Robert W. Wilson (2015) - Cosmic Microwave Background Radiation and its Role in Cosmology

Prior to the 20th century cosmology was really the study of objects in the universe, not the universe as a whole, or the physics of the universe as a whole. Then in 1915 Einstein published general relativity which established a theoretical framework for understanding the universe as a whole. There was one cosmological fact, the night sky is dark. However, Einstein understood that the universe was static, and he introduced the cosmological constant to make general relativity in some sense capable of having a static universe. He was ignoring the fact that the night sky is dark because in such a universe wherever you look there will eventually be a star in that direction, if it's infinite. The night sky would be very bright. As we'll see it turned out not to be the case that the universe is static. People have claimed that Einstein called that his greatest blunder. It's not clear that he actually did. In 1925, using Cepheid variables, Hubble showed that the Andromeda nebula and other similar nebulae or nebulous objects were not part of the Milky Way galaxy, but were actually at a much larger distance. This greatly expanded our idea of the size of the universe. Hubble went on measuring these objects, looking at the frequencies of lines in them and gathering data. Meanwhile, in 1927, Lemaître proposed a theory of an expanding universe based purely on general relativity and not on observations. The first Hubble diagram looked sort of like a standard astronomical scatterplot, except that he drew a straight line through it because the expectation was the universe should be the same anywhere. And so the expansion should be a linear thing and it ought to be a straight line. Well, there are other problems here. Hubble didn't sample a very large region and the Hubble constant he developed was 5 times too large, so that the universe was shown to be younger than geologists understood the earth is - which is not a very satisfactory situation. (Laughter) Meanwhile Bondi, Gold, and Hoyle developed the steady state theory which actually would allow such a situation because the scale of the universe is simply an exponential expansion. And you could actually have objects older than the scale of the universe in that case. As time went on Vera Rubin and Fritz Zwicky, noticing the motions of stars in galaxies and galaxies in clusters, saw that the velocities were high. You know, we can tell the mass of the Sun by looking at the orbit of the Earth. So you can tell the mass of the galaxy by looking at the orbits of the stars. They saw that there had to be much more mass than one could understand from the starlight we were seeing. Let's see how do I make this thing work? That was where the idea of dark matter showed up. Now, switching subjects a little bit, John Pierce was a polymath at Bell Labs. He wrote the book on electron beams, he contributed to communication theory with Shannon, he started probably the first computer music group and programme. And he wrote science fiction under the pseudonym J.J. Coupling. I think it was probably this latter activity which caused him to write a paper on the possibility of having ... Let's see I'm having a little trouble with this thing. Of having ... I've got something I have to setup. Of having communication satellites. This isn't working the way it does on my Mac. Anyway, in 1957, when Sputnik went up, he was immediately interested. And in 1958 when NASA proposed to put up the Echo balloon, Bell Labs jumped on it and said, "We want to use that as a first communication satellite." It would have a very weak return signal, because a wave hitting the sphere will be distributed in all directions. However, they could get enough return to actually have at least one voice channel. So Bell Labs proposed using Echo for that just to get into the business of making Earth stations and so forth. Since the signal was going to be weak, they decided to use two Bell Labs inventions to have a very sensitive receiver system. A ruby traveling-wave maser which was the lowest noise amplifier, cooled with liquid helium only about 4 Kelvin of noise temperature. Since, if they use that with a regular parabolic antenna, there would be enough pick-up from the ground around to have much more noise. They would build a very large horn-reflector antenna which would, as you can see from the shape of this, the receiver which is in the cabin there, I don't know if this how ... Anyway, is very well shielded from the ground. Both models in measurements and calculation show that there would be very little ground pick-up with such an antenna. In 1957 I went to Caltech to get a PhD. Of course, Sputnik was launched at my first semester at Caltech. It seemed important but I didn't realize how important it was to me. The next semester I actually joined a radio astronomy group there. They had just finished the major construction of the first Owens Valley interferometer. I started taking astrophysics courses. My one cosmology course was taught by Sir Fred Hoyle and philosophically I really liked the steady state theory. At that time Allan Sandage at Caltech measured the Hubble constant to be about 50 and Gerard de Vaucouleurs at the University of Texas measured it at about 100. They would argue with each other, if they had simply averaged their 2 values they would have gotten about the right value. But anyway, these things were unknown to almost a factor of 2. In my thesis I used one of the large antennas to make a map of the Milky Way. I did that by pointing to the west of the Milky Way, letting the Earth's rotation scan the beam of the antenna through the Milky Way. If you look in the control room there you won't see any computers, there's a chart recorder. The pen of the chart recorder would go along and as we got to the Milky Way it would go up and then it will go back down on the other side. Later I rip off the chart, take a meter stick, draw a baseline and measure up from there. This of course has to be wrong because we're inside the Milky Way. Any direction we look we're going to be seeing something from it, but fortunately for my thesis the Milky Way is very thin compared to its diameter, so I got my PhD. After finishing it and a one year postdoc, I took a job at Bell Labs at Crawford Hill. You can see the 20 foot horn-reflector up there at the top of the picture. Arno Penzias had been hired a year or so earlier. He had done a radio astronomy thesis at Columbia with Charly Townes. Why did Bell Labs hire two radio astronomers? I think if they probably told the management that we'd be good at understanding large antennas receiving systems, what happens when we look through the atmosphere. I think there were two other things going on. First of all Art Crawford who was our first department head and who had supervised the building of the 20 foot horn-reflector, had come to Bell Labs in 1928, at the same time as Karl Jansky. And Karl Jansky, as you know, discovered radio astronomy, if that's the right word. They had a room together and there was a feeling at Bell Labs that his discovery was not followed-up very well, and maybe it was time for Bell Labs to do something for radio astronomy. Also Art and others there were very proud of what they had actually been able to put together, and wanted to see this nice receiving system used for something more than just one brief satellite experiment. Arno and I knew about the low noise capabilities and the possibility of calibrating the 20 foot horn-reflector. It seemed like a way to do science that no one else is doing, if I may quote one of the titles here. Always a very important thing to try to do. After I was there, Arno and I got together and set out a series of plans, things we wanted to do. The first one was to measure the absolute brightness of one source. This would be actually a service to the astronomy community and to the satellite community because you could check out an Earth station by measuring the signal to noise ratio on CasA. Second one, was to look for a halo of radiation around the galaxy, in other words fix up my thesis. It also had the advantage that it was looking into something which had not been measured before. The highest frequency at which such a measurement had been made was about 400 megahertz. We were going to start out at 4 gigahertz, then build a comparable system at 1.4 gigahertz or 21 centimetres, and among other things patch up Arno's thesis of looking for hydrogen in clusters of galaxies. I think all of our projects were using the unique capabilities of the 20 foot horn-reflector. Arno made a liquid helium cooled reference noise source and I put the radiometer together to be able to compare it. Our idea was to start out at 4 gigahertz where we thought the galactic halo had to be very small, we could make a null measurement there and then shift to the other frequency where we expected to see something. When we got it all together it was actually a considerable disappointment. Does this make ... yes. Oops, it also shifts. We have the liquid helium cool load here with power increasing to the right and the helium itself is 4.2 Kelvin. Using all these numbers down here we could calculate that the terminal temperature, the wave guided room temperature, was about 5 Kelvin. And here is the antenna looking straight up with an absolute minimum of atmospheric radiation and pick-up from the ground. It was definitely hotter than the liquid helium, it was over 7 degrees. Excess noise in systems had been seen at Bell Labs before but we had a direct confrontation with the antenna because we were comparing it directly to the helium. By this time Dave Hogg and I had measured the gain of the 20 foot horn-reflector for the CasA measurement. So Arno and I spent the next 9 months understanding our receiver, seeing that we could calibrate it in several waves and they all gave the same answer. During this time the excess 3.5 Kelvin kept showing up. Every time we weren't looking at something we knew about, there was always the extra 3.5 Kelvin. Then at one point we heard about the possibility of microwave radiation left over from a hot big bang from the Gravity Group at Princeton. There's an interesting story but I don't have time to tell it today. Dicke was a physicist who had worked with microwaves in the Second World War for the radar effort and then got interested in gravity. Gravity caused him to think about a big bang universe, and we thought about it being very hot but expanding. He realized that the radiation would cool and it would be microwaves now. He got a couple of people to work on this. Anyway, Arno and I by this time had eliminated everything we could think of which might be the source of our extra antenna noise. We believed in physics, whatever was coming out had to come from somewhere, but we had no idea. We were happy to have some explanation but at that point I think cosmology had hardly ever explained anything and we were not so sure about the cosmology. We ended up writing 2 separate papers, we wrote a 1.5 page paper about our measurements and the Princeton people wrote a rather longer paper about big bang cosmology or big bang universe. However, the interpretation of the radiation as being from the big bang was rather quickly accepted. The big bang in rough outline: we start off with a very hot early universe. Was maybe quark-gluon plasma, particle-antiparticle pairs are created and annihilated. And it's expanding rapidly, cooling rapidly at about 10^-11 seconds. The energies get down to the level were comparable to accelerator energy. We can actually start to understand the physics, at about a microsecond baryons form and as it cools off a little more, the particles and antiparticles all annihilate and there can be no more generation of the pairs. Remarkably, we're left over with about 1 part in 10^10 of the original set of particles as matter rather than antimatter. At least we call it matter. At the time of a few minutes neutrons and protons can get together to form a deuterium, tritium, helium and a bit of lithium. The next interesting thing is at about 380,000 years: neutral atoms can form. The protons and electrons get together, become neutral, the universe becomes transparent. The radiation that existed then has travelled unimpeded until we see it now. In the early stage, when all of this annihilation occurred, the universe was very heavily dominated by radiation, but at about this point it became matter-dominated. They were left with minor fluctuations and structure in the universe developed from that. Here we are looking at our universe with a lot structure. After about a year there had been several points on the Rayleigh-Jeans part of the black body curve but no indication of the turnover that has to occur in a black body spectrum. It was only in the '70s that that occurred. Then in 1990 a remarkable spectrum was published by the COBE satellite. This is both a theoretical curve and the points they measured with error bars are entirely within the blue line. It was, I think, a very remarkable measurement of a black body curve, way beyond the peak. When we started out, what we said was that the extra temperature is about the same everywhere we look. The COBE satellite, as George knows well, did a lot more than just the spectrum. By this time people knew that we were actually moving through the radiation field, and that in the direction we were moving there should be a slight blue shift and in the other direction a red shift, so there's a dipole component to the radiation. George and his people were successful in removing this and this and getting this picture where we have a little bit of residual from the galaxy, but we see definite evidence of inhomogeneity and structure in the early universe. This was quite a relief to theorists because if that hadn't been there it would very hard to understand why we are here. There just shouldn't be much structure. Later the WMAP satellite did a much better job of this and people started fitting theories to it. We started seeing what I think John Peoples called „precision cosmology“ where we were actually getting numbers, not with a factor of 2 error, but with several decimal points of accuracy. Meanwhile optical astronomers were measuring the structure in the more recent universe. The structure they measure is quite consistent with what you expect from the early picture of the universe. Meanwhile the Hubble diagram has been extended. Down here in this little yellow spot is the original place where the original one was. Here in 1995 one has a remarkably good linear expansion that one could be quite happy with. In 2004 as we heard Saul Perlmutter say on Monday, suddenly there was a disturbance, it was no longer linear. The universe, while its expansion was slowing down earlier is now expanding. But he talked about that, so I'll leave that alone today. Everything seemed to fit fairly well. We have bright matter that we see being only a tenth of a percent of the mass of the universe. About 4% of ordinary matter, dark matter. And 23% of some other kind of material that we don't understand, non-baryonic dark matter. These components can be seen in both the motion of stars and galaxies and in gravitational lensing, the two measurements agree. Then we have 73% of dark energy, which I'll get to in a moment. But there had been some nagging problems from the beginning. First one being that if we look in two different directions and measure, we get essentially the same temperature. But if you go back in this sort of symbolic thing with time, the radiation released here and here was from regions that had never been in causal contact. Why were they essentially the same? Second, if you start out the universe with something other than exactly the critical density, then the expansion will be very different. Here we have a 24 digit number for the density at one nanosecond after the big bang. If we increase the density by one in the least significant digit, the expansion is slowed down by gravity so much that a structure never develops and we don't occur. If you go the other direction, the expansion goes so fast that, again, we couldn't be here. The explanation of this came about when Alan Guth heard a talk by Dicke about the density problem and he had the basic ideas of cosmic inflation. He then heard from his colleagues about the uniformity problem and realized that inflation would solve that also. In 1980 he proposed it, and then there were developments since then. The idea is that by taking a tiny bit of the universe and having an inflation, then the properties of that tiny bit of the universe are changed into our whole universe. And not only that, the density will exponentially approach the critical value. After 60 e-folding times we will have what could become our present universe. The problem with this, of course, is that it requires new physics, a metastable state of the vacuum so that this inflation can occur. But inflation invokes quantum fluctuations which can seed the structure of the universe. This is the picture the Planck satellite made which is somewhat better than the WMAP satellite. Here we have measured values of the spectrum essentially of this picture. The predicted values with a 6 parameter fit of cosmic inflation, followed by big bang expansion. It seems like a remarkable fit of theory to data. Furthermore, the measurement of optical astronomers of the distribution of matter in the present universe is quite consistent with this. Although highly successful the inflationary paradigm represents a vast extrapolation from well-tested regimes of physics, in both quantum fluctuations at 10^16 GEV and time scales less than 10^-32 seconds. Something we don't know really very much about. But inflation also predicted more. There should be polarization of the cosmic background radiation. I don't think I have time to explain how that occurs but there are two modes, a fairly simple one called E-mode, because it looks like electric fields. And one that depends on a curl that's called B-mode because it's sort of like magnetic fields. The E-mode has been measured. It pretty much has to be what it is if the Planck picture is correct and, indeed, those fit together very nicely. The B-mode is another factor of 10 less intense and it turns out very difficult to measure. It's a unique prediction of cosmic inflation, resulting from quantum fluctuations, and as I said, very weak. The idea is that during the inflation there are density waves which, at the time when the cosmic microwave background is separated from the matter, cause the picture that we see from the Planck satellite. There are also gravitational waves stochastically generated because of quantum gravity. They expand ... in principle one might be able to detect them in the modern universe but they're entirely too weak. But at this time they, due to the distortion of space, should have imprinted the polarization on the cosmic microwave background radiation. Last year the BICEP2 organization announced a measurement of B-mode polarizations. It then turned out, the Planck satellite people who had measured dust in our galaxy, the foregrounds showed that the dust could explain a lot of what they actually measured, it could generate the same B-mode polarization. Then eventually the 2 groups got together and actually compared the data. The result now is that most of what the BICEP2 people saw is probably dust, however, the most likely value is actually positive for the primordial B-mode polarization. We very much need better measurements of both the dust and the actual B-mode polarization. Those experiments are ongoing. The BICEP people are improving theirs, I think Planck is probably at the end of what they can do. But there are several other experiments measuring at several frequencies that should be able to make the separation. I can tell you that those 2 guys in 1965 didn't know how important the CMB would be. I feel extremely lucky that I had a job at Bell Labs where I could take technology developed for communications and apply it to radio astronomy. And that I got into astrophysics and cosmology at a time when it was a very young science but poised to blossom as technology improved and we were able to do more. As you gathered there was no real Aha moment in the discovery of the Cosmic Microwave Background radiation. There was a long period of trying to identify the problem. And the importance of the CMB only became obvious later as experiments and theory one after the other improved, so that we understand what it can tell us and actually can measure it. Looking back I feel good that we did our job right, that's very satisfying. It's very satisfying to see what has come of all of this. If you go outside some of these photons are going to hit you but your skin is not the same of kind of detector as our antenna. Thank you.

Robert W. Wilson (2015)

Cosmic Microwave Background Radiation and its Role in Cosmology

Robert W. Wilson (2015)

Cosmic Microwave Background Radiation and its Role in Cosmology

Abstract

In the first half of the 20th century other galaxies were recognized, their red shift measured and theories of the whole universe were developed. They included Big Bang and Steady State. Arno Penzias and I found the Cosmic Microwave Radiation (CMB) in 1964 and it was interpreted as thermal radiation left over from a hot big bang, ruling out other theories. In the 50 years since then, observations have improved so that the CMB gives a detailed picture of the universe at the age of about 380,000 years. That combined with the 3D distribution of galaxies gives high confidence in the standard model of the universe. The Big Bang must have been started in a very specific way and Cosmic Inflation is thought to explain how that happened. More evidence supporting Inflation may soon be available from the CMB.

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT


Related Content