Skip to main content

Welcome to the new Annenberg Learner website! All of the current series have migrated to our new, streamlined interface. The legacy site is available at through January 31, 2020.


Physics for the 21st Century

Dark Energy Interview with Featured Scientist Robert P. Kirshner

Interviewer: What do you do in your research?

ROBERT: We set out to find out how much the galaxies had slowed down over time. And, of course, to our surprise, we found that they hadn’t slowed down, but instead were speeding up. So the question is: what is doing that? Is it the dark energy, this thing that makes the universe speed up?

We’re trying to measure the history of cosmic expansion. We think that the universe began at a Big Bang, and was expanding. That the expansion slowed down over time due to gravity. But then about 5 billion years ago this other thing, this dark energy seemed to get the upper hand, and it’s making the universe expand faster and faster now. And we’ve actually measured that. We’ve measured the part where the universe is expanding faster, and we’ve seen back to the part where it was slowing down. We’re trying to make measurements that are accurate enough so we could tell whether the expansion history is the one you predict, given a constant dark energy, or whether it’s doing something different.

Interviewer: What is: “Dark Energy”?

ROBERT: Well, there are a lot of ideas about what the dark energy might be. Maybe the most popular one, or the best-founded one is the oldest one. Einstein, it turns out, had an idea that the contraction of the universe would be balanced out by a kind of pressure, sort of pushing out of the universe. He thought the universe was static, though, so he thought it was balanced before astronomers had shown, back in the 1920s, that the universe really is expanding, really getting bigger. Once people showed that the universe was expanding, Einstein stopped talking about this cosmological constant, this antigravity stuff, and people thought about a universe in which there was a big bang, and the galaxies started in expansion, and that they have just been slowing down, due to gravity, over time.

The picture we have for what gravity is is that it’s universal, that it’s a force that operates between everything, small particles, medium-sized particles, things the size of the earth, planets, stars, galaxies. All those things pull on one another through gravity, and that ought to slow down cosmic expansion over time. But what we see is something different. That even though all that pulling is going on, there’s something that is pushing, or, anyway, making the expansion speed up. That’s what we call the dark energy.

Interviewer: What is the Big Bang?

ROBERT: The idea is that the universe as we know it began about 14 billion years ago in a hot, dense state. Everywhere in the universe was hot, and much denser than it is now. It was in this unimaginable, almost incomprehensible moment that the expansion of the universe began. Now we’re pretty sure that this is about right, because we see a lot of clues—things that are left over from the Big Bang.

We see that the galaxies are moving apart from one another. No matter where you are in the universe, it’s as if things are moving away from you. We see that there are elements that were cooked in the Big Bang. There’s helium that we think was synthesized in the Big Bang itself. And we also see a kind of glow of the sky, no matter what direction we look, that is, we think, the leftover heat from the Big Bang itself.

So there’s an expanding thing, because we see the expansion. So we think that we understand, at least in the big outline, that 14 billion years ago the universe was a very hot, dense place, and that the expansion that we see today had its origin back then.

Interviewer: What would you expect to see following the Big Bang?

ROBERT: The picture that people had was that there would be expansion from the hot Big Bang, and then the effect of mass in the universe would be to slow things down. So as time goes by the universe would keep expanding for a while, but if gravity was in charge, you could have the case where there would be a contraction at some future time, or you could have continued expansion. But no matter what happened, you would have always slowing down. We thought we could measure that effect. But when we actually went to do it we found that it wasn’t slowing down, that it was speeding up. So you need something else. And that something else that we talk about is the dark energy.

So you throw up a ball, and it comes back down. The effect that we’re seeing is something like if you throw up a ball, and it goes up, and then it keeps going up, and it goes up and up faster and faster. That would be a very surprising result. So while it isn’t true that the gravity that you learn about here on the earth is the whole story, it is true that the basics of it, that massive things attract one another, is something that you do see here on the earth, and that we think operates on the big scale too.

Stars pull on each other, the galaxies pull on each other. And in the whole universe, all of that stuff is pulling on each other, trying to help it slow down. So gravity is something which is attractive, and making things slow down. It’s very surprising, just like it would be surprising to throw a ball up, and see it fly up into the sky, to see this other thing, this sort of antigravity-like quality that the dark energy has.

Interviewer: What data enable you to state that the universe is accelerating?

ROBERT: Well, the way we make these measurements is to look at exploding stars, supernovae. There’s a certain kind of supernova, which we can recognize, that seems to have fairly nearly the same brightness. They’re not all exactly the same, and we have to make other measurements that allow us to tell which are a little bit brighter, and which are a little bit dimmer, but those are details. If they were all the same brightness intrinsically, then you could tell how far away they are, or were, from how bright they appear.

So the idea is if the supernovae are all the same brightness, then you can figure out how far away they are from how bright they appear. And we’ve done that measurement. Searched for supernovae, we’ve measured how bright they are. And we’ve also measured the red shift, the expansion of the universe that’s taking place at the site where the supernova exploded. This allows us to construct the history of cosmic expansion. That is, how fast the universe was expanding at each time. So we can make a graph of how the universe expands with time. We can find out whether it’s slowing down or speeding up.

And what we found is that over the last 5 billion years, the supernovae appeared a little dimmer than they otherwise would have. And that means that the distance that the light had to travel is a little bit bigger than it otherwise would have been. And that means that the universe has been expanding faster over the time that light was flying to us over the past 5 billion years than it was in the past, or than we had expected. So that means there’s something else going on. The universe is speeding up over time.

Interviewer: How did you discover that the universe is accelerating?

ROBERT: We’ve been measuring these supernovae, and it was kind of slow going at first. We only had a handful. Adam Reiss was working up the data, and I remember Adam called me up and he said, “You know, I keep finding something funny. I keep finding that the mass comes out negative.” I said, “Well, Adam, you must have made a mistake, because the mass is not negative.” He said, “Well, it’s as if things are accelerating.” I said, “Yes, let’s not go there.” Because I knew that that was leading towards this cosmological constant, which it had a terrible smell for a long time. After the discovery of the expanding universe, Einstein swore off this cosmological constant.

The cosmological constant had a terrible reputation, and I did not want to go into that territory. And I remember writing an email to Adam, and I said, “Adam, in your heart you know this is wrong.” But he said, “Well, never mind my heart, you know, and never mind what you think. Here’s the data.” And it was true that if the data were right, that instead of seeing the slowing down of the universe due to gravity, we were seeing the opposite. We were seeing the speeding up. So that was very surprising. That was right toward the end of 1997. And we had a little bit of data, and we had done the analysis. And I thought this is really a big deal. If it’s right, it means that there’s something going on, which is very important. It’s a cosmological constant. Dark energy. But if it’s wrong, it’s going to be really embarrassing. And so, you know, you have to balance out your desire to be first with a new result, and your desire to avoid embarrassing yourself. And, you know, you have to weigh those, and different people would weigh them in different ways.

But we redid the analysis, or anyway we confirmed the analysis. Brian Schmidt did it independently in Australia. And together this whole group of us checked that we had a more or less reliable result before we went public, and said to people, “You know, we think we see a universe that’s accelerating.” But we finally got up the courage to do it. The data were showing us this signature.

Well it would have been irresponsible not to tell people what’s going on in the universe. But also there was another group that was working on the same problem. And they clearly had data, which was going to show the same thing, and they had showed some of their data in public places. They were kind of worried about whether they should say anything, or—because there are other things that could cause the sort of effect we see. If there’s a lot of dust along the line of sight that’s absorbing some of the light from the distant supernovae that could mimic the effect of cosmic acceleration. So you had to be able to figure out how much dust there was, which we had worked on a technique to do. So we were quite confident that the supernovae were really dimmed by this extra expansion, and not by dirt from here to the next galaxy.

Interviewer: You use supernovae for your measurements. How do use data gathered from them?

ROBERT: The idea is that the supernovae, although they’re very bright—we can see them halfway across the universe, they’re very rare. There’s about one supernova in a hundred years in a galaxy like ours. So that means, you know, if there are 52 weeks, that would be like 5,200 a galaxy you’d have to look at to see one that was kind of within a week of explosion, which was sort of what we wanted. So you have to have thousands, tens of thousands, really, of galaxies that you’re studying each night. You have to process all that data during the night, or during the next day so that you can find these things. Because after a few days—a supernova is kind of like a fish, you know. After a few days it’s no good anymore, and you can’t make the measurements in an accurate way.

Well, the measurement is a measurement of the brightness of supernovae as a function of their distance. And the idea is that you need to identify those supernovae, you have to do some tricks, because they’re not all exactly the same brightness, so you have to recognize which ones are a little bit brighter, and which ones are a little dimmer. It turns out you do that from the shape of their light curve, the time it takes to get bright and dim. And you need to make corrections for dust along the line of sight, because we’re trying to judge the distance of things by their brightness. If there’s something else that’s affecting how bright they look, then we could make a mistake.

Interviewer: How do you find Supernovae?

ROBERT: The basic idea is you have to find distance supernovae so that you can measure how far away they are. And to find them you need to take a big chunk of the sky. Make a very, very deep image that lets you see very distant objects. These are digital images. So you put all of that into a computer. You come back the next month and subtract one picture from the other. The idea is that anything that’s new in the second picture will be left over. Anything that stayed the same will just get subtracted away. Well, it’s easier to say than to do, like a lot of these things. And getting the computer to do the right thing to the data is something that people on our teamwork very, very hard to make so that it worked well. But it did work well, and it turns out with the tools that were available—four meter telescopes in Chile, the charge-coupled devices, the electronic cameras that were available at those telescopes, and the computers that we had ten years ago were able to make those subtractions, find the new supernova, and then go measure it night after night. If you didn’t find it overnight, and get right on it, you would lose the opportunity to measure it properly. Because the supernovae get bright and get dim. That takes about a month.

And so if you can catch one on the way up, and make the measurements you can get the shape of the light curve, you can get the brightness at peak, you can measure the red shift. All the things that we needed to do. But if you take the data home, and kind of work on it, in the way you do for a lot of astronomical problems, it’s too late. So the supernova game was something that required using computers to make these incredible sifting of the data, looking for the one new dot in this picture, with thousands of galaxies.

Interviewer: Do all physicists agree that there is a dark energy?

ROBERT: Now people are pretty well agreed that there is a dark energy. Something’s making the universe speed up. And there are lines, many lines of evidence that converge on this picture. So that’s good. But if somebody asks you, “Okay, what is the dark energy?” that’s when people start to wave their hands, and to kind of not really know for sure. There are several ideas, none of which seems particularly compelling, but anyway, there are several ideas.

Interviewer: What are your biggest challenges?

ROBERT: It seems like we ought to try to do better, and I think we will. I think we will—we can make the samples bigger. That’s the easy part. The hard part is making the measurement of the distance to each one more reliable. And the thing we’re having the most trouble with is the absorption by dust. The same thing that we started out having trouble with ten years ago. We have some new tools to do that better. I hope that when we measure the light at longer wavelengths, using infrared, we’re going to be able to make more precise distance measurements, and from that, make more precise inferences about what the dark energy is.

So I think that’s the path forward. There’s a very lively discussion going on now about building a specialized satellite, which would have, as one of its jobs, measuring the distant supernovae in the infrared, with very high precision, so that we could map out the history of cosmic expansion, and then compare that with other lines of evidence, and see if we can really begin to pin down what the dark energy is.

Interviewer: What is the future of dark energy research?

ROBERT: What we need to do for the next little while is to study more supernovae, build up those samples, at different red shifts. So we need a big chunk of the sky, we need some nearby, some middle distance, some far away. And a very good understanding of what the supernovae are, and how they change with time, so that we’re not misled. We have to understand what the dust is. We have to do all that stuff. And that will be fun, because we’ll find out more about how the stars evolve, and how the explosions take place.

Sometimes the observations get ahead of the theory. Sometimes the theory is ahead of the observations so it isn’t exactly like a textbook where somebody says, “Well I have an idea and now I’ll do an experiment to test my model.” Well, in fact, there’s only one universe that we’re in and so you can’t always do an experiment. Secondly, in this case we’re making the measurements and we thought we would see deceleration but we saw the opposite. And, so it’s something where the observations, the discovery, the measurement is way ahead of our current ideas. But what is missing is a really good idea. So we’re hoping that someone will come up with that.

What I would say is there’s an incredible opportunity here. It means there’s some very, very big idea about gravity that affects the whole universe whose properties we are beginning to see and maybe can learn enough about to get a clue but that we don’t yet understand. So this is really a great thing. I mean the most important thing is our ignorance after all. We have a vast reservoir of that and this is a really important problem that we know that we don’t understand. We have some ideas about how to make progress on it at least observationally by experiment. But we need a good idea. We need a good theory. And, at the moment, there are some ideas out there but it’s not clear that any one of them is better than the others.

This is an incredible adventure. We went out to study supernovae. We thought we were going to see the deceleration of the universe and instead we found this other thing. We found that the universe is not doing what we thought it was doing. It’s speeding up. We attribute to dark energy which turns out to be ¾ of the stuff in the universe and we don’t know what it is. So it seems to me that’s something we ought to put our energy into solving—that we really ought to try to find out what the dark energy is.



Series Directory

Physics for the 21st Century


Produced by the Harvard-Smithsonian Center for Astrophysics Science Media Group in association with the Harvard University Department of Physics. 2010.
  • Closed Captioning
  • ISBN: 1-57680-891-2