Popular Talk by George Smoot

Popular Talk on cosmology and observing the early Universe given by George Smoot at the (January) 1997 Partners In Science Conference sponsored by Research Corporation , a foundation for the advancement of science. This is a conference for High School teachers and their mentors.



Cosmology is the study of the universe as a whole: how it came into being, how it developed, where it came from, and where it's going. What is the most successful theory of all time? I use this question to tease my theoretical physicist colleagues. It's a theory of the universe we owe to the Greeks, but it's called the Ptolemaic picture because Ptolemy later corrected it. Its original form dates almost 2,000 years before the Big Bang theory. It's a reminder that all societies have had explanations of who they were, where they came from, and their place in the world.

The Greeks were the first to make systematic observations of the world around them to try to explain they saw occurring. They actually had several different models of the universe, including one that's very close to the Copernican system, the system that we teach students today. It was championed by Aristarchus, but it was ruled out, not because of religious or cultural reasons, but because of observation.

At the time, people said that if the Earth goes around the Sun, then our view of the stars should be affected by parallax. Here are the stars, out at some reasonable distance. If the Sun is here and the Earth is here and if it's moving around the Sun, we should see changes. When you are on one side then the angle to the stars would be different from when you are on the other side, for the same reason you have depth perception. The Greeks looked very carefully, and they saw no parallax. Aristarchus was not ready to give up his model, and decided it must mean the stars were very, very far away.

 Matters slid for a while, but a few years later Aristotle proposed a model based upon his theory of physics that things in the heavens moved in circles, the Sun, Moon, planets and stars, with the Earth at the center. This model took care of parallax, because we were in the center, and so the angles were always the same to everything else. Everything moved in circles. The model had a series of concentric crystalline spheres - one per planet.

The Greeks knew the Earth was a sphere, and in fact they measured the radius of the Earth to an accuracy of about 10 percent using the simple geometry you learned in high school. Eratosthenes, a Greek who lived in Alexandria in northern Egypt in the second century B.C., knew that on June 21 at Syene in the far south of Egypt, the sun shone directly vertically down a well. The sun made a 7.2 degree angle at Alexandria, and the distance from Syene to Alexandria was measured to be 5000 stade (one stade equals 500 feet). So using simple trigonometry, or plane geometry, he calculated the radius of the Earth. The Greeks had actually calculated the size of the Moon and its distance from Earth using simple arguments: basically by looking at the size of the shadow of Earth during eclipses, figuring out what the geometry would be, knowing roughly the angular size of the Sun, and observing how the shadow changed. (see Explanation for information, problems and solutions to these high school level geometry problems.)

 The Greeks were good observers, and they chose the Aristotelian model for a reason: by using it they were able to explain the motions of the planets with fairly high precision. It was the model that actually reigned for nearly 2,000 years (slightly corrected by Ptolemy and others).

About 400 years ago, however, our world was shaken. Copernicus, the leading astronomer of the day, was called in by the pope in 1514 and asked to reform the calendar because it was inaccurate. Copernicus told the pope he would do it, but there was one problem bothering him--the motion of the planets which could not be accounted for by the Aristotelian model. This, in fact, was something Aristotle had pointed out--the motion of the planets had to be understood for science to be placed on a firm footing.

 At any rate Copernicus studied the motion of the planets, and as you know it got to be such a controversial topic, in disagreement not only with the Aristotelian view but Christian theology, that he didn't publish it until 1543, the year he died. The Sun was the center of the universe, he maintained, and the Earth orbited it along with the other planets.

 Some years later Danish astronomer Tycho Brahe provided evidence that supported Copernicus. Although not the first to do precise work, Tycho systematically made very good observations (the best pre-telescope) over many years creating an extremely valuable data set. He was the first to realize that he would have to break those crystalline spheres to change the model. And he discovered a comet, made observations, and showed that the comet came from outside the sphere of Mars and passed inside the sphere of the Sun. It was actually passing through the crystalline spheres that the Ptolemaic model called for.

Tycho observed changes in the sky, and that things were not the way that people expected. Changes in theory were called for, but they weren't taught. You know how conservative traditional schools can be. In those days, they kept teaching the Earth-bound Ptolemaic model even though there were alternatives available. Only when you went to the university and took advanced astronomy did you hear about alternative theories. What it really took was for Galileo to come along and overturn Aristotelian physics.

 When change did come it was revolutionary, because the Ptolemaic model had been thoroughly absorbed into the culture. Here was Earth, and in Aristotle's view things that were heavy and base fell to Earth. Those things were humans in the religious view; if you were going to be wicked and be condemned to hell, you had to be on Earth. The further out you went, on the other hand, the more holy and spiritual you were. Obviously God lived up there. Everything was thus satisfactorily explained, and now these damned scientists and astronomers came along, and tried to overturn it all.

 This turn of events was in keeping with much of the history of mankind. People have always had a mythology about the creation of the universe, how they came into being as people, and their place in creation. Indeed, we're now in a period of transition between these kinds of models. But there was a revolution in the 16th century, as astronomers turned their telescopes to the heavens, they found that the Greeks, perceptive as they were, had made mistakes. Using simple trigonometry to estimate the distance to the Sun can be difficult. When you have a half moon, sunlight strikes the Moon and the angle to the Earth is a right angle. If you know the distance to the Moon and can measure the angle to the Sun between the point of the Moon and the Sun, you can use trigonometry and calculate the distance to the Sun. The angle turns out to be extremely close to 90 degrees, so a small error makes a big difference. The Greeks got a distance that was too short by a factor of 20. I suspect that they fudged the data because they couldn't believe the Sun was that far away. They knew the Sun was much bigger than the Earth, in fact had done calculations, but to be that much bigger than Earth was perhaps not credible at the time.

It was confirmed at the beginning of the Renaissance that the distance to the Sun was wrong, and people started turning their telescopes to the heavens and measuring the distances to the planets more carefully. Every time they made a measurement, the distance to the planets was further than predicted, first by a factor of 20, and then more. Then people started measuring the distance to the stars. Before Galileo turned his telescope to the heavens and saw many more stars than one could see with the naked eye, people had actually tried to estimate the distance to the stars by memory. A French astronomer, physicist, scientist and mathematician devised a technique for measuring the distance to the nearest star and chose Sirius for a test. He would go out at night and stare at Sirius for hours until he'd memorized how bright it was, and then he'd go in his house and go to sleep. He'd wake up at noon the next day and make a series of holes in his roof. Each day he put a different size hole in the roof, until he found one with exactly the same size and same image brightness as Sirius. If it weren't for the fact that Sirius was so much brighter than the Sun, he would have gotten approximately the right distance.

Once people had telescopes and new methodologies, they began to understand scale. During the 16th century, estimates of distance from the Earth to the stars increased by a factor of about 10,000. This was a theological problem, it's harder for God and the angels to look down on man when they're 10,000 times farther away; there's a psychological difference. In 1576, the pope issued a new canon. No longer was God thought to be in one place, but as infinite and all-powerful. He made stars that extended out to infinity. This also turns out to be a problem, but before I discuss it, I want to digress briefly. Just keep in mind that the pope had stars running out to infinity.


The current model of the universe

 Today's model has the universe expanding following what we call the "Big Bang" (but we don't mean a big explosion). The universe is expanding and evolving, and it's cooling down as it does so. If you looked back roughly 15 million years ago, you would find that the universe was pretty much a uniform soup and, as it expanded and cooled down, that soup coalesced like milk curdling and created stars and galaxies and clusters of galaxies. These things evolved over time so that if you were able to look back to earlier times, you would see the first generation of galaxies, you would see quasars, and you'd see lumpy things that are just forming. The universe over time is like an embryo, developing more and more features.

Roughly five billion years ago our own solar system formed, and we think that it's made of second- or third-generation materials, not the primordial material of the Big Bang. We've already been processed though the stars that created the elements we're made of. We've gone through a second generation which makes planets and life and everything else possible. We have this view of an evolving and changing universe which is very different from the static, infinite universe that the pope called for in 1576.

The Big Bang model was one in a number about 50 years ago, and became a leading one in 1965; and the dominant candidate in the last ten years. In the next decade we'll get to a point that will be quite startling: we will be able to test this model with precision. Dr. Lederman told you that particle physics has a Standard Model, and it has been studied very carefully and tested to a great extent. Now people are looking for tiny variations in that Standard Model, at the level of 1/10th of a percent, 1/100th of a percent, parts of 1000, parts of 10,000, to see if the it's going to break down. But we think we really understand how the simple particles like electrons and quarks interact with each other, and you can set up experiments and test them very precisely.

We're going to be at almost that point with cosmology in the next ten years. We have now visualized, and in fact launched in some cases, experiments that are going to test in different ways our model of the Big Bang to about the 1 percent level. And once we've tested the model, we're going to be able to extract from that same data a set of 10-15 parameters that will describe the universe in a very clean, thorough way. How is this possible?

 Just imagine that you need a new suit and the tailor is an alien from outer space. In order to get a good fit--he's never seen a human being before--he needs to measure every square inch of your surface area. If you bring in a human tailor, he measures your waist, sleeve length and shoulders. He has a model of a human, and by making a reasonable set of measurements, namely your waist, chest, inseam and arm length--perhaps six to ten measurements--he will be able to go back and make a suit that fits you extremely well. If he wanted to, he could make a dummy that looks like you. In the same manner, we're going to reduce the essence of the universe to 10 numbers, and we're going to know those numbers to about the 1 percent level. This will tell us about the universe because we have a model, just as the tailor has a model of a human being, and because we know certain laws of physics.

 If I know the ingredients of the universe at its beginning and the important laws that govern its evolution, and if the original situation is simple enough, I can make predictions on how the universe will evolve. I can't predict exactly where Mars will be, but I can predict that there will be stars and galaxies and roughly how many of each, what the life cycles of the stars and galaxies are going to be, and how fast the universe is expanding.

In short, we should now have a remarkable new model of how the universe was created, how it evolved and how it changes, and our model will have an incredible amount of precision. The analogy that I want you to embrace is that of the tailor--the suit may not be a perfect fit, but it will be pretty good.

In 1576, however, the pope had just given us an infinite universe that went on forever, and that was the beginning of the new cosmology and other dilemmas. If there are an infinite number of stars going on for infinity, and it's been that way forever, then we're faced with what's known as Olbers' paradox. If the universe goes on forever and it's been there forever, and if you look out at the night sky, then eventually your line of sight will end up on a star no matter where you look. It doesn't take very long before you realize the entire sky should be as bright as the surface of the Sun, or even brighter. The next question is, why aren't we fried?

There are two solutions: one is to assume that the universe has only existed for a finite time, so the light from really distant stars hasn't yet reached Earth. The speed of light is only 300,000 km per second, and if the universe has only been around for a few seconds, you will only see out a short distance. If the universe has been around for a billion years, you will only see out a billion light years. The second solution is that because the universe is expanding, it's shifting the frequency of the light. If it's expanding rapidly enough, light is moved out of the visible into the infrared; if it's expanding even faster, it's moved down to microwave and so on.

Although both these theories still existed in 1960, both required an expanding universe; in one case a universe that had existed for a finite time, in the other a universe that was expanding and would continue to exist forever. There had to be more distinguishing features.

Back in the 1600s and today, however, it is evident that when you go out at night you don't get fried. And even if there is light down in the infrared, the total amount of energy striking the Earth from the universe is much lower than what would be needed to cook you, from microwaves all the way through to gamma rays. We know the atmosphere doesn't boil off. So Olbers' paradox was actually a key one, because it indicated that the universe must not be either static, infinite or uniform, but that one or more of those things must not be true. It also tells us that the total amount of information coming to the Earth is finite, but that's another story.

 This picture from the Cosmic Background Explorer (COBE) satellite shows that we live in a spiral galaxy. You're going to ask how we got this picture of our own galaxy, how far we had to go. The answer is, we don't live downtown; the nucleus of our galaxy is very bright, and we live way out in the sticks. If you live near a big city, you can always see the bright lights of downtown late at night, and that's what you see in a COBE picture. If you're a galactic astronomer, you're really impressed, because the picture is actually the first good view of our galaxy. It is not quite in the optical, but in the infrared which gets through dust much better than visible light. Instead of dark regions where there are dust and gas clouds that obscure the stars, infrared reveals streaky brown regions. You can see the starlight leaking through behind them. If you're a cosmologist and you've studied Olbers' paradox, you'll be interested in the dark areas. We can see outside our own galaxy because we live far from downtown and the stars around us are not very important.

 To do our cosmology we must focus on all of the darkness out there. What we see is essentially no radiation in the visible or infrared between the stars. And if we look for any intervening material--there should be detectable absorption if there's any intervening material--we find that most of space is a vacuum. Matter is clumped in our galaxy, as far as we're concerned. It's clumped in the stars or in the big clouds. Those clouds are ordered in a very peculiar way into a flat plane or into a ring around a central bulge. There are arguments about what might be going on, but what's clear is that ordinary matter, the kind of matter that you and I are made of, is very much clumped in terms of our own galaxy.

The vacuum between us and any star is a better vacuum than anyone has achieved on Earth. We talk about these clouds being condensations of matter, but a lot of them have 100 atoms per cubic centimeter. If you should ever get to 100 atoms per cc in a laboratory vacuum, you could call the press in and make a big deal out of it. If you were able to look out and peer through the dust a bit you would see the nucleus of our own galaxy. Although we're still in the galaxy, we're two-thirds of the way out, and what's behind us is much less dense than what lies towards the center. This allows us to do cosmology in a very straightforward way.

 If we are part of a galaxy then, the next question is: are we living in an island universe? We can solve the problem of the dark night sky if it can be shown that we live in an isolated system of stars. Well, in the 1920s Edward Hubble decided to attack this problem. He wanted to understand the distribution of what were then called nebulas, how they operated, and how they behaved in the night sky. Eventually he discovered a technique for estimating how far away they were, and that there were a number of nebulae that were at least ten times as far away as typical stars in the galaxy. That meant that they were outside the spiral structure that we live in. Hubble founded a field called extragalactic astronomy, and one result is the Lick Sky Survey, a survey of the northern hemisphere, now 25 years old. It shows the location of a million galaxies.

The distribution pattern shows that the galaxies have a funny structure and they're not uniformly spread across the sky. The pope wasn't quite right, for the stars aren't uniformly distributed, although they might appear that way if you go to a really large scale. Galaxies are often together in groups and large clusters. The cluster (Coma) near the north galactic pole has a couple thousand galaxies in it. Even clusters of galaxies are often in larger groups of galaxies, and there could be clusters of clusters of galaxies. There also seems to be a string, or a connection of clusters of galaxies in this region, while in other regions there are almost no galaxies at all. If you look carefully, you find that there are huge regions, sometimes 50 million light years across, where there are no galaxies. And right next to some of these in a volume of ten million light years across, they'll be 2,000 or 4,000 galaxies. There's this grouping and clumping, and it gets more and more complicated as we study it. It's clear that there's neither uniform distribution nor just dark, and that somehow there's a mechanism behind these structures we see in the universe.

Every time we look on a larger scale, we see larger-scale structures. How is it all put together? That is the question we must answer if we want to explain how the universe evolved and how it works. We've surveyed up to four million galaxies. The photos show there are clusters, and the clusters are grouped, and there are voids. But now the voids are beginning to fill in, and we can see galaxies behind them. One needs a three-dimensional plot, and I have one that includes observations done in New Mexico by the Center for Astrophysics from Harvard.

Work on this plot was started by Margaret Geller, John Huchra and their students. Basically they took a series of pictures with a telescope as the Earth rotated. As you continue taking these pictures you get a wedge of the sky, and one series of wedges seems to show the stick-figure of a man. There are also elongated structures, known colloquially as the "fingers of God" because they're pointing at us. We now know that the reason they look like structures is because the method used to estimate distances was based on either the red shift or intensity. When you have a cluster of galaxies and you've estimated that some galaxies are different than others in the cluster or have speeds relative to other members, you have a cluster that gets stretched out along the direction in which you're estimating the distance. This happens because, while you can get angles very accurately, you can't estimate the third dimension very accurately. So the "fingers of God" are just an experimental error, and the stick-figure of the man is simply anthropomorphic.

Another incredible structure is a big flat sheet that has been given the imaginative name, "the Great Wall." The Great Wall is a big plane of superclusters of galaxies, and it's at least 300 million light years across. That means that if the universe was fairly uniform in the beginning, you had to move this thing at close to the speed of light over the whole history of the universe in order to create it. So it was an awesome engineering job. The voids are much clearer when you see them in three dimensions. There are these bubbles in the universe, and they look like foam.

So we've gone from Olbers' paradox, to knowing that the universe isn't static, to discovering there are galaxies outside of our own with a complicated distribution in the sky. But there's another thing: Hubble wasn't satisfied with discovering galaxies and starting a whole new field of astronomy; he wanted to know where the galaxies were going, so he decided to measure their motions. He moved from Arizona and New Mexico to Palomar on Mt. Wilson to make observations of galaxy brightness versus recession velocity. The technique he proposed to use? The Doppler shift.

 My explanation of the Doppler shift for high school students is the following: I have a Doppler dog, and he's trained to bark very uniformly--bark, bark, bark--once a second. As long as he's sitting still, uniform circles of sound come from him, and wherever you are, you hear the sound bursts come to you at the same speed. Now if the Doppler dog runs and barks at the same rate, his mouth moves forward and the center of each of those concentric spheres changes. His first bark was at one point, his second at another and so forth. And you can see that if he's moving toward you, the barks are going to reach you more rapidly. And if he's running away from you, they will be further apart. If you knew he's trained to bark once a second and you measure the arrival rate, you can tell his speed along the line of the sound you are hearing. You can't put a dog in every galaxy, but fortunately we have trained atoms, and they all have characteristic frequencies. If you look at the frequency shifts of spectral lines from atoms, you can tell how fast they are moving.

With the data he collected, Hubble estimated the distance to the galaxies in millions of light years. And the recession velocity of the galaxies, as represented by the fractional increase in their spectral lines or redshift, was 0 to 1/3 percent. There were actually two sets of data, because you can treat galaxies as either individual galaxies or as clusters, but both were close together. So, when Hubble looked at this data, did he say "we have a new cosmological law?" The data were best fit by a straight line, now called the Hubble Law. Well, I think the data are adequately fit by a straight line because you don't have good enough data for anything better.

 Fortunately the Carnegie Institution thought Hubble was a great guy and hired Milton Humason to help him out. Two years later, by 1931, they had measured out many times as far to 100 million light years, and up to almost 7 percent redshift or speed of light. With that many data points, a straight line was a pretty good observation. So the conclusion in those days was that the nebulae (they still didn't call them galaxies) and the entire universe were in recession. They were moving apart. This led to the Big Bang theory that the universe is expanding.

Unfortunately "Big Bang" is a misnomer, and I'm going to try to persuade you that it's just a catchy phrase. The idea of the Big Bang is that there was a huge explosion, things flew out, and those moving fastest went furthest. When you measure their distance and their velocity, you find a linear relationship--and that's exactly what you would get in an explosion or a race. If we race a tortoise and a hare and measure the distance they cover in a certain unit of time, and each has a constant velocity and the tortoise is moving half as fast as the hare, you will get the same relationship whether you're talking about one unit of time, two units or three units of time. There will be a simple linear relationship, and it is important to know the distance scale in order to know the velocity. Indeed, this is the major problem we still have in modern cosmology. If everything was crunched together at the beginning and there was a giant explosion, then you could explain what Hubble saw that gives this relationship.

Well, I told you the Big Bang theory was wrong, so now I'll tell you what I think is right. We've got to go back to the ancient Greeks, actually to the Egyptians, because geometry was very important to them. They lived by the Nile, and needed the Nile floods to replenish the land. They had to get the crops planted fast enough so that they could eat and so the Pharaoh could collect taxes. This required rapid survey of field boundaries so crops could be planted. Geometry in ancient Egypt was an honored profession for improving surveying. Eventually Euclid condensed everything down to five rules called Euclidean Postulates. For the next 1800 years, people tried to show that Euclid's fifth postulate was not necessary, and so a lot of mathematicians wasted a lot of time. The fifth postulate, I'll remind you, is that parallel lines stay a constant distance apart, and the sum of the angles in a triangle is 180 degrees.

In the 1820s, Gauss began to speculate about a geometry in which parallel lines could diverge. If you think about a funnel, for example, lines spaced out uniformly around the bottom will diverge as they travel outward. One consequence to this is that the sum of the angles of a triangle or a funnel will be less than 180 degrees.

Eventually, Gauss, Lobachevsky and Bolyai independently came up with a hyperbolic, or equivalently negatively curved, geometry. Gauss was so convinced that it was true that he hired surveyors to go out and measure the sum of the angles of a triangle between mountain peaks that were 100 kilometers apart. Of course they were close to 180 degrees when you correct for the curvature of the Earth and just measure straight-line distances. But Gauss was convinced this might be the real geometry of the universe. Though it seems obvious, the idea of converging parallel lines (e.g., on the surface of a sphere) took a while. It wasn't until Georg Riemann came along and developed the theory that allowed us to calculate geometrical curvature that we got some idea of what a closed universe might look like. A sphere is a good example: if you have positive curvature, the sum of the angles in a triangle will be greater than 180 degrees.

 Riemann was convinced that he not only understood geometry, but that he understood all physics, because all physics in those days was electrostatics and gravity. What happens with electrostatics and gravity? Well with gravity, two things attract in proportion to mass. And with electrostatics, opposite charges come together and same charges move apart. Reimann thought he could geometrize all of physics but he was wrong because he forgot about the fourth dimension: time. Then Einstein came along with the theory of special relativity, and eventually general relativity. Einstein came to the realization that there is a relationship between the density of mass plus energy in space, and the curvature in space-time. A massive object like the sun will warp space-time and cause light passing near it to follow a curved path.

 The general theory of relativity was published in 1915 and tested in 1919 by an eclipse. An eclipse was needed because observations of starlight passing near the Sun would otherwise be impossible. With an eclipse, it was evident that Einstein's theory was right: the position of the stars close to the rim of the Sun appeared to shift as compared to observations of the same stars without the Sun. The amount they were further apart was very small, but it was a measurable quantity. People were able to measure the position of the stars when they were on one side of the Sun, and when they were on the other side of the Sun, and show that in fact light curved, and it curved by the right amount. Einstein's original prediction, which was made in 1915, didn't take time into account. He did the rough calculation and it was off by a factor of two, because he only took the bending of space into account; taking space and time together, you get the full factor of two. The full factor of two was the curvature of space-time; when you allow space and time to have curvature, you allow them to have a scale. That scale can change with time.

The Expansion of Space

We now have a different view of how the universe is expanding; it's not due to a giant explosion, but because space itself is expanding. Imagine that space has only two dimensions (it is too hard to visualize it with three). And imagine that the spatial dimensions of space are like the surface of a balloon. We ride along a spot on the balloon representing our galaxy, which is being blown up at a relatively constant rate. As you can imagine, any two spots on the balloon will move apart. Depending on where you are, you'll still find a relatively linear relationship between the distance between these two spots and the rate they appear to be moving apart. So that the rate they're moving apart is just their separation multiplied by the fractional rate at which the balloon is expanding. You get exactly the same results as with the Hubble law, but it's a much different result in terms of psychology.

What's happening is that the space between the galaxies is increasing and growing, rather than the galaxies themselves moving out into space. The wrong view is that space already exists and the galaxies are being blown up into it, either on a balloon or in an explosion. What's really happening is that space is being blown up and the galaxies are going along for the ride. All of the problems--things moving at the speed of light--disappear in this view because locally one galaxy isn't moving relative to any other galaxies around it. It's only when you look over the long distances that space is expanding--like a rod that's being heated and expands. Nothing local moves very fast, but the ends move apart fairly rapidly. Remember that it's really space that's expanding.

 We also have to study the structure of matter. The history of physics starts with matter at low energies, adds energy to it, and breaks it apart to see what it's made of. For the history of the universe, it's the opposite. You start with a hot universe and put together building blocks. At each epoch, those building blocks are bound with much less energy than the previous. As the universe cools down, you get more and more fragile objects.

Water, as an example makes some very complicated and fragile things--snowflakes, snowmen, ice houses, ice sculptures and so forth. When you have frozen water as a solid, heating it up gains a symmetry; that symmetry is what we call a liquid. If you heat water up even more into steam, you have the molecules bouncing off each other. If you look at it, you realize that water is made up of two hydrogen atoms and an oxygen atom. And if you add more energy to the system to break it apart, you get individual atoms; looking at the atoms you find that each one is made of electrons and a nucleus. The nucleus is made of protons and neutrons. Now you've got a really pretty picture: everything you know of in the universe is made of three things--protons, neutrons and electrons.

This would explain everything, but life didn't come out this way, because now we know things are made of quarks, too. And if we look down inside a proton and a neutron, we know they're made out of quarks. But the nice thing about it is that there is a hierarchy--the energy levels are usually an order of magnitude or several orders of magnitude apart. And so, in each regime you're dealing with one class order.

The universe starts out at a very high energy. It was very simple, because quarks interact in a very simple way; protons in a more complicated way, atoms even more so, and liquids and solids are even more complicated. As the universe cooled down it made more and more complicated structures, building up those levels of complexity. The history of science starts with the complex and progresses toward the simple: the world, is made of molecules, the molecules are made of atoms, the atoms are nucleons and electrons, and the nucleus is made of baryons, and the baryons are made of three quarks. That's the layered structure that Professor Lederman discussed in his talk. But in our cosmological view of the universe, we must start with those quarks, and work our way up until we make the Earth.

Our idea of Big Bang cosmology is to start out with quark soup at the beginning and build to make protons and neutrons. We then burn the protons and neutrons to make the light elements, which are going to form atoms. Those atoms form clouds, and the clouds form galaxies and stars. Finally those stars are going to burn the light elements into heavier elements. Then the stars generate the elements that we're made of. In the beginning there were just quarks and electrons, which built up to the things that we see today. That's the whole process of how we think the atoms and the materials that we're made out of were generated.

 Bringing us to this stage were Einstein with his general theory of relativity, and George Gamow and his co-workers, Ralph Alpher and Robert Herman, who worked out the idea of how the elements formed. There was an alternative theory still around during this period, and people didn't understand the structures of the nuclei before the 1940s. That's why we didn't have the Big Bang until our physics was far enough along so that we could understand it.

These were the pioneers who gave us the foundations for a complete theory of how the universe was created and formed. The high-energy physicists working with accelerators are learning about quarks, leptons and how they interact, and how things fit together. We were all so disappointed when the SSC didn't get funded because the SSC corresponds to the energy level at 10-12 seconds, micro-microseconds after the Big Bang. That's the point at which we think a big phase transition occurred in the universe, the time at which baryons started to form and particles gained mass.

Another thing happened back in the 1960s, a very funny thing. Somebody had the bright idea that you could save money by launching satellites and using them to relay telephone calls. Bell Labs hired Arno Penzias, now head of the successor to Bell Labs and Robert Wilson, a scientist there, and asked them to measure the brightness of the sky and the background noise, because satellites in those days were tiny and weak, and didn't have strong signals.

Penzias and Wilson made systematic measurements and found that the brightness of the night sky was about 20 degrees. Part of it due to the atmosphere, part of it due to the signals from our galaxy, and part of it due to their receiver. There were three degrees out of the 20 degrees that were left over. Now most people would have just swept that under the rug because they didn't understand it, but Penzias and Wilson studied it carefully for a year, and concluded that since the signal didn't vary regardless of where they looked, it either came from their instrument or from the universe as a whole. They went to the trouble of proving that it came from the universe as a whole, and 12 years later they won the Nobel Prize for being so careful and precise.

What they found was the light from the early times. The universe was much denser and much hotter in early times, and it had to be as hot as the Sun, or even hotter--a million degrees or more-- in order to cook the elements. There would be some radiation left over from that high temperature, and that relic radiation would be cooled by the expansion of the universe. Its wavelength was stretched and therefore its temperature was stretched as the universe expanded. If the universe wasn't expanding, the night sky would be as bright as the Sun. That expansion had moved the relic radiation from the hot phase of the Big Bang down into the microwave region. Right at the region in which they would want to make cellular telephone calls, only they didn't have cellular telephones in the 1960s.

That's how they discovered relic radiation coming from the early universe while exploring the possibility of satellite links. After they published that finding and there was confirmation, I realized that it was like light from the Sun: if you're careful with your instruments you can make a good image of the Sun. And if we were very careful we could make a picture of the early universe, and find out what it looked like when it was on the order of 10,000 to 100,000 times younger that it is today. That became one of my primary goals in science.

Cosmic Background Explorer Satellite: COBE

Other people had this realization too, and eventually three different groups proposed what became the Cosmic Background Explorer Satellite (COBE), NASA's first satellite dedicated to cosmology. We would make a picture of the early universe to measure the thermal spectrum, and to look for light coming from the first generation of galaxies. Our satellite included three experiments with all the instruments stored inside a shield. We were well above the atmosphere, and were able to make very good measurements of the whole sky.

We cooled our equipment down to the same temperature as the radiation, 3 degrees Kelvin above absolute zero; liquid helium was needed for some of the equipment. Any fingerprints or other contamination would be absorbed onto the cold equipment, so we had to make sure everything was clean and wear "bunny suits" while working. We wanted our satellite to be on the terminator line so we could look away from the Earth; so that we could always rotate the spacecraft so it looked away from Earth; and to keep it at 90 degrees to the Sun to shade it. And then we rotated the whole thing like a chicken on a barbecue to keep it a uniform temperature, and that also let us scan circles in the sky. We covered a lot of the sky on every orbit, and nearly half of the sky every day. After six months we swept the entire sky, and now we have four years of data.

A 1979 look down at the Earth shows a huge, but not surprising temperature variation: from 220 to 300 Kelvin. It's hot at the equator, no big surprise, and especially hot in the deserts; and of course it's cold at the poles. These measurements were made by satellite using equipment similar to the COBE instrument. They are checked by what we call "ground truth." You have weather stations all over the Earth, so if you wanted to check your satellite results you call up Buenos Aires and ask for the average temperature for 1979. You can get a grid of points to calibrate a map, and you know whether you have the right temperature or not. Now we had a problem: we don't have ground stations all around the universe, and so we had to test our equipment very carefully before and after launch to make sure we were getting the right values. It ended up taking about 20 people 20 years to make our picture of the universe.

At a wavelength of six millimeters approximately, the universe shows a diffuse uniform glow, and is different from what you see at any other frequency. There is a mean temperature of 2.73K, or 2.728K to be more precise. There's hardly any variation at all, it's a very uniform sky. That's because the universe on a really large scale is very uniform. If you look at the early universe, it has hardly anything at all in it, it's just a uniform soup. And yet the modern universe has a complicated structure, so our goal is to measure the properties of the early universe that allowed these structures to develop. Only at this wavelength--and it's really different from every other wavelength band, from the gamma rays down to radio--has nature provided a window that allows us to look out and see the early universe with incredible precision.

Pictures taken in the other wavelength bands show we do live in a spiral galaxy. Now we can look though a series of pictures taken at different wavelengths. The stars in the central bulge are around 3,000 degrees, and because they're burning very slowly they show up in images as a white puffy cloud. If you go down a factor of ten in wavelength to look at something with a mean temperature of 300K, roughly the temperature of the Earth, you can see the plane of the galaxy showing up and one spiral arm that encompasses Earth. There is also a structure with a temperature slightly below 300K--dust in the solar system. There's a lot of dust in the solar system because there are a lot of asteroids and comets. When the asteroids orbit and comets come in near the Sun, the asteroids grind against each other to make dust, and comets are heated by the Sun and spew out dust. In our studies we find that it's roughly 50-50: at least 30 percent due to comets and the other part due to asteroids.

If you go down another factor of ten in wavelength to 30K, you see the dust in our galaxy showing up very strongly. The Large Magellanic Cloud and the Small Magellanic Cloud show up, and there's a big cloud in the galaxy relatively near us where stars are forming that shows up pretty well. Then there is the planet containing solar system dust. You can see the Moon, Jupiter and the other planets in this plane. This is a very different picture than you see in the microwave, because the microwave is so uniform.

You can't go down another factor of ten in wavelength because that would be 3K, but you can go down by a factor of ten. A picture in this range shows cirrus dust which looks like the high, thin clouds that are highest in the atmosphere. These are the clouds that are the highest in the galaxy so they're the furthest from the stars and the coolest. These wispy clouds show up in regions at the edge of where the stars are. You can see what the galaxy looks like, and that roughly 100 times as much radiation comes from the plane of the galaxy than comes from far off the galaxy.

 Let's look in more detail at this radiation from the early universe. We have precise measurements over this whole long wavelength range, and can see that the spectrum is very, very precisely thermal, and that means that it really was thermal at the beginning. On an early map we see a region that's a little warmer than average and a region that's cooler than the average by a part in 1000. The galaxy shows up in its more intense regions at about a part in a thousand of the intensity of the microwave background.

We believe this stronger intensity is due to the motion of our galaxy and our solar system relative to the cosmic microwave background radiation. It's like driving in the rain, when you're driving forward your windshield will hit more raindrops than your rear window. In the case of radiation, you increase the energy of the photons coming toward you, and you collect more, so it looks hotter. You're decreasing the energy behind because you're escaping; photons are not catching up with you quite as fast, and you collect fewer of them. So you see a temperature variation, which explains everything, including the galaxies.

We believe that our galaxy is moving at about a part in a thousand of the speed of light. We know our galaxy's rotating, and that its speed is roughly two-thirds of a part in a thousand of the speed of light. But this is in the opposite direction, so that means our whole galaxy is moving at slightly more than a part in a thousand of the speed of light, and that's pretty strange because none of other galaxies around us--we're in a group of about 14 galaxies--have a very large velocity. That means the entire group is moving at about a part in a thousand of the speed of light. It turns out that the group is moving off towards a very large cluster of galaxies in the distance. It is given a very unusual name, "The Great Attractor," because when the clusters were discovered people started looking for a great attractor that was pulling our whole group of galaxies toward itself. They finally found it and the name stuck.

You can see the galaxy in a picture of the early universe; you see the spiral arms and the blackened center, and there's structure along the galaxy. Off the galaxy are regions that are cooler, and regions that are warmer, and unfortunately the contrast in such a picture isn't very great. We're talking about parts of 100,000, so the universe is incredibly uniform. That's about a factor of ten more uniform than a billiard ball. If you pick up a billiard ball, you don't feel any structure. If you could pick up the early universe you wouldn't feel any structure. You can see the seeds for those galaxies and clusters of galaxies on the Great Wall, these parts in 100,000 variations that we're seeing. And what we want to do now is go back and measure these extremely precisely on all angular scales, because that's going to give us a lot of information; the early universe is very simple, when things only change by a part in a thousand, linear theory works great. You can understand the universe very well if you do precise measurements, and most of the information we're going to gain in cosmology will be through precise measurements of these variations in the early universe.

If we shift the color scale of our picture of the early universe, we can emphasize the very large regions that are cooler, and the very large regions that are hotter. What you see is a very long wave that is cooler than the average, and then smaller bumps that are cool show up on it. It turns out that this is characteristic of what we call a scale and variant spectrum. There are equal size bumps on every scale, and they cover roughly equal fractions of the sky. So there are roughly equal, very large-scale structures, medium-scale structures and small scale structures, and so you'll have a long wave of something that's hot or cold and then the bumps that are the same size will show up on top of them. You'll have regions where there will be a whole lot of clusters together, because the clusters will come from small-size bumps, but they're going to be riding on a middle-size and a long-size wave. We think we understand it, and it's the only theory that makes sense: otherwise the universe would be filled with things bigger than a great attractor, and everything would fly around the universe and come crunching together. Or if it was tilted the other way, the universe would be full of small black holes and you'd see them too. So it turns out that scale and variant is pretty close to what we need in order to explain how the universe came into being.

That's why we're confident we're on the right track. Our theories are developed to the point where we have a model and we can really go out and make these observations and learn the parameters of the early universe. We have a concept of an advanced version of the Big Bang which includes a thing called inflation, and NASA and the European Space Agency have both approved and begun missions to make the high precision maps that we need. The NASA launch should be around the year 2001, and ESA will be around 2004.

 There's an exciting future ahead.


 1997 PIS Conference Summary/Smoot/