors to develop the atomic bomb years ago. And we're we're not we have nothing to do with the deal. We accept that we're on the land and it's high desert.
As you can see from the picture, the one on the right is the second detector we have. It's in Livingston, Louisiana, and it's in commercial pine forest, where they cut down trees every 10 years. Let them grow, cut them down every 10 years to make paper. And we live in those, too. And as I'll show you, the sensitivity of them are identical, although they're very different. The one on the right, by the way, is swampy. So you can see water here.
There's water if you go kind of one millimetre below the surface in Louisiana. And we built this up. You can't see it very well in the picture, about six metres above the surface so that it wouldn't get flooded. The dirt that we got to build it up here, we took from here and it immediately filled with water, fish, alligators and everything else. And then the locals kill the alligators, so it's interesting, OK?
This pictures and what happens, you put all this together, what limits you and what enables you to make a measurement? What I'm showing you on the left is the sensitivity. So first, looking at the scale on the scale, on the right side is this sensitivity. This is the 10 to the minus 20 what, 10 to the minus, 20 to 10 to the minus 20. The sensor to the shaded region is where we should be sensitive. All these lines are the different ways that affect our, our limit, our sensitivity.
So you'll see that we're limited by three different lines immediately. Again, as an experimentalist who was taught when he was in school that you should control all the variables. I'm already not doing that. I have a terrible situation experimentally where we're limited, not by one problem, but by three problems. So we have three different experimental problems to be at low frequency. We're limited by what I wrote here is seismic noise. That's the shaking of the Earth.
So this is how will we get rid of the shaking of the Earth that affects the low frequency and that's shown here. It falls like frequency to the fourth power. So even if you get a little better, you don't gain a lot at high frequencies. This is the frequency band at high frequencies. We're limited by what people that work on interferometers called shock noise, or somebody they say comes from particle physics or quantum physics.
We call for statistics. So basically, we're limited by how many photons we have, which is less and less as we go to higher frequency. And in the middle level, we're limited by something else, which we call thermal noise. And that's basically just Caity noise because we're working at room temperature, so we're limited in these three regions. You'll notice that the frequency band is the frequency band, the audio band.
So we have nothing to do with audio. But if we're living here on Earth and we want to make the most sensitive instrument, our ears have learnt for us through evolution that the place where you're most sensitive is in the audio band. So we're in the same band of frequencies where the Earth, if we're going to lower frequency shakes too much at higher frequency, you have to sample too quickly and you don't get any photons.
So we're basically limited to the same frequency band as the audio band that we all communicate with. So for that reason, people have taken our signals, for example, and make an audio signal out of it and transmit it, and some of you have probably heard that. I don't do it because I'm too much of a purist because the frequency that we saw in the very first event was very, very low frequency. The ones you heard on the radio actually were multiplied by four to make them better for the years.
And but it's in the audio band. The graph on the right is kind of the experimental history. So we started in the early 2000s and we had the same plot as this, you know, notice it has the same characteristic shape. That's good. But the sensitivity is up here somewhere. This was already the most sensitive instrument in the world, so we were able to try it run for maybe six months, write some papers that we didn't see anything more sensitive than ever before.
Lick our wounds. Put in some new devices and try again, and then we come down to here and come down to here. Subsequently, we came down to what's shown down here and the dotted line is basically when we came to the design limit, or we knew we would be limited by truly by photo statistics or the remaining shaking of the Earth or thermal noise. So at this point, we hadn't seen gravitational waves, but we were ready and we had already proposed to the NSF how to go to better.
Fortunately for us, the National Science Foundation stayed with us, even though we've been doing this for quite a few years and we started a major upgrade. We called the upgrade LEGO to natural thing that happens in experiments. They rejected that. They said If you're proposing Lego to your, you're inferring there's going to be a Lego three. I go for it. Therefore, we changed it to what's called advanced LEGO, which you've probably seen in the in our papers and newspapers and so forth.
Now we have a new version that we're working on, and we call it Advanced Sligo Plus for this kind of twisting it around. And so it's called A-plus often, which is advanced Lego plus, and that'll be the next few years anyway. The improvements were many. We had started getting experience with an interferometer. But the NSF was supportive enough that they let us develop the technologies to make it better while we were doing the graphs that I showed, the curves that I show you here.
And so we were ready to do a major upgrade with the R&D programme and engineering programme that we had carried out that. Involved lots of changes, we had done one smart thing, and that is that we built the infrastructure for Lego from the beginning so that it was bigger than it needed to be, had more ports and flexibility so that we didn't have to go back to the NSSF to rebuild a lot of very expensive infrastructure,
but rather for instruments and techniques. So that was the smart decision that we made early. Anyway, we essentially rebuilt all the parts of Mica, which to make Advanced Sligo, which I show here on a chart to show you there's more to it than what I'll show you in a minute. That enabled us to have a more powerful laser to have a quieter thermal test, mass and so forth and so on. But the key one that enabled that, I promise you that it enabled the discovery is shown here.
So let me describe the problem. I already said that we have to isolate ourselves from the Earth by one part in 10 to the 12. So how do we do that? We didn't invent anything magnificently clever. We just did a good job of something that seems pretty mundane, which once you understand it. The first is that we work in this frequency band from 10 to ten thousand hertz, the audio pan, and we want to get rid of the shaking of the Earth.
So what's the technique that we all know pretty well? Many of you used it today in your car shock absorbers. Shock absorber doesn't get rid of the energy. It just takes the energy and moves it from the frequency band. That gives you a big bump when you go over it to low frequencies where you smoothly feel your car move along. And so we did the same thing. We made the world's fanciest shock absorbers for different layers of the shock absorbers.
The shock absorbers themselves have just the right squishiness. We did a lot of engineering and we put that all in, and that was what was in for the measurements that I already showed you and we didn't detect gravitational waves. So we had to do something more than that and we took another. We knew this in the 1990s, not 2000, but we didn't hadn't developed the technology yet.
So we asked the NSF during that period, we were making those measurements to develop the technology, and we had it developed by about two or five. And that is again, a technology that many of you are familiar with, and that is, you get on an aeroplane and you don't like the roar of the engines. And so you get one of these Bose or somebodys earphones that cancel the noise of the engines. What that's doing is measuring the ambient roar of the engines and basically cancelling the ambient noise.
The stewardess comes up and asking you if you want a drink and you hear perfectly well because that's not ambient, so you're able to take out that noise. Again, we're not doing anything audio, but it's the same problem. So we basically decided we had to do something that wasn't totally passive equivalent of what's done in the earphones that is cancel actively the shaking of the Earth.
So what we did is embed in these shock the system with shock absorbers, active size seismometers that measure the amplitude and the direction of any remaining shaking of the Earth, and we cancel it by pushing a force against it that took a long time to develop. We've developed, I think the most stable tables in the world could be very good for microelectronics or something, but they're incredibly expensive. So somebody else has to make them cheap anyway.
With that, we got this factor of 10 to the 12. And one of the thing that happens is we use a pendulum and a pendulum. You all know when you're holding a pendulum and you move the top, it doesn't move the bottom very well. So we get rid of most of this kind of motion with shock absorbers and this motion by having a pendulum.
The pendulum itself is four levels of pendulum so that all the controls for the bottom, which is the mirror, are controlled from above, and it's rather fancy to be able to do that. But that's what we do so that we have a pure, very, very fancy mirror on the bottom. I don't have time to show you our other technologies today, so I'll move on because I want to show you a little bit about the potentials of multi messenger astronomy.
OK. So this is where we were on the left with two the two interferometers at the time I brought you to where before we installed the improvements. And and it has this characteristic noise curve. By the way, these little lines here are the natural resonances that are in any system, whether they're electronic.
In our case, at multiples of 60 hertz or whether they're mechanical because of the mounts and so forth, those represent stable little lines that we notch out, and it represents less than one percent of the area. So don't concentrate on those, they don't matter. These are the two best sensitivity curves we had once optimised for low frequency, one for higher frequency. So they're slightly different, but that's where we were. No gravitational waves and been detected.
Then we decided to do this improvement programme and our goal was to improve by a factor of 10 everywhere. At least a factor of 10 means that we measure this little edge, which is an amplitude, not a power. It means we are sensitive a factor of 10 more. If we're a factor of 10 better, we look for a factor of 10 further out into the universe, which means there's a thousand times our ten cubed as many galaxies stars potential sources.
And so our goal was to get in a factor of 10. At high frequency, we do that by making a higher power laser, which we've done at at the middle frequencies where we're limited by thermal noise, by making a better, quieter test mass suspension system. For example, we moved all the electronics off the final test mass made it bigger and heavier and better engineered test mass and so forth. And at the lowest frequencies, we basically are limited, as I told you, by the shaking of the Earth.
Seismic noise. We were approved. We did this upgrade programme between twenty eleven and twenty fourteen or fifteen. And we turned back on turning back on. We didn't immediately try to get all the improvements in, but went partway partway by not turning the laser up to full power, not completely in all coordinates, being able to cancel in the active seismic everything. You have to do a lot of measurements to do that. But we had gained everywhere.
At least a factor of three and a factor of three for us is a factor of twenty seven and rate are possible sources because of the cube. So it didn't take very long to have more data than we had before. But you'll notice the low frequencies here. We actually had gained a factor of 100.
That was because we put it in the active seismic isolation. So it's the elevator take-home speech if you want, because at that point, we had gained a factor of 100 cubed and our ability to see gravitational waves at low frequency. And that's why within a few days of turning this detector on, we could see something we hadn't seen in ten years of previous running.
So that's the final message experimentally. We then turn this on here, even though we're not all the way down to the bottom because we're anxious. We are anxious to try to get experience with it and then continue to improve it. Then on September 14th, 2015, we saw this. So this is the picture that many of you have seen, and I've even seen it on Madison Avenue in a dress shop where they put it on fancy dresses for a woman and made ties for us that we wore in Sweden.
What you see is first, the scale from here to here is the 10 to the minus twenty one. So we can see something that happens then that scale the scale. This direction is about two tenths of one second. And the signal is this we're going it gets narrower and bigger as it propagates along. That's the so-called chirp signal as they come together to objects is I'll show you the frequency gets higher and the signal gets larger until they merge and then they merge and come together.
This is the signal in Hanford, Washington. This is the signal in Livingston, Louisiana. This one was seven milliseconds before that one. This is the two put on top of each other, and they look very much alike. So at that point, we. We're pretty sure that we had detected gravitational waves immediately looking at the data even more impressively.
If I take that data and I calculate the expected signal for gravitational waves using Einstein's equation, then the signals themselves are shown in the top frame. Here, the Hanford, Washington and Livingston, Louisiana, and these are the calculated signals for two, essentially 30 solar mass black holes. Merging into each other to give a gravitational wave signal as calculated using Einstein's theory of general relativity.
Then I subtract one from the other down here, and what you see is just wiggles. That's the noise in the system. Nothing systematic, and that's the easiest way to look at what we saw initially immediately before we did anything much with the data. I just show this to illustrate that it's the low frequencies. I've started off everything below 40 hertz or so. Uh, the signal. This is a time frequency plot that it was only by improving the low frequencies that enabled us to make this measurement.
Interestingly, even though we only make one measurement that she has an incredible amount of information that we can calculate the shape, we call it a chirp pass, but it doesn't matter. I've just shown you the shape can be written down in an analytical form that's quite detailed. There's there's a second set of terms that I just don't show here that allow you to measure all the detailed parameters.
So when people have wondered how we can take one event and tell so much, it's because there's really a form, a formulation that's in detail where we can tell everything from the masses how far apart they are, how far away they are, the chirp mass, the ratios of the spins, how redshift than it is, and blah blah blah, so we can measure almost everything from this shape. There's also orbital perception and spin misalignment and so forth.
We don't see any evidence for that yet. That's not in this term. It's in the next terms and they're harder to measure. We also get the sky location from the time the signal arrives in the two detectors from the amplitude, we get to the distance away and we get the binary orientation information from the time delays between the two. So basically, we get all that information for the detectors.
And from this first event, we were able to publish the first paper that told us that what we have seen is two black holes, one thirty six solar masses, the other twenty nine solar masses with a reasonable error on them giving a final black hole after the merger of sixty two solar masses. You'll notice there's some left over here. That's the energy that was transmitted away in gravitational waves, which is significant.
And in that, two tenths of a second was the brightest object in terms of energy in the universe. Then we can measure something about the spins, how far away it is and so forth. So all of that we knew from this first measurement, which was the discovery of the gravitational wave signals. We generally can look for gravitational waves much more sensitively. Then what you saw there and we expected and use for most of our events since that time, a template bank.
And the idea is to calculate those forms for as many from the from for the whole space of possible masses and and and mergers that we might detect. And then compare that with the data we use what's called a match filtering technique that is to take the shape that we calculate.
Multiply it by the background. And when you do that, if you run into some real signal like I'm showing this illustrative one, you actually find you can pick it out even though it's smaller than the noise because it's there for a while. And by seeing that spike, we then go in and analyse the data and pick out the merger by its shape and signal much below the actual noise level itself. That's the technique that's used for most of the signals that we've seen.
This is the second event that we saw, which was in December of the same year, 2015. And you can see that it's amplitude is lower than the noise level. But there are many, many oscillations, not like the very first one that you saw that had about six oscillations. The reason there's many is that these were not as heavy. And so they took went to higher frequency before they finally merged.
So using that, we basically have done a lot of work to look at that detail wave form and ask how well does it really in detail fit or not fit Einstein's equations of general relativity? The fit doesn't disagree at all, and everything we've done so far and the best tests are the ones that oscillate many times like these. Not the first event which was bigger, but we see no evidence so far of any deviations from Einstein's theory. OK, now let me move on. We have two detectors, which I showed you.
And in August of 2017, we had a third detector join us and that was a detector in Italy made by the Dutch, the French and Italians called Virgo. And we immediately saw an event on the 14th of August Twenty Seventeen, which is shown here. And what to notice is this is a picture of the Earth on the right and how well we can tell from when the signals arrived, where the signal source was, where it came from.
And you'll notice as long as there was Leo itself, which is all these top banana looking shapes, we couldn't tell very well. But then when we added this third detector, not surprisingly, we get a much better localisation of where something's coming from. That's a key to enabling us to be able to work with astronomers and with astronomy to try to actually know where any signal we have comes from in the universe. We could do that to about a thousand square degrees when there was only like.
And now it's a 10 or 20 square degrees with the addition of another particle. I'm sorry. Last thing on General Relativity, before I got into astronomy, we are in a situation where you'd like to test Einstein's theory of general relativity against any sort of other possible theories of general relativity.
Unfortunately, most of the very variation of theory theories that are variants of Einstein's theory are not predictive enough so that we can use those directly to ask whether they fit what we see better or worse than Einstein's theory. So instead, we look for deviations from Einstein's theory. I'll just show you one. We haven't seen any. If we look and just ask, is there any evidence of dispersive term?
What might give a dispersive term if we had some mass where people like to call it gravitation connected to the gravitational waves that would put a dispersive term in if we added dispersive term to Einstein's equations, that changes the waveforms? How much mass can we allow to put in before that? They don't fit the data very well. It turns out we're able to limit the mass of a graviton to 7.7 10 to the minus twenty three f overseas squared.
So we don't have any evidence in that formulation of any deviation from general relativity. But we're just beginning to detect. We've now reported 10 examples of mergers of black holes. They're all shown together up here and in the mergers of the black holes you'll see. They look different. If you look at the wiggly lines, some are longer, some are shorter. Now they're starting to be different masses, different orientations.
And we're starting to develop in the information to see, for example, the distribution of masses, distribution of distances and so forth that will, we hope, enable us to understand the origin of these heavy black holes. They weren't expected by astronomers, so what we discovered was not expected by everything we knew from electromagnetic measurements. They're heavier than what was expected.
We're now trying to understand what the origin is of those by making enough measurements to constrain the problem. What is the distribution of the masses? Spins are the spins of the two or aligned or not aligned and so forth, and we're just starting to get enough data to do that. I want to move on from that because these events have nothing to do at this point, at least with multi messenger astronomy, since they don't give any electromagnetic signal. Ah, so far none has been seen.
If you look at the bottom right, you see the 11th signal that we've seen since the original, and that has a timescale that goes four minutes. Instead of two tenths of a second, it goes for like 60 seconds. Very different. And it goes to very, very high frequency, it turns out. And you can see the signal here. This was a signal that we analysed and showed was a binary neutron star merger instead of a binary black hole merger.
First, let me say I'm a black holes, I said this already. So. Once we saw that, then there's the immediate question, once we see something like a neutron star merger that it's a nuke, it's not a black hole, it's a nuclear physics system that combines together. What, if anything, can be seen by electromagnetic instruments? The first step in doing multi messenger astronomy? In order to do that, we have to be able to develop and give information very quickly to the electromagnetic community.
And so what's shown here is that we can develop quite quickly candidates the sky location by what happened in the three detectors, and that's within a few minutes and we can alert astronomers. We then try to do event evaluation, which can take up to 30 minutes just to make sure that we can avoid it immediately if we if they're starting to turn their telescopes.
And if we then try to understand more of the physics, we have to fit into Einstein's equations and so forth, and that can take hours or days or weeks or even a month or so. So this process for us in establishing it from the gravitational side has an immediate alert and as best information we have to give astronomers. And then if upon analysis wasn't real, we say oops and cancel it. But otherwise, if they are interested, they can turn their instruments.
This first event that we saw had the feature that we were able to locate it in the sky quite well. So originally, as I said, the sky locations we had were typically hundreds or even more than a thousand square degrees. That's not exactly vocal enough for astronomers to want to turn and use their instruments. But once we added the third detector just in time to see this event on August 14th and on August 17th, two or three days later, we got lucky.
All these events, by the way, with black holes have been looked at to some extent with electromagnetic devices. Nothing has been reported. No one's seen anything that they've reported with from any of the 10 events we've seen the 11th event, the one that I just showed you, we saw in all three detectors.
Because the Virgo detector was then working, we were able to locate the position in the sky to 31 square degrees, and we could tell by the amplitude of the signal there was 40 plus or minus eight mega parsecs away. That's pretty good. So at that point, this is just what that signal sounded like. So it's already detectable here and the signal you can't see very well, but this this time versus frequency.
And as we get near the end, you should be able if your ears are good to hear the final what we call chirp signal. But I'm not changing the frequency. You're going to hear the real frequency. So as it gets near zero over here, just in here, you'll hear the cheer. I don't know how well you heard it, but that was it. So we saw this signal, you can see that the three devices, these are the two like ones, this is frequency versus time.
So it goes to higher frequency and time. That's his character characteristic shape. The signal is much weaker in the Virgo detector, but fortunately for us, it was located in the sky in a position that made them able to tell the position, help us tell the position very well. We also could tell where it was, what the distance was. And with the help of astronomical measurements, we could tell what galaxy it actually came from, which was important in the following measurements.
Two seconds later. A signal we're seeing by the Fermi satellite and the same region of the sky. They're looking at high energy photons coming from what are called gamma ray bursts. So what the gamma ray bursts were suspected or thought by many people to come from the merger of two neutron stars. So we immediately within two seconds out of 100 million light years, by the way.
So the first thing to mention is that the speed of light gravitational waves is the same to two seconds out of 100 million light years. The signal from the electromagnetic signal from the merger is supposed to be lighter because it happens not from the merger itself,
but from the nuclear physics after it emerges. So a couple of seconds later, they saw this signal that enabled us to alert the astronomical community, and something like 2000 out of 4000 instruments in the world pointed in the sky over the coming seconds, months, days. And that included gravitational waves. Of course, it included visible infra-red, radio waves, X-rays, gamma rays, even neutrinos. Oh, no, signal was saying with neutrinos. Then. This was done over some period of time.
So first, we saw that short signal 1.7 seconds later in the family satellite, which was presumably from a short gamma ray bursts. Within five hours or so, we were able to get those details sky location from another electromagnetic observation and so on out through the days and through the different bandwidth that people do astronomy.
I can't show all this, but it fits taking the measurements that happen over a period of time to a picture of a neutron star merger that happens in a phenomenological scan, which we call a kilonova. That kilonova predicts even how the wavelength will change with time. That fits pretty well with the light curves were that were detected, which I show here, and even with the emission is like as a function of wavelength, and all of this fits the picture quite well.
So I'm not going to go through that in detail. I'll show you one result from this one event, and that is that we as. Physicists try to describe everything in nature, always, whether we can or not. And so I grew up learning that. The heavy elements were a problem, how do we how do the heavy elements get into the Earth? We know that most of the universe is hydrogen and helium.
We know that stars burn using the fusion process, and that works up to about iron, and then it burns up the fuel and stars collapse, calling us making a supernova so that the elements up to something like iron. We understand where they came from and how they got into the Earth. But where did the heavier elements come from? That was always a puzzle. The explanation that we've been taught or believe is that there was an enhanced process in the burning of stars called the R process.
I won't talk about it in detail, which enhanced the production of heavy elements in the burning of stars. That was never very satisfactory, actually artificially making the table. How was that done? That's done generally by going into a laboratory, taking a heavy element, bombarding it with lots of neutrons, and you make a heavier one. And if you're successful, you make something that somebody didn't see, and that's the end of the plot.
So how did they get into nature? Into the Earth has been kind of a not a very well established and not a very satisfactory explanation. Well, the merchant way they're made artificially with neutrons bombarding some elements sounds a lot like what happens when neutron stars collide. So the phenomenology of what happens when neutron stars collide has been put together, and it's only as reliable as a single event, which is all that we saw.
So it's not statistically reliable at that point at this point, but it's consistent with all the your favourite heavy elements say platinum and gold being made by the collision of neutron stars. So if I look at the collision of this neutron star, presumably the ones in the Earth happened a long time ago, it actually produced about 100 hundred earth masses of gold. Now, we don't know for sure all of this is this one event, as I said, but that's the take up. So where are we going from here?
We basically now have three detectors working two in the US, one in Italy. The one in Italy, we hope, becomes as sensitive as legal instruments. But that'll take a while. We're also commissioning one. The Japanese are commissioning one in Japan, and that has some innovations in it too. That might eventually be the kinds of things that we do in the future, which is mostly to go from room temperature to a cold detector. And we're building one in India that'll be finished by twenty twenty five.
That'll enable us to do much better in localisation. So the first problem and doing multi messenger astronomy is we have to be able to localise what we see. This is where we are now and this is where we'll be in about twenty twenty four. That is, we got lucky being able to see where the event we saw now was because of where it happened in the sky. But we'll be able essentially anywhere in the sky, be able to tell to roughly 10 to 20 or 30 square degrees where an event came from.
So that's the first necessary ingredient to be able to compare with astronomy. There's lots of other sources besides the one that I've talked about today, and as we get better sensitivity, we'll have a better chance. Let me just list them a little bit. We can see if they're close enough and we don't know very well a supernova. So certainly a supernova that might happen in our own galaxy.
We can see in the most optimistic conditions where we improve the detectors we might see out as far as the Virgo cluster, where there's one two year type two supernova. And within that range, there are periodic sources note spinning neutron stars in our own galaxy. To the extent that they're not perfectly spherical, they'll have a quadrupole moment and they'll give a signal.
We've looked for those we've limited for, known Paul, the best known pulsars that they don't have mountains on them bigger than, say, a millimetre out of the 12 miles or so across. So that's small, but we haven't seen yet a signal. That signal will be a continuous signal, and we're not looking in the most sensitive way. If you expect to see a signal for gravitational waves, the neutron star should be not perfectly symmetrical.
The best ones known by radio telescopes are, by definition, the ones that have been around so long they give a very stable signal. They're therefore pretty Saracho. So what we need to do is to detect the ones that are young and not seen yet by radio telescopes. That's a tremendous computing and technical. Problem, but we're working on that. Lastly, and I'll mention that at the very end is to see cosmological signals.
So. We will be improving. My gosh, we have so far over the next decade, but we're still not limited by nature. So when people are impressed by how well we've done, we still can go further and we know that more for LEGO itself, roughly a factor of 10, which is a factor of a thousand. And right before we had some problems that are more fundamental that we will limit us. They're technical at this point. We then can move and we're starting to design.
What I show here is a European design of what a third generation detector might look like. Third generation detector, the one they talked about here, may not be what's built, but looking on the left. These are some of the parameters they're all likely going into a third generation detector, presumably in the twenty thirties, and the Einstein telescope has almost all those features. I show that first you might want to go underground.
Going underground gets rid of a lot of the seismic shaking of the Earth, especially at very low frequencies. So going underground, if it's important enough to go to yet more frequency, that's heavier a mass black holes. You go underground. The arms can be longer than what we have at the ideal length is actually 100 kilometres. So I'm sorry, 40 kilometres. But they are talking about making 10 kilometre arms. The present ones are three and Italy and four kilometres in the US.
You can have a configuration that, instead of being L-shaped, is triangular. That has some advantages in determining the two polarisations of gravitational waves. And also in being able to internally be able to crudely but be able to point somewhat. A big step for us and one that we're working on really the hardest is to get rid of the noise in the most sensitive region by cooling the detector becoming cryogenic.
That's a big challenge, but only a technical challenge. It means that we have to be able to get the heat out, which is made all the time when the laser beams going through without shaking it because everything has to be quiet and we have to add optics that will work at cold temperatures. So we have to develop the right materials and the right coatings to reflect the light. And we're working on all those problems and the these people are as well.
Lastly, we show up when I showed you was a broadband instrument. It looks at gravitational waves at all frequencies when in fact, if you want to, you can do better in different parts of the frequency band. So for example, in the case of the neutron star merger, we'd like to understand what happened to the nuclear physics after they merged, not just the general relativity that these two objects merge. And then a lot of light came in.
But what happened in the nuclear physics that happens if we to detect that at higher frequencies than we now are very good at? So you want to optimise things at higher frequency. So anyway, you'll see them increasing. Our proposal in the US is to make a maybe longer version of, like we call it, the cosmic explorer. It's to go instead of four kilometres on an arm, something like 40, which is pretty ideal in terms of the parameters that you're measuring.
And with that, and using the same technologies that we're developing over the next decade, but put into a new detector of longer length, we can see we can become an instrument that starts studying cosmology that looks at high redshift. Right now, we're not. And for example, for the black holes, we can see that the edge of the universe. And that's the future. Lastly, the whole game is not on the Earth, just like an astronomy.
And I started by talking about astronomy doing done in different wavelengths. You'd like to do a graph and we will do eventually gravitational waves and the different frequency bands. So everything I've talked about so far is on the Earth's surface, and I show that on the right side, that's legal.
And it's that is the audio band. There's a proposal and there's a approved space experiment called Lisa being built through the European Space Agency with small collaboration from the US that looks at not ten hertz to ten thousand hertz, but 10 to the minus one to 10 to the minus four hertz by having three satellites that bounce the beam back and forth the distance between them two and a half million kilometres.
And that does the region in a lower frequency. There's also ideas how to do the region and between the two, but there's no approved experiment now. It's certainly doable. And lastly, on the left. You see that I have something called PTA, that's what that is, is fine is pulse pulsar timing a pulsar timing array, meaning that we see these pulsars.
They're very good clocks. If you measure the whole array of them and a gravitational wave comes through, it changes the relative timing and you can pick out a low frequency gravitational wave. The ultimate goal this is the space experiment, 2.5 kilometres, 2.5 million kilometres apart on a triangle like the experiment below ground.
The Pulsar timing array to see things. And lastly, I'd say not immediate, but the dream that certainly I had for the moment I got into this is how do we really learn about the early universe? And I think the ultimate technique is gravitational waves, everything we know and it's a tremendous amount. We've learnt from the cosmic microwave background experiments, which only probe what happened up to four hundred thousand years after the Big Bang.
Before that was a fault. So I can go back before that. We project back, but we can't measure it back. If you want to go to earlier times and probe it, there's two choices. The two choices are neutrinos or gravitational waves. Neutrinos. They go back to a few seconds after the Big Bang. The problem is that neutrinos that were made at that point in time have thermals in between. They're very low energy and experimentally that makes them almost undetectable.
Gravitational waves aren't absorbed almost at all. So they're our ultimate probe, I think, to go back to the very first instance of the Big Bang. We don't yet have the detector to do that. But I told you, I took 400 years to get where we are from. Mark Galileo was. So if you give us 400 years or maybe less, maybe we can measure signals from the very earliest moments of the Big Bang and understand really how it all began. Thank you.