YUVAL GROSSMAN: Yes, it's on. Welcome to the Bethe lecture. And my name is Yuval Grossman. I am a professor of physics here, and I will say a few words about Hans Bethe and about our lecture today. But before I do so, I want to remind you or let you know that after the lecture, after the question and answer session, we're going to have some reception, and that's going to be in PFB 401.
I guess most of you know where it is. If you don't know where it is, just follow the crowd. We're all going to get there, OK? So join us.
So let me say a few words about Hans Bethe, and then I will move to today's lecture. So Hans Bethe was one the most original, productive, and influential physicists in the 20th century. He was born in Germany in 1906 and came to Cornell in 1935, where he actually stayed until he passed away in 2005.
He retired from being an active faculty there mid-1970s, and back then, after he retired, the university actually established this set of lectures to honor his contribution to physics, to Cornell, and in general to the world. He made seminal contributions to many areas in physics, and that includes nuclear physics, condensed metaphysics, particle physics, and astrophysics, and he actually received the Nobel prize in physics in 1967 for his work where he explained how stars actually burn hydrogen into helium and how they actually shine. OK?
And while unfortunately I never met Hans Bethe because I arrived to Cornell in only 2007, actually even coming here, I still kind of have his legacy around me. Every time, when things happen in the department-- and said, when Hans Bethe was here, he would solve this problem like this in a very friendly and nice way. And so we always try to keep this friendly atmosphere in the legacy of Hans Bethe. So he was not only a great physicist. I can see how his legacy-- not only a physicist here with us today.
And actually, it's not surprising that today's lecture, which is going to be about neutrinos and neutrinos from the cosmos, actually relates to some of Hans Bethe's works. Actually, I think almost every talk in physics somehow relates to what Hans Bethe did. So I was thinking that actually, it would be quite amazing for Hans Bethe if he was with us today. And he would have seen actually how much progress there had been in the area of what we call particle astrophysics.
We actually do astrophysics with particle physics-- what's happened in the last 10 years. And a lot of it that's been done is the experiment of IceCube that I was speaking on in a talk today. So actually, that brings me to the introduction of our speaker, Professor Francis Halzen. Professor Halzen is the Hilldale and Greogry Breit Distinguished Professor of Physics at the University of Wisconsin.
He received his PhD from the University of Louvain. I've tried saying it in French accent and in English accent and in Flemish accent. But probably it's not-- let me try the Hebrew accent. In Louvain, Belgium in 1969, he'd been in the physics faculty in Madison since 1971, and he started his career in particle theory. And some of you know his book, the Halzen and Martin book on particle physics that we are still using a lot this year.
Until one day he came to this extremely crazy idea that the best way to look at the northern sky is actually to go to the South Pole. OK? After the amazing, crazy idea, he actually decided to actually do this idea, and then he became the leader in the first experiment that actually looked for neutrinos coming from the northern sky in the South Pole, called AMANDA. And later on, after this experience was quite successful, they actually scaled it up to what is called now the IceCube experiment, where he is now the leader of these experiments.
The IceCube experiment produced many interesting physics results and received many awards. I will only mention the last one that they received. It's called the 2013 Physics World Breakthrough of the Year Award, which is supposed to be the most impressing result of the year in 2013. And that was given to them for the first observation of cosmic neutrinos. And I'm sure that in today's talk, he's going to say much more about this. So it's time for asking you to help me to welcome Francis Halzen.
FRANCIS HALZEN: Thank you for this very kind introduction. It's obviously a great honor for me to be here, and I must thank the department for allowing me-- it's not-- again-- ah. Yeah, I don't speak through my feet.
We had trouble with this. I know what I always do. I wear this tie, like, once every three years. So I'm going to tape it to my tie. OK. Let's see if that worked.
So yes, I really enjoyed my visit here intellectually, and I thank you for the warm reception I received here in the last few days. So coming to the topic of this talk, I'm going to talk about the IceCube adventure. And in case, to make sure not to disappoint you, this is a talk about Antarctica, but there are no beautiful pictures like this. So I show one to make up for it.
Frank Hurley was the doctor in the famous Shackleton expedition. And if you like photography, you really should look at these pictures sometime. There will be no penguins.
This is what it's about. This is the South Pole. Someone made this picture, actually, for me. And what's special about it for us is-- and that's why we're here today-- you stand here on three kilometers of crystal-clear, transparent ice, and that's what made this experiment possible.
That's a picture taken more remotely. And so you see this is the important ingredient for us at the South Pole. There is a station that is run by the National Science Foundation and that has cranes, bulldozers. You can do things, actually. This, for those of you in this business-- that's the South Pole telescope. And IceCube is a mile deep in the ice, but this is the computing center at the top of the ice.
Well, maybe I'll show you one nice picture. Well, I'll tell you that the great thing about this station-- there are many stations in Antarctica-- it operates the whole year round. And in fact, at this time, two people are stuck at the South Pole operating our experiment, and I should probably dedicate the talk to them.
So this is a long subject to introduce, but somehow none of it is really fairly difficult, so don't get scared. But we have to introduce a lot of concepts. What's a neutrino? Then I have to tell you about cosmic rays, which kind of is one of the leading motivations or was one of the leading motivations to do this experiment. I'll finally tell you about IceCube, and I'll show you the first neutrino maps of the sky. And that comes with the first surprise, which I will tell you about.
What's a neutrino? Well, you were told in school that the world is made out of protons and neutrons that make nuclei, and nuclei with electrons make atoms, and that's it. And that's true but not the whole story. In fact, a neutron and a proton occasionally can change into one another. For instance, a neutron can decay into a proton and an electron.
And so if I were a neutron and I decayed, the proton goes this way. The electron goes that way. That's momentum conservation. Energy conservation. And Shadwick realized, when he studied radioactivity of nuclei, where neutrons change into protons, that this was not true.
And in fact, there was an experiment done-- very beautiful experiment-- by Ellis and Mott in 1933, and they actually showed that when the electron does that way and the proton goes that way, some other particle must go that way. And they looked for this particle. They measured its energy. It was the missing energy, the energy that was missing to balance energy momentum conservation. And this is a standard technique these days to look for particle, this missing energy technique.
And so they discovered the electron. They even measured that it's mass was very small. But at the time, only the three other particles were known, and they couldn't imagine they had discovered another particle. And so it took until 1956 for the neutrino to be rediscovered.
And in fact, it had been proposed especially for the purpose of saving energy conservation by Pauli in 1930, even before this experiment. And he's famous for his quote, "I have done a terrible thing. I have invented a particle that cannot be detected." And when we were developing our project at the South Pole, I would often quote this in meetings for my experimental friends were present, and they didn't think that was funny, but-- they didn't laugh.
So we now know that this is not the full list. This is the full list. And without a neutrino, there wouldn't be any nuclear physics. In other words, changing neutrons into protons-- that's nuclear physics. The sun wouldn't burn without neutrinos. Stars wouldn't explode, and we wouldn't be here, because the universe would just be protons and helium and a little bit of something else. And so it's thanks to nuclear physics that we are here in the sunshine. So what can be more important?
So in fact, you find neutrinos-- neutrinos actually, among these four articles, are the most common one in the universe. And so they are wherever there is nuclear physics. I have mentioned exploding stars. I have mentioned the sun. The Big Bang produced neutrinos. There are about 200 in this room. Every cubic centimeter. We are sure they're there. If you know how to detect them, you go straight to Stockholm.
You are made of neutrinos because you have salt in your body that decays. And I have forgotten the flux, but it's big. And of course, we make neutrinos at accelerators, but for this talk, neutrinos are made when stars explode. Very important. That was actually Hans Bethe's favorite topic at the end of his life. And the only time I met him, we had the discussion about supernova.
But an important source of neutrinos in this talk is the atmosphere. There are particles called cosmic rays that hit the atmosphere. They come in from the cosmos. When they hit the atmosphere about 20 kilometers above your head, they interact with nitrogen and oxygen and then make pions, and these pions decay. The charged pions decay into neutrinos and muons.
And so this I will refer to as cosmic-- all this stuff stays in the atmosphere most of the time, but the neutrinos are not stopped by anything. So they come down to ground level as well as the muons. About 100 muons go through this desk every second. You're not aware of it. It's OK. You won't die from it, but they are there, as you will see in our experiment.
So actually, I want to go back, not to the neutrinos. The sky, because of the existence of these cosmic rays, glows in neutrinos, which is-- when you want to do astronomy, you realize. You see it coming. That's going to be a big problem.
In fact, IceCube sees a neutrino, as I'm standing here, every six minutes. And most of us wish we didn't. And so we'll come back to it, but the cosmic rays are actually as interesting as the neutrinos.
And so they were discovered. They are charged particles, mostly protons, some nuclei. They were discovered in 1912 by this man. So notice that the discovery's more than a century old. And he took up a measurement device in a balloon to 17,000 feet without oxygen but with a tie on.
And he measured-- he had two measuring devices, very simple measuring devices, and he noticed that the radiation became stronger when the balloon went up. The common idea, I guess, was that the radiation was some radioactivity of the Earth, from the Earth, but that's not what he saw. He actually saw the radiation increase, and he discovered cosmic rays coming in from the cosmos. We know, these days, that they come from galactic sources and extragalactic sources.
And sorry, let me go back. As you see in this slide, in 1991, an experiment in Utah, which was actually designed by Greisen, who was a professor here-- and the experiment was finally done successfully in Utah-- detected a cosmic ray that had this energy.
Now I have to introduce my units. A TeV is 10 to 12 electron volts. It's a tera-electronvolt. It's the energy of the Fermilab machine. So the protons coming at Fermilab have an energy of one TeV-- the accelerated proton. So that's what I'm going to use as my unit. But you see, this particle has an energy 300 million times that.
So if you wanted to build-- in fact, I'll make the analogy with the LHC, which is Large Hadron Collider. If you took the Large Hadron Collider's magnet and wanted to make a ring, an accelerator, and accelerated a particle through that energy, you would have to fill the orbit bit of Mercury. I stole that slide from someone, but I checked it this afternoon. It's correct.
Obviously, the universe doesn't line up magnets, and so the ideas of how these particles are accelerated to these incredible energies are very few. It's a challenge to think of something that accelerates these particles. But the basic ideas go back to Fermi in the 1950s.
And you see here, this is a solar flare. Nothing to do with the nuclear reactor I was talking about. This is a solar flare where, for a short time, the sun managed to accelerate particles. And we have studied these, and we begin to kind of understand the principles by which the sun does that. And you can calculate that the sun accelerates during a solar flare like that.
You have relativistic articles that move. If you're moving charged particles, you get magnetic fields. You have magnetic fields, you build magnets, in a sense, that accelerate. And so 10 GeV-- of course, that's 100 of my unit. But indeed, when you wait for a day or so, you see these protons arrive with that energy.
Now it's so difficult to imagine how to accelerate these particles that there is basically only one good idea, and that is these particles are accelerated when stars collapse. And when they collapse, like this star-- this is a star that collapsed 300 years ago. You see the debris of the star spread out in a galaxy. And you see this filament, a high energy particle, just like the ones-- the sun. So it's similar, but now you have bigger magnetic fields. You have to move over big distances. And so the theory is that this are the accelerators of the cosmic rays inside our galaxy.
Now if this star had collapsed to a black hole, it would have done this. And it's exactly the same thing, but it happens in a few seconds instead of in 300 or more years. By the way, it only looks different. You saw a simulation. It only looks different. You see this alignment. That's the spin axis of the black hole that was formed by the collapsing star. But it's the same mechanism. So that's the theory.
So the problem, however, is that there's absolutely no evidence for this. And so one of the motivations of IceCube was to find-- and I remember giving this talk. And I say, we'll discover the sources of the cosmic rays before the 100th anniversary of that discovery. It's a great line, but we couldn't deliver. We are a few years late.
So you may say, what have neutrinos to do with cosmic rays? Well, cosmic rays-- they don't tell you where they come from because they're charged particles, and before you detect them, their trajectories have been bent by the magnetic field of the galaxy. Or even if they come from outside the galaxy, by extragalactic magnetic fields. So they can be produced there, but actually you detect them coming from there. So that's why this problem is unsolved.
But as this slide explains-- so you notice the accelerator is a collapsed object. So when a star collapses, you produce a large amount-- you release a large amount of gravitational energy. And so if you can convert a few percent of it in acceleration, you explain the cosmic rays. And so that's the standard model.
But these accelerators-- so the black hole is the accelerator. But these objects are surrounded by radiation and other stuff, but mostly radiation. And so that means that these particles, when they accelerate, can interact with the light, the photons that surround the black hole of the neutron star. And that's what we particle physicists refer to as a beam dump.
This is how you make neutrino beams at Femilab. You shoot the beam in a target, a big target that absorbs everything except neutrinos that go through everything and come out at the other end. So it's the same here. Cosmic accelerators make neutrinos. And so the neutrinos-- they have no electric charge, so they come straight from the accelerator to you. And if you detect the neutrinos, you can tell what the sources are.
So that's the idea of-- that was kind of the slogan on which we build our case to build IceCube. But neutrino astronomy existed long before us. And a little bit about history. You know, this is a cliche, but I brought this slide back because it's still the best way of introducing neutrinos in this context.
It's a John Updike poem. It says neutrinos-- they are very small, whatever that means. But there is no charge. So that's the key point. He immediately brings in the key point. They have no mass. This is not true, but in this talk, they have no mass. Their masses are so small we don't care. And they do not interact at all.
So what you read from this is that neutrinos are just like photons, like the particle or light. No charge, no mass. Moving at the speed of light. And so they are like a photon. You can do a astronomy with them.
And so by the way, you only shown the first page of this poem. The second page you should go and read, and then you will see why it's usually not shown in a public lecture.
So they are electrically neutral, essentially massless, and essentially unabsorbed. So the difference is that light doesn't go through walls. Neutrinos go through walls, and they go through everything. And of course, they only are created where nuclear physics is happening, like in particle accelerators, and so that's what we are interested with here. But of course, they are difficult to detect because they also go through your detector, no matter how big it is.
But already the neutrino was rediscovered in 1956 by Reines, and he told me that everyone the same day had the idea that you should do neutrino astronomy. And in the 1960s, several people, among whom Greisen, from this university, wrote very beautiful papers introducing the topic-- but it's simple. This is a picture of a star that exploded 1054, and this is that same picture when you look at you observe it in wavelengths-- in different colors of light, different wavelengths.
So for instance, this is infrared. That's the picture with the Hubble telescope, what it looks like in radio. This is an x-ray picture. Blue, if you want. And so in this picture, actually, it was taken by a satellite not so long ago, and you see this particle B. That had never been seen before until an x-ray picture of this object was taken. Also, usually you see sources-- not only do sources look different, but you see things in the sky you had never seen before.
So everybody understands the argument that we have exploited all the wavelengths of light. Let's try neutrinos. And nobody ever questions that we should do this or try to do this.
And so the other motivation, when we built this instrument, we very much built it as a discovery instrument. It's as big as possible and as good as possible for the money they would give us. And so this is the kind of thing-- this table is actually taken from some textbook. The history is horrible, but I just want to make a point.
And I understand the first point. I grew up near the sea in Flanders. And the first telescopes were built in the area of Bruges and in London, and they were used when the ships arrived from London. And you could first see what was on the ship. You could sell it.
And there's no evidence that any astronomy was done with it, but Galileo found out that there was these two lenses. And he knew optics, and he immediately built an instrument that was thirty times bigger than the so-called spyglasses you could buy in Flanders and in England.
And I showed you this picture of a star collapsing to a black hole. In the astronomy trade, it's called a gamma ray burst. That was actually discovered by two military satellites that were looking-- built in Los Alamos-- looking for nuclear explosions. And they never saw one, but they would calibrate their satellite by looking at Vela.
And then they saw this explosion in the sky, about one a day, that they couldn't detect in the Soviet Union, and that was a gamma ray burst. And in fact, it was a military secret for years. We didn't know they had discovered this, but eventually-- they were the ones, actually, that discovered gamma ray bursts.
X-ray astronomy-- I actually read a proposal written by Rossi. He wanted to see the moon. In fact, the moon wasn't seen for decades, but he discovered the source in the sky that was so big-- it was Sco X-1-- that the referee of Physical Review Letters turned it down because it was impossible. And so he wrote a letter back to Physical Review Letters-- this observation is correct. And it was published. Can you imagine me writing this? So the point of this is that you may not discover what you're looking for, but what you discover is more likely to be much more interesting.
So this is a quote that was given to me, of whole people, by someone from the Office of Management and Budget when I was talking to them. And he said, you cannot show that slide. He said, because it looks like you are taking the taxpayer's money and gambling it, which is kind of what we were doing, right? And I said, what do you propose? I was kind of irritated. And this is what he quoted to me. He was very useful to us later, as well.
So we're going to build a neutrino telescope. You cannot detect neutrinos with mirrors. It's a particle. You must build a particle detector. And how to do this? It was actually known since the '60s, as well-- late '60s-- that you better built something very big. That's what the early pioneers didn't realize.
And the word "kilometer cubed neutrino detector" is around since 1969. I had papers, already in the '70s, about this. And so how to handle this was an idea of this man, Markov. And this is the idea.
So first of all, let's make clear you cannot see neutrinos. There is no electric charge. So they just go through your detector. The only way you can detect them is-- also, they don't see atoms. That's why they go through walls. So they don't see an atom. So it's only when they crash into the nucleus of an atom, which is very small, that they will occasionally make a nuclear reaction.
And when they make a nuclear reaction, they emit blue light. You saw on one of my slides-- I should have pointed this out, but you know that-- you see a nuclear reactor. It's covered with water to absorb the radiation. That water is blue. That blue light comes from the particles coming out of the nuclear reactions, the charged particles.
So when a neutrino crashes into a nucleus, it makes a spray of particles. They are charged and emit blue light, just like in the reactor picture. And we detect that blue light with these objects, which are photomultipliers. Basically, they are light sensors. And from the shape of the light, we can reconstruct the direction and the energy of the neutrino.
And so we love these neutrinos that are called muon neutrinos, and they make a muon. And these muons can travel through the ice for a kilometer or more, and then you can see them from far away. And when it travels through the ice, it makes a Cherenkov cone because it goes at the speed of light, but the light goes at the speed of light in ice, which is not the speed of light in vacuum.
So if you don't get the physics, just two pictures. So this is the blue light, and this is what we call the Cherenkov cone. Look at this duck. It goes faster than the speed of the waves it creates. And so it creates a Cherenkov cone. And so if you measure this cone, you know which way the duck is going without looking at the duck. That's what we are doing with neutrinos.
And then finally, the easy part of the experiment. These things here are light sensors, and you just buy them in Japan. They cost slightly over $1,000, and they work forever.
So this was tried, a heroic experiment from which we learned a lot. They developed a lot of the techniques we still use, and they made mistakes that we then knew how not to make. And they are friends of the DUMAND experiment.
So they tried to put photomultipliers 4 and 1/2 kilometers deep off the coast of the main island of Hawaii. That experiment was in progress and development for about two decades, and it failed. They deployed a string, a buoyant string of photomultipliers, and it worked for 24 seconds.
And so that's the time that we started working on the idea of changing water into ice. And so you know, why not-- if it's difficult to put them in water, it may be easier to build them into ice, which is what we finally ended up doing. So you go to the South Pole, and we had this great luck that there was this research support that made this experiment possible.
By the way, this is the IceCube project. This is the station, which I showed a picture earlier. And this is where the plane lands. At the time of IceCube, everything went in and out of the South Pole in planes.
Now it's amazing to think of how this project started. I actually started it with two graduate students from Berkeley. In fact, they started it and forced me to go to NSF. And this was our group in Madison-- a scientist, a postdoc, and a student.
In fact, if you think this student looks unhappy, he transferred from Illinois to work with me on something called QCD, and I said, well, we are going to build optical modules instead. And that's what he did the rest of his life. He's now building a detector in [INAUDIBLE]. He's at the University of Valencia. [INAUDIBLE] is a professor at the University of Delaware but mostly is happily retired in Hawaii.
But you see, these are the optical modules, and they worked until we turned them off many, many years later. This was an incredible-- in fact, the room-- I had no lab, but the theory group had a table tennis room. And so that was the table tennis room, and they still hate me for removing the table.
So we built this experiment, this small experiment. It's actually a 1% prototype of IceCube, if you want, of 700 photomultipliers. And if you go to the top, you see here? That was the lab, the counting house.
At the time, the experiment was completed 2000. This is what the South Pole looked like for those of you-- this is the dome, and this is the geographic South Pole. This is about 800 meters away.
And we detected neutrinos. This is a neutrino. You see, the yellow and the blue line are fit by the computer. These dots are photomultipliers that detect light, and you see the light is detected on the track that emits the photons. And you can bet your house that this is a neutrino. The muon comes through your feet, and nothing but neutrinos can bring muons to under the detector.
And so then what you do is-- that's a well known technique. You publish in Nature, which we normally never would think of. Make extravagant claims and ask for money. And by the way, we got it instantly.
This is Scientific American. I tried to make the font so that you couldn't read it. And so we were declared one of the seven wonders of astronomy in the category of the weirdest telescope. That's the Hubble telescope, by the way. But we're on the same page.
You would think now, after the failure in Hawaii and the success at South Pole, that it would be easy. I mean, we always knew-- in fact, if you take the original AMANDA, the first AMANDA talk I gave, it says that we wanted to build a kilometer cubed detector. And so you thought that now it would be easy.
I mean, we have proven. And it was quite a shock to us that people were doubtful. They couldn't accept that putting photomultipliers a mile deep into South Pole in the ice was the easiest way to build a neutrino experiment. But it is. It has been shown over and over again.
I come back. The people who do this in water are not totally unsuccessful, but they are years behind, and it obviously is harder. And this was another print that-- someone, an administrator at Wisconsin, knew I was struggling selling this idea for the ultimate detector and knew that I was a fly fisherman. And he said, you have found the hardest possible way to do ice fishing. And so that let to the title of this talk.
So to make a long story short, this is IceCube. So you go 1.5 kilometers deep, and at that point, there are no bubbles in the ice. The ice is perfectly clear. And so you just think of a kilometer cubed of ice in which we put 5,000 tons of 64 photomultipliers. That's it.
And in fact, you say, is this the straightforward or-- we spent two years thinking of intelligent ways of building this detector, and the only thing that came out of it is that it's hexagonal instead of, you know, a square. It doesn't matter anyways. It has to do with corner clippers, if you know what this is.
So here is the two-minute movie on how you put photomultipliers in ice. The top 80 meters is snow, and you just melt it. Then comes in this thing that you call the hot water drill, and it puts out water, hot water under pressure, and melts its way to the bottom. And if you wait two days, it's 2 and 1/2 kilometers deep, and you have a hole this big. It's not a hole. It's water instead of ice.
And so what it takes to do this-- so you need a five-megawatt heating plant that puts out 200 gallons per minute, under pressure, of boiling water. And so this is what the setup looks like. So it's like a circus where every element of the hot water drill is built on sleds. And so you move it to where you want to drill.
This is the hose that brings down the hot water. It's 2 and 1/2 kilometer long. It's about that big. It was built in Italy. And this is the power plant. The power plant is 40-plus car wash heaters driven by normal generators that operate on fuel. And so you see, everything is on these sleds, which are snowed under under at the moment.
And this is the moment when the hot water drill comes out. So you see, there's nothing to it. You will see at the end there is a nozzle. This nozzle, actually-- shape is critical. It was Jeff Cherwinka who figured this out. There's a lot of ingenuity in this drill which is not transferred by this movie.
But then, when the drill is out, this water acts as an insulator, so this water stays liquid for quite some time. And here are the optical sensors waiting. So these are photomultipliers, which you then attach to a cable. This cable brings the high voltage, and you put on a photomultiplier every 17 meters until you have instrumented a kilometer cable. And so this is the last one.
Now the whole thing sinks to 2 and 1/2 kilometers, and that's it. So that's the last time. It's like launching a satellite. You see your optical sensors.
So if you could go inside the detector, it would look like this. So this is a kilometer-long string of optical sensors. This is 2 and 1/2 kilometers. If you look 125 meters away, there would be another string. And so if you do this 86 times, you have IceCube. That simple. You fill the kilometer cube.
So here is the photomultiplier. This is actually how it's deployed. If you look, I'm going to open it for you. So you see here this light bulb, which doesn't emit light but detects it.
And here. What is this? This is basically a computer where you capture the signal and you digitize it. And so the whole experiment is about collecting digital signals from these 5,160 photomultipliers. And of course, that you do by each of these strings.
This was, for instance, a season where the crew had deployed 20 strings. So each of these cables go into this two-story building, and if you look inside the building, there's nothing. There are computers. And so these computers collect the signals and put them together in light patters from which you determine the energy and the direction of the neutrinos. And that's the telescope.
So here is a movie of this. This is a muon going through the detector. And you see, the blue is the Cherenkov light, and the light lights up these modules. So you see, we have more money now. The event display looks much more impressive than before.
But that's the idea. If the muon comes up, which happens one in about a million times, then it's a neutrino. So the gain of this experiment is to reject these down-going muons, measure them in real time, and reject them about at the billion and one level so that they don't represent the backhand. And so that-- you have to understand the ice very well, but that's a different talk.
Yeah. So I will show you what this looks like in real life. Remember, this is what you detect-- neutrinos that are coming through the Earth, muons that come from the atmosphere above you, 100 per second through this table. Now if you go a mile deep, some get absorbed, but the detector is a kilometer squared, not the size of this table.
And so you can see the detector at work. So we see muons coming from the back and then neutrinos coming from all over the Earth, including the sky above Cornell. And so this is what data taking looks like. You see all these signals arriving from these photomultipliers, and then you see the computers making tracks, fitting tracks to it. And occasionally, if you wait, you will see a big bundle of cosmic ray muons go through here.
You see, this movie is 10 milliseconds long. So the outcome is that we detect 100 billion muons per year which we have to reconstruct to make sure they are coming down and not up. We detect about 100,000 neutrinos, and we are looking, of course-- neither of these things we are looking for. This is all backhand. We are looking for neutrinos coming from the cosmos.
And so building this detector, I probably don't have-- [INAUDIBLE] give a better talk about the construction than I did. But it was really exciting, sometimes and many times much too exciting. And so we had no idea, when the detector was finished in 2011, end of 2010-- and we suddenly realized, we now better discover something. And I remember-- I mean, it never bothered us. Were so engaged. I had never thought. It came as a mystery to me, suddenly, that you had to discover something.
And at that time, The New Scientist, who's a respectable scientific magazine, put up a website where you could bet on the probability that big projects would discover something. And you see, it's interesting. ATLAS here-- 6 to 1 for discovering the Higgs. LIGO, 500 to 1 for discovering gravitational waves. AMANDA was also 6 to 1. It's off the screen here.
And you know, it suddenly hit me, because 6 to 1-- you may think are good odds, but that means five chances out of six, I would be teaching 15 courses a week for the rest of my life, right? And you know, these odds-- it's not like they were very generous. Look at how they calibrated them. Finding the Loch Ness monster, 66 to 1. And finding Elvis alive, 100 to 1.
So you know, truly, I didn't think it was funny, actually. It was painful. But OK, it worked out, or I wouldn't be standing here.
I have to tell you, we have these weekly phone calls. The collaboration. I'll show you the institutions later. We are a group of about 300 people. And once a week, you get on the phone. Everybody's on the phone describing the important developments of the week, and this is incredibly boring. And it goes on for hours. And suddenly, someone showed this picture. And so I'll let you look at it.
Now, I knew what this was, but I also knew I'd never seen anything like this before because the amount of light represents the energy of the neutrino. And so if you look at this neutrino and you know where it is in the detector, we can immediately tell-- this was, remember, the 1 TeV unit. This was a neutrino that had 1,000 TeV energy, which we can actually measure with a 10% error. And you cannot produce that in the atmosphere.
In fact, in the sample of events we were looking at here, there was another event like that. And so as I said, I knew what this was. Here is a simulation of the event which we had done to develop the detector, actually. And you see, it's like someone turns on a light bulb in your detector.
But remember, color is the time. So you see, where the thing is uniform, while it's spreading out, the photons on one side reached earlier to the photomultipliers and the photons on the other side, and that's how you can tell the energy of this neutrino. You see, these arrived first, these later. So this neutrino came in that way.
And so what it is-- it's a electron neutrino. And I have to tell you, electron-- there are three flavors of neutrinos. Electron, muons, and tau. And this is an electron neutrino which gives its energy in tracks.
But a nuclear reaction is very simple. It gives 80% of its energy to an electron, and that electron makes an electromagnetic shower in the ice which is five meters long, from here to the wall. Less, actually. And that's a point source of light. So it's like you switch on a light bulb in a kilometer cubed detector, which is what you saw.
Also notice these events-- this is the shoreline of the Madison campus, which, if you haven't been there, you should come sometime. And that's this event superimposed. So we have information. There are 100,000 photons in this event. We know to two nanoseconds where they are, which is about like that, and the size of the event is like five city blocks in Ithaca.
And so you can do this measurement very well, so we were rather confident that we were not goofing up here. And so we published this. And notice, if these events had been made in the atmosphere, it should come with muons that came along was the neutrinos. These events are totally isolated, and so we didn't claim any discovery. So I already said that.
The idea we had is just to take this two-year sample again and just look for events starting in the detector. And so you look for a neutrino that starts here, has a nuclear reaction in the detector. You reconstruct it and then look back and saw that no light was coming in.
And so we ran the whole two years through this system, and we found another 26 events. I'll show you all of them. I cannot resist. Here is the third. There is the second.
And then we published in Science. You know why now. You make extravagant claims and go and eventually ask for more money. In our case, it was more to operate the detector, but we will ask for money, of course. And we got immediately-- in fact, we published a paper in December of 2013, and within a few weeks, we had this award which you've already mentioned.
And people tell me, you must be incredibly excited. And in fact, I remember sitting in this press conference, and the only way-- of course, questions I had answered so many times. It doesn't involve any intelligence. You know, I was all the time thinking, if something is wrong here, how are we ever going to take this back?
And you don't get excited. I mean, for one year or two you don't sleep. In fact, I spent most of the last year making sure that some exotic backhand that could be there actually can be ruled out. So we're still in that mode, but I don't think there's anyone in the experiment who doubts we have seen cosmic neutrinos.
So by the way, remember half an hour ago? I told you this was the way we were going to discover cosmic neutrinos. Well, as usual, somewhere in some office, some graduate student is sitting doing this analysis, and nobody talks to him because we are all excited about these events starting in the detector. And he also was analyzing these two years of data, and he came up with exactly the same signal.
And in fact, I show you the highest energy events we have seen, 10,000 PeV. It comes from below the horizon, barely, and, you see, goes through the detector. So the energy of this is typically-- the highest energy neutrinos we have produced at a lab are much less than 1,000th the energy of these events. So you can imagine there's particle physics in these as well.
OK. Back to astronomy. So this was the first map of the neutrino sky made by AMANDA in the year 2000, when I showed you the picture of the pole. And you see, if you show this to astronomers, they immediately point to this thing.
So you realize what is this. You're at the South Pole. You look at the northern hemisphere, and so it's like a pixon map, and every pixon-- pixel-- is the rival direction of a neutrino we detected. This, of course, was detected-- that was too small. We knew were atmospheric neutrinos. And this here is a statistical fluctuation. In fact, it's 3.6 sigma, I think. But in a map like this, you have to see a 3.6 sigma effect, and fortunately, we knew that. Nobody got excited.
So this is the map of the events I showed you one by one before. So this was two years of data, and the first event we measured was this one, and it was within-- so it takes about a day of computing time-- I mean, a cluster, a big cluster-- to reconstruct this event. And it was within a degree of the center of our galaxy.
So we thought, ah, we already know all the answers. In fact, we now think this is accidental. We think this map is so totally isotropic, and in fact, this cluster is close to the center of the galaxy. This is, in this projection, the plane of our own galaxy, and there is a weak correlation.
But we doubled the data. In fact, we tripled the data, but only half of it is unblinded, and so you see the evidence. These things don't grow. They stay at the same statistical level.
So I added one year. I added another. So this is a totally uniform flux, which means it comes from cosmic ray accelerators outside, distributed through the galaxy.
But we were supposed to know what we were going to find, right? So, supernova remnants for the galactic cosmic rays, which we don't seem to see. Of course, they must be there at some subdominant level, and I bet we'll see that eventually.
And so this was the explanation. Gamma ray bursts. But that, actually, is fairly easy to prove, because the way we do that is astronomers see a gamma ray burst, they can tell you when it happened, and they can tell you where it happened. So you look at one direction in the sky for 10 or 100 seconds and see if a neutrino arrives. Then you don't build up this enormous backhand of down-going muons.
In fact, we can look up and down that gamma ray burst, and so we have looked at more than 1,000 gamma ray bursts and never seen a neutrino. And so in fact, the upper limit on the flux of gamma ray bursts is less than 1% of the flux we see, and so it's not gamma ray bursts.
I want to make a point that this is the main result. It is a real surprise of this experiment. It takes a bit of math to prove it, but-- you build a detector on which you spend 20 years of your life. You see a handful of events. You must be looking at something very esoteric in the sky, right? Some really weird object or-- not at all.
If you take into account that for every event we see, there are more than a million that's we don't see, this is an enormous flux. And in fact, the flux in neutrinos we see at high energy above 100 GeV or so is actually the same flux that astronomers are seeing in photons, raising the possibility that we are actually looking at all the same sources. We see the whole of astronomy, not something exotic.
And so these fluxes are exactly the same level. Now, the astronomers, however know what they're seeing, unlike us, so you just go and ask them, and what they see is this. And I must say, to be fair, many theories had speculated that this was a source of the high energy cosmic rays.
So this is a galaxy not like ours. It's one that has an active and a supermassive black hole in it. We have a six million solar mass black hole that's not active. And what it does, it-- remember the critical thing was to set up flows of charged particles so that you could construct acceleration by shock waves or reconnection? And so in the inflows and the outflows of this black hole, there is the opportunity to accelerate particles to this astronomical energy I mentioned before, that you had to build with an accelerator filling the Mercury orbit.
And so the best bet now, if you want to start betting, is that this is the answer. This is far from proven. There are a lot of papers written on this, but clearly we need more events.
And so we can get more events in two ways. We can run this detector for 20 years or make something bigger. Either way, we'll find the answer. Instead of getting 20, 30 events per year, you have to get to 300 per year to go by a factor of 10. But we will prove this eventually. And of course, the simplest way is to take the muon neutrino data, which we can reconstruct fairly well, and make correlations with the direction.
By the way, at this high energy, the number of photons astronomers see are very small, too. So we are comparing data now with the photons arriving at the Fermi satellite, and these are similar data samples that we have in neutrinos and gamma rays. Of course, it's not because you see the same sky that you have to see exactly the same sources because the neutrinos you collect from everywhere. The photons have a horizon. But eventually, we'll figure this out.
Now how do you build a 10 times bigger detector? You say from after this talk, that's not possible. Well, it is possible because one of the things about AMANDA is that these photomultipliers were spaced by something like from here to the wall, and we understood the optics of the ice over this distance, and that's how we could analyze this data.
And in fact, what we figured in IceCube-- we saw a light traveling over hundreds of meters. We had never seen this before. In fact, the absorption length of blue light, the light, actually, that we collect, is more than 100 meters in the bottom half of the detector and more than 200 meters in the-- no. The other way around. More than 100 in the top half and more than 200 in the bottom half.
And you see these excursions. That's because it reminds you this detector is pretty complicated. It's frozen and compacted snow that fell on Antarctica 100,000 years ago, and each year is a layer about this big, and so you have to take this into account. And so there are better layers or worse layers optically, and the reason-- it all has to do with how much dust there is in the ice, which has to do with the climate in South America. And you can actually see these same excursions and map them on top of each other with glaciologists that measure icicles.
And so we understand this ice very well. That's to remind you that distilled water has an absorption rate of eight meters, and the cleanest water-- that's in the Super-Kamiokande detector for the purpose of doing neutrino detection-- has an absorption rate of between 60 and 80 meters depending how old it is. And so you have to-- it means, actually, in the end, to make a long story short, we could have put our strings farther apart.
And so we have now-- oh. One more slide that I'm very proud of. Of course, the dust is important, as you heard from my story. We have to know the amount of dust, and we have measured this to the point that we occasionally beat the people who are analyzing icicles.
For instance, this peak here is less than a centimeter thick and is the dust deposited by one volcano 74,000 years ago. And it was never seen. It was seen in the geological record, had never been seen in the icicles. So we settled this business. I would say, not that we really care, but we care. We publish these things.
So we are developing this detector, replacing 20-year-old technologies by now, to basically-- instead of spacing our strings by under 25 meters by 250 or more. And then, with the same number of photomultipliers, you can build a detector 10 times the full view. So it's like on the market in Europe, right? 10 times for the same price. How can you resist? So that's it.
What's the outlook of this? We, of course, want to capitalize on this discovery. By the way, if you collect 10 times more events, you detect multiple neutrinos from the same source, and then you are doing astronomy with neutrinos only. And the astronomy is guaranteed, because don't forget the big surprise was how huge this flux is.
And anyway, I'm sure we'll discover something else. Neutrinos are never boring. Inspired by this, which is good news, actually, people continue to try to develop the technique to do this experiment in water, where the DUMAND project failed. And as we speak, actually, a detector that's trying to reach to deploy, eventually, strings at a level of IceCube a few weeks ago deployed their first two strings in the Mediterranean. And there is also an effort that's ongoing in Lake Baikal to instrument the bottom of the lake with a telescope like this.
I am three seconds over time. I apologize. So this is a picture, actually, by a famous artist, [INAUDIBLE], who I think works for Atlantic magazine. I like very much. Thank you very much.
YUVAL GROSSMAN: Questions? Yes.
SPEAKER 1: I have three very short questions. First is-- you said that its light detectors essentially last forever, and I noticed you have some electronics on it. So the question is, have you had any failures? The second question is related to that. How long do you expect IceCube to continue running? And the third question is, how long have you personally spent at the South Pole?
FRANCIS HALZEN: The last one is the easiest. I've never been there. I wouldn't be useful, and they wouldn't let me go. The second one-- the failure rate, as far as we know, is not the photomultipliers. It's the electronics. And it has been built to last for 20 years.
The first question-- we haven't lost a single module in the last three years, and since we deployed them, we have lost less than 10. It's some number like five or six. But it had to do with failures of the power supply to the experiment.
But the experiment has been stable. It runs 100% percent of the time. We take data. 99.87% percent of the time. And in the last two years, no module has failed. So of course, there's always the risk that some component on the data board goes wrong and it repeats in other modules. You're not at Fermilab or at CERN. This is still risky physics.
YUVAL GROSSMAN: More questions? Yes.
SPEAKER 2: In the movie, you showed the muons that go through the ice, and it looks really nice. I was wondering, what if there are many particles that go through? Then it wouldn't be making a nice cone like that. You would have to somehow do data analysis to solve that problem. Is that a problem?
FRANCIS HALZEN: Are you asking, are all these muon tracks, are the data clear enough so that you know exactly that these reconstructions are correct?
SPEAKER 2: Is it one muon event or many?
YUVAL GROSSMAN: If you had [? a pipe. ?] If you have two muons at coming at the same time, can you actually--
FRANCIS HALZEN: Oh. Well, in the movie, you saw a bundle go through that are probably 100 muons that come through at the same time. That's easy. You throw that away immediately. It's not a neutrino.
But it's not like you collect the data. The high energy physicists are very familiar with this. We have to do the same thing they do, right? You look for the Higgs. One in 10 to the 9 events or so was interesting. That's the same ratio for us.
And the computers at the South Pole don't do that for you. You collect the data. They do a first reconstruction. Then they are sent by satellite to Madison, where we put cuts on the data and do more sophisticated reconstructions to finally come down to a sample where you can even begin looking for up-going neutrinos.
So this is a very long process which I could show you slides of, but you would immediately run away if I showed them to you. Yeah. It's not that simple, but it's like any particle physics experiment. You have to do your homework, and succeeding means that you find that one event in 10 to the 9 or so.
YUVAL GROSSMAN: More questions? Yes?
SPEAKER 3: How does the ice froze-- it expanded. How did you compensate for the crushing effects of the ice or for the motion of the detectors?
FRANCIS HALZEN: Well, that's-- in fact, there is another part to your question. The way the ice refreezes, it's cold at the top and warm at the bottom. The bottom of the bedrock, which is about 400 meters from the lowest modules, is at zero degrees C, approximately. There are lakes there. And so the hole freezes from the top, and so by the end of the process, there are these pockets of water where there are enormous overpressures.
And for instance, these pressure vessels can take 10,000 PSI, but for instance, when the cable freezes in first and then the module is sitting in one of these liquid bubbles and gets pushed down, you lose the module. There's nothing you can do. And as [? Dawn ?] knows very well, we predicted we would lose one in 100, and we did. Slightly less. There's nothing you can do about it. Yeah.
YUVAL GROSSMAN: More questions? Note, before we thank the the host speaker again-- remind you that we go and have our reception. And we hope to see you there at that thing, Francis, again.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
The IceCube project at the South Pole has melted eighty-six holes over 1.5 miles deep in the Antarctic icecap for use as astronomical observatories. The project recently discovered a flux of neutrinos reaching us from the cosmos, with energies more than a million times those of the neutrinos produced at accelerator laboratories. These neutrinos are astronomical messengers from some of the most violent processes in the universe--giant black holes gobbling up stars in the heart of quasars and gamma-ray bursts, the biggest explosions since the Big Bang.
Francis Halzen, Gregory Breit Professor and Hilldale Professor of Physics at UW-Madison and the principal investigator of IceCube, discusses the IceCube telescope and highlights its first scientific results, March 23, 2016 as part of the Department of Physics' Bethe Lecture Series.