Saul Perlmutter (2015) - What We Learn When We Learn that the Universe is Accelerating

Good afternoon. I'm now trying to picture what will be the next scale up beyond this talk, the next for the afternoon, afterwards. Today, I want to tell you the work we've been working on, which I think may represent research... Make sure I'm here, yes. May represent research which answers questions which I've always imagined the very first human beings found themselves asking. In fact, it almost may define what it means to be a human being, that you would walk outside at night from, I imagine, your cave and look up at the sky and found yourself seeing the starry night and wondering, does it go on forever in space, and will it last forever in time? I've always thought that this might be the sort of question that anybody would have found themselves asking even before we knew how to answer. In fact, there were no really good ways to answer this kind of question except to ask the philosophers for almost all of human pre-history and history, until just the beginning of the last century, when Einstein set down his theory of general relativity. And, for the first time, we had some of the conceptual tools to address this problem. And, Einstein had a rather interesting moment in the summer of, I think, it was 1917 when he tried to apply his theory of general relativity to the case of the universe that we live in, and he ran into a, I always imagine that he's rather excited by that moment, but this photograph that you see here was taken that year, and that doesn't quite show the degree of excitement. He ran into a problem, which was that when he tried to work out the equations, he could get a universe that would be expanding, or he could get a universe that would be... Let's see, poor reception here, okay. Expanding, or he could get a universe that would be contracting, but he could not get a universe that would just sit still. And, as far as he knew, and as far as the astronomers that he knew would tell him, he thought we lived in a static universe. So he did something, which I think most of us as students have been tempted to do at some point or another in a physics course, which is, if you can't get the problem to work out, you might want to put in an extra little term into the equation to make it balance. When Einstein does it, he uses the Greek variable lambda and he calls it the cosmological constant so it sounds good, but if fact, it really was just a fudge factor in the equation, and it was not a very good solution. It would make a universe that would just barely balance when standing still, and it was only about a dozen years later that the astronomers measured the distances to what we now know are distant galaxies, and realized that we were living in a universe that really was expanding, and Einstein famously kicked himself and he called the introduction of that cosmological constant his greatest blunder. But, as you'll see, even Einstein's greatest blunder ends up being a fertile idea, which we'll explore today. Now, when I say that the universe is expanding, at that point, you should find yourself feeling somewhat disturbed. I was going to try and provide you a tool as you leave the meeting today and when you go home, talking to your friends and family, which is that this particular bit of very fundamental physics is one of the few examples I know of, of fundamental physics where, in principle, you can explain it to almost anybody, and I've tried it on airplanes, you know, when people have asked me what I work on, and so you should try this when you get home, and you'll find that there are only a few basic concepts that you'll really have to take a moment to explain. One of them is this one: the idea that the universe is expanding. And, I think most people find themselves disturbed by the idea, for good reason, because they've heard the term, "the big bang," and they picture an expanding universe as something of an explosion into empty space, and the empty space itself, to most people, would seem like that's part of the universe, too. What do we mean when we say that the universe is expanding? I'll try and give you a somewhat simplified picture of the universe, and, as you can see, it's perhaps oversimplified. Which is, I think, useful to think about in this situation. You're supposed to imagine now that we live in a universe that's just a series of galaxy after galaxy after galaxy, and these blue spots are supposed to represent galaxies. It's supposed to be infinite, which means that if I could draw better, you would see the dots going on forever in that direction and forever in that direction. They'd be going towards you forever, into the floor forever, and we're just positioned in this sea of infinite galaxies, and the only thing characteristic about this universe is that there's a typical distance between the galaxies. If you're traveling across the universe in a spaceship, every now and then, you would stop for lunch at the next galaxy, and that is, when we say that the universe is expanding, all we mean is that the characteristic distance between these galaxies in an infinite universe, just gets a little bit bigger. It's just as infinite as it was before, there's just as many galaxies as there were before, it's just that now, there's a little bit more space between the galaxies. If anything, if you ask where is the expansion happening, we're pumping extra space, extra vacuum between points in space, between the galaxies. So, with that picture in mind, you'll find it's actually, well, it's mind-boggling as well, but, if you just think about what the alternative is, it's the least mind-boggling of your options in expanding universe. Now, with this in mind, though, an idea that the universe is actually changing in this fundamental way, allows you to start asking about that original question of the fate of the universe in a more nuanced way because now you start asking: if the universe is changing and expanding, will it go on expanding, and will it continue at that same rate of expansion? You might imagine that all of these galaxies, all the mass in the universe would attract all the other mass and that would slow that expansion down, and, in fact, right after this discovery that the universe is expanding, the next question was: how much is the universe slowing down because of gravity? People wondered, could we make that measurement, and could we find out if it's slowing enough so that someday it could even come to a halt and then collapse, so that's a way to determine the fate of the universe in a more rigorous way. At the time, we did not have the tools to do that, although there was, already, the suggestion that perhaps we could use a exploding star as a tool, and this is saying that it was already known that there are very distant stars and distant, we now know, galaxies, and you can see them explode in a sudden burst of light, and then they fade away. They brighten in a few weeks, they fade away in a few months, and that one star when it explodes can be brighter than the entire galaxy of 100 billion stars in which it's seen. So, these are dramatically bright events, and therefore, you can see them across vast distances of the universe. In astronomy, the next idea that you might want to point out to the person when you start trying to explain these ideas is a very simple one, which is just the idea that light takes time to travel, and most people you talk to will know this, but they probably will not realize how long it takes for light to travel across the distances that we see in the astronomical world. So, our own sun is relatively near. The sun's light is traveling to us in about eight minutes. If the sun were to go out, we wouldn't know about it for about eight minutes, then we might notice. The nearest star is further away, so we travel around the sun four times, four years go by while the light travels to us from a distant star, so that's interesting. But, it becomes much more interesting when you look at the nearest galaxy of stars because now we're looking at light, when we see these images, these beautiful images of a distant galaxy, we see light that's been traveling to us for 150 thousand years, so the light that we see there from that distant galaxy left when, here on earth there was some of the first evidence for human culture. The nearest cluster of galaxies of stars is even dramatically further than that. There, the light's been traveling for 65 million years to get to us from these distant clusters, so the light we're seeing there left when, here on earth, the dinosaurs were going extinct. Yes, my extinct dinosaur. That's nothing compared to how far you can see these very, very bright exploding stars, so the most distant of these super nova that we've seen now, this light's been traveling to us for some 10 billion years. That's two-thirds of the age of the universe, so now we're really talking about a dramatic time, and that's really what you want if you want to study how the universe has been changing over time, how the expansion has been, we thought, slowing. So, with that in mind, the only other question that you might ask is, can we tell how far away those points are, and it turns out that the new tool that became available when we started doing this research in the late 1980's was a sub-type of supernova called the Type Ia supernova that you can recognize by its spectrum. As those became available, that meant that we found a tool where they are almost always the same brightness, we think that these are a triggered explosion, so it brightens to a certain brightness that's about the same every time and then fades away. Therefore, you can use the brightness of the peak of these supernovae as what we call a standard candle. You know how far away it is by how bright it appears. So, the fainter it is, the further away it is, and hence, the further back in time that particular supernova exploded. So each supernova's brightness tells you the date in which that particular explosion occurred. There's one other piece of information you'd like to know. You'd like to know how much has the universe expanded since that date? And, that's something that we can learn just from this very simple trick that the light from the supernova comes to us primarily blue. If you were able to stand next to a supernova and one exploded, it would look mostly blue. And then, the light travels to us in the expanding universe and anything that's not nailed down in an expanding universe stretches just with the universe, and that includes the very wavelength of light of that supernova that exploded, so by the time that it reaches us, it's been stretched to red, we call that the "redshift," and how much it's been stretched tells us how much the universe stretched. So, putting this together in a distant galaxy far away, some supernova explodes, and then its light, which is blue, starts to travel to us in an expanding universe, which is stretching, and the light's becoming redder and redder, and then we read off, when it reaches us, just exactly how much the universe has stretched since the time of that explosion that we know from its brightness. That's the entire measurement that you have to try when you get home, you can email me afterwards, tell me if it works. As you can see, this is a relatively simple concept. In principle, this is an easy measurement to make. In practice, however, it was very difficult at the time to see how to do this, and that was partly because we needed to be able to find these very, very distant exploding stars, and at that point, nobody had found a supernova far enough away to make any of these measurements. When we began working on this project, the first job was to figure out how to find these in a systematic, guaranteed way. It involved developing new technologies, new cameras that would allow you to look at thousands of galaxies, all at one time. This image here is a ten minute elapsed exposure from a 4 meter class telescope, this particular one in Chili. All of these little faint blue specks here in the background are what I want you to focus on. These are the distant galaxies in which we wanted to look for supernovae. The bright foreground things are nearby and were not interesting for this purpose. In those distant blue specks, we had to find a new spot of light that wasn't there before, and that, obviously, was a very challenging job, even your most patient graduate students couldn't do it by eye. But, it meant that we could take advantage of what was then becoming possible with the growth of computing power. What we were eventually able to do is, have the computer home in on a spot like this and show us that in a difference of three weeks, this spot is brighter than it was three weeks before, and if you subtract this image from this one, you're left with just the supernova light as the difference. So, this technique took a while to develop, and eventually we got to the point that you could guarantee these discoveries, and you could even eventually observe these from space so the Hubble Space Telescope came online around that time, and there you can see this little white speck is the supernova light that's shown against this galaxy, where from the ground they're all blurred together. Now, after we had developed this technique, we were ready to start using it and going out to the big telescopes around the world, we were observing once every semester, and every semester, we would get a half-dozen to a dozen of these galaxies, and we built up a sample of some 42 supernovae over about three years, and for those of you who've read The Hitchhiker's Guide to the Galaxy, 42 seems an auspicious number. Those who haven't will have to ask that later. The question then was, if we plotted those 42 supernovae on a plot of the expansion history of the universe, what would we see? Would we be on one of these curves here that would be decelerating and slowing forever, but expanding forever, shown here; or, was there enough mass in the universe to slow the universe to a halt and bring it to turn over and collapse into a big crunch at the end? Those were the options, and we were quite excited to actually start plotting data on a plot like this. So, nearby supernova, at the time, were not showing you the difference between these different lines at all, but they already were showing you that the brighter supernovae were closer, the further supernovae were further. That tells you how far back in time you were from today, shown in the bottom axis. The supernovae's color, the special features as they redshifted, told you how much the universe had stretched since that time, so this is the average distance of the galaxies, relative to today's average, and then we were ready to plot the 42 supernovae and see which of these curves we found ourselves on, and the answer turned out to be none of the above. It missed all the curves, so that was the surprise at the time, and apparently, this data is telling us that we live in a universe that's very different. Apparently, it was decelerating for the first half of its life down here, but then, in the recent half of the lifetime of the universe it's been accelerating, and as far as we know, it could go on accelerating forever. So, that's a dramatically different universe, and it was a big surprise that we were seeing. Actually, two different research groups were now racing to get this result in these last few years of the project, and they both were actually doing this analysis at Berkeley, one at the bottom of the hill in the astronomy department, the other at the top of the hill at Lawrence Berkeley laboratory, where my group was. I was leading one team, and Brian Schmitt, in Australia, was leading the other team, and these teams by this time were international, each of them had members from seven different countries around the world, and we were all struggling with the question of, "Is this possible, are we actually seeing a universe "that's expanding faster?" There was no reason to expect this. You assume gravity would be slowing. This has become one of the key questions, now, for our day in cosmology: What is going on? Why do we live in this universe? One possible explanation is that there is some new form of energy that we haven't earlier explored that is pervading all of space. We are calling it "Dark Energy," just to represent our ignorance of it, it's nothing to do with the color of the energy. The possibility is that most of the universe is in this form, it could be that as much as almost three quarters of the universe comes in the form of dark energy, and we've never seen it before. So, that would be fascinating if true. Now, it's also possible that we've just misunderstood something about how gravity works, and perhaps it will turn out to be that we need to modify Einstein's theory of general relativity, which, of course, would be, if possible even more dramatic since Einstein's theory of general relativity has been an amazingly successful tool in understanding gravity, and we can predict things to many digits of precision. It's remarkably mathematically symmetric, and it's very difficult to modify and preserve all these wonderful features that we know about general relativity. It's become a very exciting moment while we try and explore what's going on and how this is happening. Since that time when we saw this result, there has been paper after paper after paper from the theory community. In fact, at one point, I was estimating that there has been a paper published every 24 hours since the time of that discovery. If you ask any of these theorists, do they have the answer to what's going on, they will cheerfully tell you, At one point, we were thinking that one of the best things about them was they were coming up with wonderful names for these theories. I love things like "Big Rip Cosmology," and, "Ekpyrotic Universe," which you can barely pronounce. It's actually a nice theory, too. The theorists then say to us that, "No, the ball is back in your court "as the observers, you have to tell us some more And, this is not true; we have not just been sitting around as the observers, the experimentalists. We have been making progress trying to understand what's going on. The original plot of the expansion history of the universe that I showed with the two teams' data on it had some just 50 or so supernovae. If I flip this around and turn this upside down, the way the astronomers usually plot it, you can see that those original 50 supernovae has now morphed into over 700 supernovae in a recent dot plotation here. One that we're finishing up right now has well over 1,000 supernovae. Every single one of these supernovae is much better measured than the original 50 supernovae, so we're learning a lot about this expansion history of the universe, but when you ask, One of the properties we'd love to know is, how springy is this dark energy? How much does it make the universe want to expand faster? And, that's what's captured by this equation, the state variable, "W," which is a ratio of the pressure to density in Einstein's theory of general relativity. You find that you've very tight measurements only in relatively recent times, very low redshift, and as you get further and further away, the range of possibilities expands until almost any theory can fit in there, and we know very little about the properties of dark energy as you go back in time. Why is this so difficult? Well, it turns out that the measurements, although they've really sharpened this picture significantly, as you can see, as we go from the original 50 supernovae to today, we have a much better tracing of the history of the expansion of the universe. The problem is that all the theories that we have to compare predict very, very slightly different versions of this history and so slight are the differences, that they all fit within the thickness of that green line on the plot. So that means that we need some 20 times more precision in the measurements. The supernova community has gone back through every step in the story as we go from the original explosion of the supernova down through the astronomical route that it takes to us and through the properties of the detectors that we use, and each of them is getting improved in terms of our understanding and our ability to control the errors in the measurement. I'll describe just one of these today, just to give you a feel for the kind of thing that we're working on, and some work that I've personally been involved with lately. This is, in the very first step, when you ask, "Is it possible the supernova is changing its brightness What's useful is that the supernova, as it explodes, it sends you a spectrum that tells you a lot about all the elements and their physical states at any given moment. So, first the outer layers of this explosion of gas send you a spectrum, but then, over time, you see deeper and deeper and deeper in to this exploding ball of gas as the outer layers expand and become transparent. Then, different chemical and molecular, well, in this case, elemental features appear at different excitation states, different temperatures, and this allows you to, essentially, do a CAT Scan through the entire ball of gas, and you can now compare distant ones to nearby ones in great detail. We've been working on a project to then study relatively nearby supernovae in that degree of detail, and so these are time series of different supernovae, and what's interesting in these plots is that each of these plots shows, not one supernova, but two supernovae overlaid on top of each other at the same dates in their CAT Scan, the same dates in this explosion history. For the upper right one, for example, you can see the blue with the black time series, and a green time series, and you can see that those look almost identical, those two supernovae. If you look over at the top left, you see a red time series and a black time series, and they look identical, but those, if you overlaid them on top of each other are not the same as the ones on the far right. So now, we can treat these supernovae as cases where individual supernovae can be matched against individual other supernovae in great detail, and those become our new standard candles for this kind of work. This is one of the kind of improvements that we can make to get that extra factor of 20 that we need in the precision to learn more about dark energy. We cannot yet make these same precision measurements for very, very distant supernovae, and this is why we're now in the process of developing the next generation of, for example, space telescopes. There's a new project called WFIRST that can do these kinds of measurements, and it looks very likely that this will be approved by NASA within the next year. We think we're going to have a chance to do this measurement now, very soon from supernovae. And, I should also mention that there's other techniques, now, finally coming on the horizon. Just in the last few years, we finally have another method for making the measurement of the expansion history that's actually been published with the result that this technique called baryon acoustic oscillation, it's abbreviated BAO, and it measures distances away from the early times, around the time of the glow left over from the big bang, the cosmic microwave background, and moves its way in using the fact that there's over-dense and under-dense regions that we see back then, which eventually turn out to be where the over-dense and under-dense regions are today that we call galaxies and clusters of galaxies. You can actually follow those distances forward in time, just like the supernova allows you to work backwards in time from relatively nearby supernovae. So, those techniques are coming in, there's also another technique called weak lensing, and we are hoping that we'll be able to start taking advantage of all three together to pin down this question, and that we have a fighting chance to really get to the bottom of this fascinating question of, what is actually causing the acceleration? What kinds of dark energy or changes in Einstein's equation of relativity? This is an unusual period, I think, in our history, in this area because we're about to move a number of these projects to the next level of precision, and I think we're still at the stage in this field of cosmology where, every time we learn how to make a much more detailed measurement of this history of this universe that we live in, we've found surprises, and I think the bets are, probably, that we will continue to find surprises. So, we're very much looking forward to this next series. Coming back to my title, what can I say that we've learned about the process of learning while we were doing this study? I want to just point out a couple of aspects of this work. First, there's the fact that if you take a really good challenging problem, not every challenging problem is doable in a few month period, and this was a very good example of one. We thought this problem was going to be very hard, we thought that it would take us three years, in our original proposals, to get it done. In fact, after three years, we had not found our first distant supernova; it took five years to do that, six years to demonstrate the technique, nine years to collect the data, and in the tenth year is when we actually saw this result. I don't think any of us were disappointed in that time frame because we were learning things all along the way. I think a really good problem is well worth that kind of stick-to-it-ness. It takes many iterations, you get things wrong, you figure out better ways to do it right, and this is something which you will find over the years: that it's not just a matter of your having that confidence and that ability to follow through on a problem that you see progress in, but it's a matter of also being able to convince the funding agencies and the larger world around that they should know how to have patience and be willing to stick to problems that are really worth sticking to over a longer period of time. So I think this was an important point, here. Another that I think is worth pointing out is that you, in this kind of work, the really important element for a lot of it, understanding that you are looking for how you're making mistakes, you're looking for how is your big theory potentially wrong? In this case, it turned out something very deep about the universe that we had not quite gotten right was in play. And then, on the small, how every measurement you're making could be wrong, and that you're spending, I would say, or your measurements are fooling you, or your measurements have some mistake in them. So, when people sometimes ask, "What did you set out to prove "when you worked on this project?" I always feel like, no, that's really not the way that my experience with science has been. It's really been setting out to be open to finding what is wrong about your current understanding, and being able to tell whether you trust the current measurements and the current results you're making. So, I think that was an important element of the story. And then, I want to point out that there's this aspect of all this work that is highly social, in my experience. Some of this is the groups that you're working with, and these projects depended on these amazing teams of scientists who were capable of working together over time. This is also not an easy aspect of doing science, but it's also the fun aspect of doing science. It's an activity that really can bring people together in ways that you would never expect, and doing that, figuring out how to get groups of people to work well together, and how to make sure that people can all have great careers coming out of it, almost requires the same ingenuity as the science itself does. I think it's an important part of the story to remember, as well. With that in mind, I find myself thinking that, just to finish this talk, I should mention that I think that transmitting the culture of looking for our error, not trying to prove yourself right, looking for how you work with groups of people in ways that's productive, all these things, I think, are worth teaching, not just to the next generation of scientists, but actually, to almost everybody. As part of almost what it means to be a critical thinker, and I've been wondering, can we develop ways to teach this? And, I want to finish with a couple of examples of ways to think about that. I want to ask this group about, over the week, if we have time: what would be the list of things that you would want to teach such that our society is better able to make decisions and solve problems where it often feels like they don't know how to talk to each other and to address questions together? Where does our authority come from in science, and why should anybody believe it? I think it's coming from the fact that we've developed these collections of heuristics and tricks that allow us to side-step our various mental weaknesses and play to our strengths, and often, looking for where we're wrong and where one's fooling ourselves, and I've been trying out a set of ideas that I've categorized under some of these things have to do with the fact that we think probabilistically, that we're, not that we're 100% sure of things, but we figured out how to quantify how sure we are and how not sure we are. We have a lot of elements of our thinking that are very skeptical that come along with that statistical thinking, and yet, we also have to balance that skepticism with that "can-do" spirit that allows us to stick with a problem for many years. And, of course, this whole aspect of how the groups work together, I think, becomes an important part of that story. I'll just mention that, if we have time this week, I would love to talk about these, trying out these different ideas of what could we teach that might be useful for people to learn when they do science, but also when they ask other questions as a society, and I have no idea whether this is the comprehensive question, the set of topics, but it will be the kind of thing that will be interesting to talk to people and find out, in different fields, what they would consider to be what they would want to teach that's not content, but how do we think as scientists. Let me leave you, then, just with that thought that I think we live in a very exciting time, both for the sciences where we can actually, I think, get at some very deep fundamental questions as we develop these next generations of tools. We have a fighting chance, I think, to understand why it is that we live in a universe that's accelerating and maybe even in our own generation, we'll get a chance to learn the answer to that. Maybe it will tell us something very deep about how the four forces of the universe, gravity and the other three forces fit together. And while we're doing it, I hope we also learn more about how we do critical thinking as scientists so that we can make a difference in the world in many ways beyond the science that we're doing at that moment. Let me stop there.

Saul Perlmutter (2015)

What We Learn When We Learn that the Universe is Accelerating

Saul Perlmutter (2015)

What We Learn When We Learn that the Universe is Accelerating

Abstract

The 1998 discovery that the universe's expansion is accelerating was not only unexpected, but it also led to the postulation of a previously-unknown “dark energy” forming almost three-quarters of the "stuff" of the universe. How was this discovery made? What has been the progress since in understanding this dark energy and the accelerating universe? And what does this tell us about how science works.

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